0001
 1   
 2   
 3   
 4   
 5   
 6   ******************************************************
 7   
 8                  TEXAS LOTTERY COMMISSION
 9                           MEETING
10   
11                      JANUARY 13, 2006
12   
13   
14    ******************************************************
15   
16   
17   
18            BE IT REMEMBERED that the Texas Lottery
19   Commission meeting was held on the 13th day of January
20   2006 from 8:30 a.m. to 5:44 p.m., before David Bateman,
21   RPR, CSR in and for the State of Texas, reported by
22   machine shorthand, at the Offices of the Texas Lottery
23   Commission, 611 East 6th Street, Austin, Texas 78701,
24   whereupon the following proceedings were had:
25   
0002
 1                    A P P E A R A N C E S
 2   
 3   Chairman:
 4            Mr. C. Tom Clowe, Jr.
 5   
 6   Commissioners:
 7            Mr. James A. Cox, Jr.
 8   
 9   
10   General Counsel:
11            Ms. Kimberly L. Kiplin
12   
13   Acting Executive Director:
14            Mr. Gary Grief
15   
16   Charitable Bingo Operations Director:
17            Mr. Billy Atkins
18   
19   
20   
21   
22   
23   
24   
25   
0003
 1                          I N D E X
 2                                                    PAGE
 3   Appearances......................................   2
 4   
 5   AGENDA ITEMS
 6   Item Number I....................................   6
         Meeting called to order
 7   
     Item Number II................................... 101
 8       Report, possible discussion and/or
         action on 3rd quarter bingo conductor information
 9   
     Item Number III.................................. 110
10       Consideration of a petition for rulemaking
         to amend 16 TAC 402.600 pertaining to
11       Bingo Reports
12   Item Number IV................................... 118
         Report, possible discussion and/or action on
13       lottery sales and revenue, game performance,
         instant game closures, new game opportunities,
14       including Powerball, and trends
15   Item Number V.................................... 133
         Report, possible discussion and/or action on
16       the agency's financial status
17   Item Number VI................................... 135
         Report, possible discussion and/or action on
18       the selection of jackpot payment option for
         prize winners
19   
     Item Number VII.................................. 156
20       Report, possible discussion and/or action
         on the Mega Millions game and/or contract
21   
     Item Number VIII.................................  15
22       Report, possible discussion and/or action
         on Gtech Corporation
23   
     Item Number IX...................................  54
24       Report, possible discussion and/or action on the
         lottery terminal functionality, including quick
25       pick feature for all games and/or impact to players
0004
 1   Item Number X.................................... 157
         Report, possible discussion and/or action on
 2       The lottery advertising and promotions
 3   Item Number XI................................... 164
         Report, possible discussion and/or action on
 4       HUB and/or minority business participation,
         including the agency's mentor/protege program
 5   
     Item Number XII.................................. 166
 6       Report, possible discussion and/or action on
         the agency's contracts
 7   
     Item Number XIII.................................  34
 8       Report, possible discussion and/or action
         regarding the demographic study of Texas Lottery
 9       players
10   Item Number XIV.................................. 167
         Report, possible discussion and/or action on
11       the 79th Legislature
12   Item Number XV...................................  93
         Consideration of and possible discussion and/or
13       action on external and internal audits and/or
         reviews relating to the Texas Lottery Commission
14       and/or the Internal Audit Department activities
15   Item Number XVI.................................. 168
         Consideration of and possible discussion
16       and/or action on the appointment and
         employment of an executive director
17   
     Item Number XVII........................... 6, 14, 90
18       Executive session
19   Item Number XVIII.......................... 7, 15, 93
         Return to open session
20   
     Item Number XIX.................................. 176
21       Consideration of and possible entry of orders
22   Item Number XX................................... 197
         Report by acting executive director
23   
     Item Number XXI............................  197, 201
24       Report by Charitable Bingo Operations
25   
0005
 1   Item Number XXII................................. 202
         Public comment
 2   
     Item number XXIII................................ 210
 3       Adjournment
 4   
     Reporter's Certificate........................... 219
 5   
 6   
 7   
 8   
 9   
10   
11   
12   
13   
14   
15   
16   
17   
18   
19   
20   
21   
22   
23   
24   
25   
0006
 1                      AGENDA ITEM NO. I
 2                 CHAIR CLOWE:  The Texas Lottery
 3   Commission will come to order.  Commissioner Cox is
 4   here.  My name is Tom Clowe.  This is January the 13th,
 5   2005 (sic).  It's 8:30 a.m.
 6                    AGENDA ITEM NO. XVII
 7                 CHAIR CLOWE:  Commissioner Cox, with your
 8   approval, I would like to make a motion that we go into
 9   executive session.
10                 COMMISSIONER COX:  Certainly.
11                 CHAIR CLOWE:  At this time, I move the
12   Texas Lottery Commission go into executive session to
13   deliberate the appointment, employment, evaluation,
14   and/or duties of the executive director, acting
15   executive director, and/or deputy executive director
16   pursuant to Section 551.074 of the Texas Government
17   Code.
18                 Is there a second?
19                 COMMISSIONER COX:  Second.
20                 CHAIR CLOWE:  All in favor, please say
21   "aye."
22                 COMMISSIONER COX:  Aye.
23                 CHAIR CLOWE:  Aye.
24                 The vote is two-zero.
25                 The Texas Lottery Commission will go into
0007
 1   executive session.  The time is 8:31 a.m.  Today is
 2   January the 13th, 2006.
 3                 (Executive Session from
 4                  8:29 a.m. to 10:52 a.m.)
 5                   AGENDA ITEM NO. XVIII
 6                 CHAIR CLOWE:  The Texas Lottery
 7   Commission is out of executive session.  The time is
 8   10:53 a.m.
 9                 Is there any action to be taken as a
10   result of executive session?
11                 COMMISSIONER COX:  Mr. Chairman, I have a
12   motion.
13                 CHAIR CLOWE:  Yes, sir.
14                 COMMISSIONER COX:  I move that Gary
15   Grief, effective January 17, 2006, be returned to his
16   position as deputy director and his annual salary be
17   set at the level it was before he moved into the acting
18   executive director position in January of 2005 but with
19   the four-percent legislative increase so that his
20   annual salary is set at $118,116, which is $9,843
21   monthly.
22                 CHAIR CLOWE:  I second that motion.
23                 All in favor, please say "aye."
24                 COMMISSIONER COX:  Aye.
25                 CHAIR CLOWE:  Aye.
0008
 1                 The vote is two-zero.
 2                 Is there another motion, Commissioner
 3   Cox?
 4                 COMMISSIONER COX:  Yes, I have a further
 5   motion.
 6                 I move that the evaluation of Gary Grief
 7   for the period February 28th, 2005 until January 13,
 8   2006 as acting executive director and as the deputy
 9   director be approved.
10                 CHAIR CLOWE:  I second that motion.
11                 All in favor, please say "aye."
12                 COMMISSIONER COX:  Aye.
13                 CHAIR CLOWE:  Aye.
14                 The vote is two-zero in favor.
15                 Is there a further motion, Commissioner
16   Cox?
17                 COMMISSIONER COX:  Finally, based on his
18   job performance and productivity, I move to approve the
19   promotion for Gary Grief effective Wednesday, January
20   18, 2006.  He would be promoted to a director five,
21   class number 1624, salary group 821, with his annual
22   salary to be set at $125,088.
23                 CHAIR CLOWE:  I second the motion.
24                 All in favor, please say "aye."
25                 Aye.
0009
 1                 COMMISSIONER COX:  Aye.
 2                 CHAIR CLOWE:  The vote is two-zero.
 3                 Now I have one.  I make the motion that
 4   Anthony Sadberry, effective Tuesday, January the 17th,
 5   2006, be named acting executive director of the Texas
 6   Lottery Commission with his annual salary to be set at
 7   $114,996.
 8                 Is there a second?
 9                 COMMISSIONER COX:  Second.
10                 CHAIR CLOWE:  All in favor, please say
11   "aye."
12                 COMMISSIONER COX:  Aye.
13                 CHAIR CLOWE:  Aye.
14                 The vote is two-zero in favor.
15                 I'm going to call on Mr. Sadberry.  But
16   before Commissioner Cox and I do that, I would like to
17   state on the record that we are very pleased to be able
18   to offer this position to Mr. Sadberry.
19                 As many of you know, he served with
20   distinction as a commissioner of this agency for a
21   number of years and during a time that there were a
22   great many problems that this agency dealt with.
23                 I have had the pleasure of serving with
24   Mr. Sadberry as a commissioner.  And I admired his work
25   in that role and appreciated his service to the state
0010
 1   of Texas from a firsthand observation point.
 2                 I would like to say that I approached Mr.
 3   Sadberry and asked him if he would be interested in
 4   this position as acting executive director.
 5                 He responded positively.  The
 6   commissioners have visited with him in executive
 7   session and have now just voted to ask him to assume
 8   this role.
 9                 Mr. Sadberry brings to this position,
10   assuming that he accepts this role, a history, a
11   knowledge and experience of dealing with the lottery
12   that is rich and I think unexcelled.
13                 He has served with distinction on the
14   executive search committee that continues to function.
15   And most recently, he chaired the last meeting and
16   brought us to a point where we reached last month and
17   will continue in the future.  My understanding is that
18   he will no longer serve on that executive search
19   committee upon taking up these duties.
20                 Now Gary Grief was limited, because he is
21   an employee of this agency, to a term of six months as
22   acting executive director.  Because Mr. Sadberry was
23   not an employee of this agency, if he accepts this
24   role, he will not be limited to six months.
25                 That's a question that the commissioners
0011
 1   would like to answer before it is asked.  At the same
 2   time, I want to state that the search for the executive
 3   director continues.  We expect to hear from Diane
 4   Morris shortly with a report on those activities.  And
 5   she'll bring us up to date on the level of activity
 6   there.
 7                 Having said that, do you have anything to
 8   add, Commissioner Cox?
 9                 COMMISSIONER COX:  No, sir.  I do not.
10                 CHAIR CLOWE:  Mr. Sadberry, are you here?
11                 MR. SADBERRY:  I am, Chairman.
12                 CHAIR CLOWE:  As the man who hopefully
13   came to work on Friday the 13th -- really it will be
14   the 17th -- I'd like to ask you if you would accept
15   this position and give us any remarks that you may
16   have.
17                 MR. SADBERRY:  Chairman Clowe and
18   Commissioner Cox, I -- after -- Chairman Clowe and
19   Commissioner Cox, after due deliberation that has gone
20   on for some time and our discussion this morning, I do
21   gladly and willingly accept this position on the terms
22   in which you have laid out in your motion, that is as
23   acting director, executive director, as your search
24   committee, in my absence now, continues to function in
25   the process that you set in motion to look for an
0012
 1   executive director to serve this agency.
 2                 And you correctly state that I was
 3   requested to consider it.  I am honored by that
 4   request.  I'm honored by the confidence you have
 5   expressed in me in your motion and your vote this
 6   morning.
 7                 You're correct.  I sometimes try to add
 8   it up.  I believe -- the record will certainly show,
 9   but I believe it was some eight years or thereabouts
10   that I served as a commissioner.  I was one of the
11   initial three commissioners appointed in 1993 when this
12   agency began its initiation as an independent and new
13   agency in the state of Texas.
14                 Every day since that time and up to this
15   time, including the time that I was away from the
16   agency, I've had a strong feeling for this agency, for
17   the -- and understanding of the very important role
18   that it serves in this state and that utmost we all
19   must realize and appreciate is that we serve the state
20   of Texas and the people of Texas.
21                 And that's where, in coming to this
22   position, I acknowledge and accept the responsibility
23   of doing that and bringing to this effort the best that
24   I can bring and offer to this agency.
25                 As you know and as we've discussed, my
0013
 1   philosophy is a team approach.  I think the most can be
 2   gained from that by approaching leadership and
 3   functioning of an agency in that regard.
 4                 So I look forward to working with the
 5   leadership, management, administrators and staff of
 6   this agency and with you, as commissioners, and with
 7   the leadership of our state.  I look forward to it.
 8                 I look forward to the -- the tasks and
 9   responsibility and challenges and the opportunities
10   that will be brought by this effort.  I look forward to
11   going to work and working each day to reach our goals
12   to make this agency continue in the excellence which I
13   believe has been achieved.
14                 Thank you for your services,
15   Commissioners.  I understand what that entails and
16   appreciate where you've brought this agency from the
17   point at which I left it as commissioner and I join it
18   again at a different capacity equal of importance and I
19   very much look forward to it.
20                 So again, I accept.  And thank you for
21   the opportunity.
22                 CHAIR CLOWE:  Thank you, sir.
23                 COMMISSIONER COX:  Thank you.
24                 CHAIR CLOWE:  Anything further,
25   Commissioner?
0014
 1                 COMMISSIONER COX:  No, sir.
 2                 CHAIR CLOWE:  We will give Mr. Sadberry
 3   an opportunity to speak with those of you who are in
 4   the room that want to visit with him.
 5                 But in regard to additional business,
 6   Commissioner Cox, with your approval, I'd like to make
 7   a motion that we go back into executive session.
 8                 COMMISSIONER COX:  Absolutely.
 9                   AGENDA ITEM NO. XVII
10                 CHAIR CLOWE:  At this time, I move the
11   Texas Lottery Commission go into executive session to
12   receive legal advice to -- pursuant to Section
13   551.071(2) of the Texas Government Code regarding the
14   lottery operator contract.
15                 Is there a second?
16                 COMMISSIONER COX:  Second.
17                 CHAIR CLOWE:  All in favor, please say
18   "aye."
19                 COMMISSIONER COX:  Aye.
20                 CHAIR CLOWE:  Aye.
21                 The vote is two-zero.
22                 The Texas Lottery Commission will go into
23   the executive session.  The time is 11:02 a.m.  The
24   date is January the 13th, 2006.
25                 (Executive Session from
0015
 1                  11:02 a.m. to 11:59 a.m.)
 2                   AGENDA ITEM NO. XVIII
 3                 CHAIR CLOWE:  The Texas Lottery
 4   Commission is out of executive session.  The time is 12
 5   noon.  Is there any action to be taken as a result of
 6   executive session?
 7                    AGENDA ITEM NO. VIII
 8                 CHAIR CLOWE:  And if not, let's move on
 9   to item eight, report and possible discussion and/or
10   action on Gtech Corporation.
11                 Mr. Grief?
12                 MR. GRIEF:  Commissioners, there have
13   been additional developments regarding the potential
14   stock ownership change for Gtech Corporation that's
15   been mentioned previously.
16                 On January 10th, Gtech issued a press
17   release in which they announced that Gtech and
18   Lottomatica, an Italian-based company, entered into an
19   agreement for Lottomatica to acquire all the
20   outstanding shares of Gtech stock for 35 dollars per
21   share or approximately 4.8 billion dollars.
22                 That press release also said that the
23   transaction is expected to be completed by mid-2006.
24   And the press release went on to say that Lottomatica
25   is headquartered in Rome, Italy, operates the Italian
0016
 1   Lotto, and is a market leader in the Italian gaming
 2   industry.
 3                 Our understanding is that Lottomatica is
 4   controlled by De Agostini, an Italian diversified
 5   industrial and financial holding group with significant
 6   international operations and De Agostini controls 58
 7   percent of the Lottomatica shares.  My understanding
 8   from Gtech is that Lottomatica shares are traded on the
 9   Milan Stock Exchange.
10                 As you know, with the approval and
11   assistance from the Office of Attorney General, the
12   Lottery Commission previously issued a request for
13   proposals for outside counsel services related to this
14   transaction.
15                 And this RFP covers a broad range of
16   legal services related to assisting the commission and
17   determining the appropriateness of the transaction and
18   the suitability of the new shareholders.
19                 An apparent successful proposer, Graves
20   Dougherty, a Texas-based law firm, has been announced.
21   Contract negotiations ensued shortly thereafter and the
22   contract has been executed by both parties.  My
23   understanding is that that contract has now been
24   forwarded over to the Office of Attorney General for
25   their final approval.
0017
 1                 We do have representatives from Gtech
 2   with us in the audience today.  And in the event that
 3   you have additional questions of them and I'm sure
 4   that, if I've said anything inaccurate, they would be
 5   happy to correct me.
 6                 CHAIR CLOWE:  Thank you, Gary.
 7   Gentlemen, will you come forward, please?  I want to
 8   ask you to identify yourselves for the record.
 9                 But before you do that and the commission
10   hears from you, I spoke to you this morning when we
11   began this meeting and informed you that we would call
12   on you as early as possible.  And in light of
13   misleading you, I need to offer you an apology.
14                 We understand that you are busy
15   executives and you have commitments.  And we are
16   appreciative that you would come to Austin and appear
17   before us.  And at the same time, we are sorry that we
18   have been unable to call on you before this hour of the
19   day.
20                 We've had a number of matters that have
21   been pending before this commission for some time.  And
22   it was absolutely necessary to deal with those and,
23   therefore, the delay in coming to this item of the
24   agenda.  But I want you to understand it was clearly
25   not our intention to have you sit here and wait through
0018
 1   the morning.  So we apologize to you for that.
 2                 And now you have our full attention, so
 3   if you will identify yourselves and tell us what you've
 4   come to say.
 5                 MR. SWEITZER:  Thank you very much, Mr.
 6   Chairman.  There we go.  Thank you.
 7                 My name is Donald R. Sweitzer.  I am the
 8   senior vice-president of Global Business Development
 9   and Public Affairs for Gtech Holdings.  Mr. DeSocio
10   will speak in a minute.  This is Walter G. DeSocio, who
11   is senior vice-president, general counsel and corporate
12   secretary of -- of -- of Gtech.
13                 We're not offended at all that we've --
14   that we've had to wait a little while.  Your work is as
15   important as ours.  And we actually very much
16   appreciate the opportunity to come here so soon after
17   our announcement.  And we appreciate your invitation to
18   be here today to answer any questions that you may
19   have.
20                 We -- we have a great respect and a long
21   relationship with this lottery.  The -- the Texas
22   lottery -- we talk about the Texas lottery model around
23   the world, frankly.  When we -- in my own
24   responsibilities, I spend a lot of my time outside of
25   the country talking to countries and states and
0019
 1   jurisdictions that are thinking of putting in a
 2   lottery, whether it be Sri Lanka or Thailand or Russia.
 3                 And I don't know whether you-all know how
 4   much the Texas lottery is invoked around the world
 5   because when we are advising governments of best
 6   practices and things along those lines as to how they
 7   should set up a lottery, the -- very early in that
 8   conversation is a discussion of the Texas model because
 9   the -- the innovative way in which you-all have run
10   this lottery over the years is something that is
11   pointed to around the world as a model of best practice
12   that ought to be the model for any kind of a new
13   operation.
14                 So we are pleased that we are your vendor
15   and we intend to continue this relationship.  We -- we
16   don't have any formal statement to make because I think
17   what we'd like to do is respond to -- to your
18   questions.  And Walter may say a few things when I'm
19   done.
20                 But the main message that we're here to
21   say to you is that this -- this acquisition is one that
22   is basically a vertical acquisition.  This is not as if
23   one competitor is buying another.  And the classic
24   thing that happens in those kinds of situations is
25   there's a lot of change, there's a lot of duplicate
0020
 1   work and so on and so forth.
 2                 Generally speaking, there is turmoil in a
 3   lot of the change.  As Gary said, this is the --
 4   Lottomatica is the operator of the Italian Lottery.
 5                 They are more like you than they are like
 6   us.  And therefore, the manner in which we will
 7   continue to operate the Gtech Corporation will be very
 8   much as you understand it now.
 9                 We will not be losing our identity.  The
10   Gtech Corporation will still be headquartered in United
11   States, in Rhode Island.  We're building a new building
12   in downtown Providence right now which will be open
13   in -- in November.  And the manner in which we serve
14   our clients will be very much along the lines of how it
15   is now.
16                 The personnel, as -- as you probably read
17   in the announcements, Bruce Turner, our CEO, who has
18   appeared before you before, will be -- will continue to
19   be the CEO of Gtech and he will also be the CEO of the
20   new entity, Lottomatica.
21                 The -- one of the things that the
22   acquirers have asked all of us is for the entire
23   American team, senior management team, to manage the
24   combined operations.  And so, therefore, there will be
25   Lottomatica Holdings Company of which Bruce will be the
0021
 1   CEO and there will be two units in that company,
 2   Lottomatica and Gtech.
 3                 And Bruce will continue to be the CEO of
 4   Gtech.  All of us will continue as officers.  There
 5   will be -- there will be literally no change in the
 6   personnel that -- that serve you.
 7                 And we think there will be many strengths
 8   after this acquisition is completed, as far as being
 9   able to strengthen us even further to allow us to do
10   the things that hopefully we are chosen for, and that
11   is the research and development that we do to allow us
12   to continue to be the market leader.
13                 And so our basic message to you -- and
14   we'll respond to any particular questions you have.
15   And I know Walter may have some exchanges that -- as
16   far as the description of how you intend to examine
17   this -- this -- this acquisition which, of course,
18   you -- you should do.
19                 Our main message is that this will not be
20   turmoil or major change that you -- the -- the way that
21   you know that we operate now will change very little.
22   We think it strengthens us both financially and as a
23   market leader.
24                 And we look forward to completing the --
25   the -- all -- all that must go into that and being open
0022
 1   all through this process to any inquiries or
 2   investigations or research that you need to do.  That's
 3   the basic message.
 4                 And I know Walter wants to say a few
 5   words.  And then we'll be happy to respond to any
 6   questions that you may have.
 7                 CHAIR CLOWE:  Thank you.
 8                 MR. DeSOCIO:  Thank you very much.  I
 9   support everything else that Don said and add that,
10   over the coming months -- and this will be a month -- a
11   multimonth process as we work through all the
12   conditions necessary to consummate this transaction.
13                 We intend to work fully and cooperatively
14   with all of our customers and that obviously includes
15   the Texas Lottery.  So I want to make it clear that we
16   are going to be at your disposal and at the disposal of
17   other customers to provide whatever information it is
18   in order for you to carry out what you believe to be
19   your duties to the citizens of your state.
20                 The transaction, as we view it, is really
21   one very much of continuity.  As Don said, we are going
22   to be keeping things in place.  There were -- there was
23   a time when we talked to several members of the staff
24   as a way of updating them about a transaction that our
25   board had been considering for many, many months.
0023
 1                 And during the course of that
 2   transaction, we had some discussions about, frankly, a
 3   transaction that looked very different from the one
 4   that we announced earlier this week and that, in my
 5   view, might have given one reason to look at the
 6   delivery and financial capabilities of the resulting
 7   company.
 8                 We believe that the acquisition
 9   transaction announcement that was described very well
10   by Gary that was announced earlier this week results in
11   a combined company that's very healthy, that has --
12   will have an investment-grade rating and we believe
13   will have the delivery capabilities to provide you with
14   what you expect from us under the contract and even
15   beyond what we deliver under the contract.
16                 So with that, we'd be happy to either
17   take questions now or take away things that you'd like
18   us to do later and come back and work with your staff
19   on.
20                 CHAIR CLOWE:  Well, gentlemen, I want to
21   thank you again for being here today and tell you that
22   we appreciate the timeliness of your visit with us this
23   morning.  I think we are fortunate that we had this
24   meeting scheduled and had you on the agenda.  And we're
25   appreciative of you coming from Rhode Island and being
0024
 1   here with us.
 2                 I'm also appreciative of your kind
 3   remarks about the Texas model.  There are some that
 4   think we're somewhat contentious and hard to do
 5   business with.  My view is that the focus of this
 6   agency is to protect the people of Texas and to make
 7   certain that the players are treated with honesty and
 8   fairness and we -- pretty stringent on that
 9   requirement.
10                 I've been both the acquirer and the
11   acquiree in these kinds of deals.  And, you know, you
12   really don't believe there's going to be any change,
13   you better look at that one again because there always
14   is change.
15                 MR. SWEITZER:  Well, we -- we know that.
16   We're speaking, of course, in the manner in which we
17   run the operation.  We -- we have also been acquirers.
18   And in this company, we haven't been acquired in our
19   experience, anyway.
20                 But no, we fully understand there --
21   there will be cultural differences and changes and just
22   different personalities.  The advantage that we have
23   here, though -- and frankly, we didn't have it in the
24   original -- as Walter mentioned, the original proposed
25   deal that was negotiated for some months and then
0025
 1   folks -- and then it moved on was with people who we
 2   didn't personally know but we came to know.
 3                 We know the people at Lottomatica because
 4   we're in the business and because we also operate in
 5   Italy.  We have come over the years to know most of the
 6   people that we're dealing with.  And that gives you a
 7   little bit more comfort.
 8                 But we're also big boys and we know that
 9   there will be different folks and different ideas and
10   certainly we know that there will be some changes and
11   perhaps even a little head-knocking from time to time.
12   But that's -- that's human nature and life.
13                 CHAIR CLOWE:  Well, you know, what brings
14   that about is you take their money.  And when you take
15   their money and they own you, then they're going to
16   call the tune.  And I think this deal is different than
17   the first deal that we heard about.  And I think that
18   represents just how quickly change can come about.
19                 And these deals are interesting.  And
20   certainly a deal that involves the acquisition of a
21   domestic United States corporation by a foreign
22   corporation which is traded on a foreign exchange where
23   58 percent is controlled by private entity is a complex
24   and a very interesting business arrangement.
25                 And our interest, as I think is justified
0026
 1   by our contract, is a determination of the suitability
 2   of whoever the survivor in this transaction might be
 3   regarding assuming the role of the operator to the
 4   Texas Lottery.  And that's our interest.
 5                 We certainly intend to, in an obligation
 6   to the state, to look into this matter.  And I would
 7   ask you the questions of:  Do you intend to be
 8   supportive to that effort insofar as, specifically
 9   comes to mind, the retention of outside firms, attorney
10   and the outside firm of investigators and are you open
11   to the posting of monies that would enable those
12   activities to go on and be funded by Gtech?
13                 MR. DeSOCIO:  The answer to that is very
14   clearly yes.  And we've -- I've had some discussions
15   with your general counsel about the mechanics of
16   posting appropriate funds that will allow you to begin
17   to do your work.
18                 But clearly that is -- that's -- that's
19   beyond doubt that we'll -- you know, we want to
20   facilitate what process you feel you need to undertake.
21   And we're at your disposal in terms of financial
22   support to get it done and in terms of informational
23   support to help your advisers do their job.
24                 CHAIR CLOWE:  And in regard to that,
25   there are categories of states that you are the
0027
 1   operator for.  And as I understand it, there's a
 2   condition of the deal going forward that 87 percent of
 3   those entities in some form must give some sort of an
 4   assurance -- and I'm not clear on the definition of
 5   that and I don't want to be pinned down to stating what
 6   specifically that is because I don't know.
 7                 But my question doesn't go to that.  It
 8   goes more to your attitude towards those entities, the
 9   states where you are the lottery operator, working
10   together and being a single entity in this
11   investigative and determination of their feelings about
12   suitability.
13                 What's your attitude about that at this
14   point in time?
15                 MR. SWEITZER:  I think our attitude is --
16   with all of those other entities as -- is the same as
17   what we're saying here today.  We -- we look forward to
18   and will cooperate with and encourage a thorough scrub
19   by all of these states into this transaction.
20                 So I might also add that Lottomatica is
21   not an unknown entity in this industry in the United
22   States.  As you may know, Lottomatica actually owned a
23   significant percentage of one of our competitors up
24   until a couple of years ago and, therefore, has been
25   through many of the processes that a company must go
0028
 1   through in various jurisdictions in the United States.
 2                 So they are not an unknown entity, I
 3   think, even to the Texas Lottery.  But we look forward
 4   to a thorough look into the and monitoring of the
 5   transaction as it goes forward by all of the entities
 6   that we serve.
 7                 CHAIR CLOWE:  Well, I think then what I'm
 8   hearing you say is you're open to that and you don't
 9   have a plan working right now to either work with
10   individual states or try to bring the states together
11   and have them act in consort.
12                 Is that correct?
13                 MR. DeSOCIO:  I think that's largely
14   correct.  As you probably know, when the first
15   iteration of the transaction was under discussion and
16   we had -- that is Gtech and its representatives had
17   discussions with, I will say, 20 to 22 or 3, maybe
18   virtually all the domestic lotteries for which it
19   provides services, we -- we received a letter from the
20   New York General Counsel's Office advising us that it
21   had -- it had made an arrangement.
22                 And we were taking the letter at face
23   value.  It had made an arrangement to pool resources so
24   that the various domestic lotteries can take advantage
25   of efficiencies and understanding what, at that time,
0029
 1   was an exceedingly complex transaction and that the
 2   decision-making would certainly be done by each
 3   domestic jurisdiction but the pooling of resources
 4   would be an efficient way of having experts hired and
 5   information obtained about the transaction.  This is, I
 6   would submit, a different type of transaction.
 7                 And at the end of the day, we operate by
 8   the customer rules principle.  And if states have
 9   specific question with regard to their lottery
10   relationship with us, we intend to deal with those
11   bilaterally.
12                 And I would -- I would suspect that, now
13   that the transaction has become publicly-known, that
14   many lotteries are now beginning to consider what is
15   the appropriate level of resource commitment and
16   analytical commitment to a transaction that looks very
17   different than the one that they were preliminarily
18   advised about a couple months ago.
19                 CHAIR CLOWE:  Well, and I think there's
20   yet to be more information developed on the answer to
21   my question regarding that issue.  And you've taken
22   your answer as far as I believe you have it at this
23   point.
24                 MR. DeSOCIO:  That's correct.
25                 CHAIR CLOWE:  In regard to the
0030
 1   transaction that you have made at this point in time
 2   that I read about, how far to the controlling entity --
 3   and that would be the majority stockholder -- do you
 4   anticipate the willingness to disclose information,
 5   insofar as suitability, is going to be available to
 6   ourselves, for example?
 7                 MR. DeSOCIO:  Well, I can't speak for
 8   them.  But I think I can give you a pretty -- a pretty
 9   informed and educated answer.
10                 I think De Agostini, the controlling
11   shareholder, is committed to providing information to
12   the -- the degree necessary on its operations and on
13   its directors and stockholders in order to provide
14   satisfaction both to the lottery customers and to the
15   gaming customers, gaming jurisdictions, that will have
16   similar types of investigatory exercises.
17                 As Don stated earlier, we understand that
18   they have been through the process before in connection
19   with their Scientific Games' investment.  There were
20   both lottery and gaming licenses issued and
21   investigations conducted.  So we feel very, very
22   confident.
23                 And, in fact, I'm sure -- I wasn't part
24   of the board process, but I think I can probably speak
25   for the board in saying that one of the considerations
0031
 1   that would have been important to the Gtech board would
 2   be to understand that the buyer has a commitment to do
 3   what's necessary to consummate the transaction.
 4                 Otherwise, it wouldn't be, in my view, a
 5   deal that would be as -- as positive to the -- for the
 6   Gtech shareholders.  So ascertaining that level of
 7   interest and commitment, I think, was a gut-check
 8   decision by our board.
 9                 Certainly the management members of Gtech
10   feel very confident about the commitment.
11   Historically, in the last couple of years, they have
12   evidence of the commitment to submit themselves to the
13   appropriate level of investigation in different
14   jurisdictions.
15                 So I believe you'll see cooperation to
16   the extent that Texas law or California law or New York
17   law or Canadian law requires.
18                 CHAIR CLOWE:  I think that you've
19   answered preliminarily the issues that I have in my
20   mind.  I think that clearly we view this as a major
21   change in your company.  It's a change of control as I
22   see it.
23                 And I think, as I've said before, it is
24   our responsibility to determine that there is
25   suitability and that's our responsibility.  We will
0032
 1   need to work closely with you if you want some kind of
 2   a response from us and I -- I couldn't begin to
 3   describe what that's going to be until we get further
 4   into it.
 5                 But again, we appreciate your being here
 6   and giving us this initial indication of your
 7   willingness to work with us in dealing with us.
 8   Commissioner Cox?
 9                 COMMISSIONER COX:  (Shaking head)
10                 CHAIR CLOWE:  Anything --
11                 MS. KIPLIN:  May I -- just to be clear in
12   my mind, may I ask a question of the general counsel?
13                 CHAIR CLOWE:  Sure.
14                 MS. KIPLIN:  I just want to make sure
15   that I'm clear.  And I appreciate the discussions we've
16   had off the record.  But I want to make sure that Gtech
17   is committing to paying the costs of whatever we need
18   to do, including the cost of our outside counsel that
19   we've retained and any outside investigators, and you
20   will post funds with us from which we can draw to pay
21   for those costs and --
22                 MR. DeSOCIO:  That --
23                 MS. KIPLIN:  -- will replenish those
24   funds as -- as we need, that we need to.
25                 MR. DeSOCIO:  That is our agreement, yes.
0033
 1                 MS. KIPLIN:  Thank you.
 2                 CHAIR CLOWE:  That's what I thought he
 3   said.  Feel better now?
 4                 MS. KIPLIN:  I do.
 5                 CHAIR CLOWE:  Anything further,
 6   gentlemen?
 7                 MR. SWEITZER:  I think the only thing I
 8   would add in closing, Mr. Chairman, if you'd allow me
 9   to, is to reiterate our commitment to the state of
10   Texas.  Another point that I would make that I think
11   demonstrates it and will continue in the -- in the new
12   arrangement is what we've done here in Austin, Texas.
13                 We are a worldwide company.  We have
14   three technology hubs in the world.  And, as you know,
15   a couple of years ago we chose one of those technology
16   hubs to be Austin, Texas.  Austin, Texas, West
17   Greenwich, Rhode Island and Warsaw, Poland.
18                 And we are very proud and very confident
19   of what we've done here in Austin in not only serving
20   the Texas Lottery from this technology center but also
21   other states in the United States and some foreign
22   lotteries that are served from a facility right up the
23   street.
24                 And so I think that further demonstrates
25   our commitment to this state, which will continue in
0034
 1   whatever the new arrangement is going forward.  We are
 2   proud of our association with Texas and very pleased to
 3   be here today.
 4                 CHAIR CLOWE:  Thank you very much.  Thank
 5   you both.
 6                 MR. DeSOCIO:  Thank you.
 7                 CHAIR CLOWE:  Mr. Fernandez, are your
 8   folks still here?
 9                 MR. FERNANDEZ:  Yes, sir.  They're still
10   here.
11                 CHAIR CLOWE:  Commissioner, with your
12   approval, we have some outsiders that are here that
13   want to cover an item and then we'll go into executive
14   session.
15                 Mike, what -- what item is yours on our
16   agenda?
17                 MR. FERNANDEZ:  It's item number 13, Mr.
18   Chairman.
19                   AGENDA ITEM NO. XIII
20                 CHAIR CLOWE:  Let's go to item number 13,
21   report, possible discussion and/or action regarding the
22   demographic study of Texas lottery players.
23                 And gentlemen, we appreciate you-all
24   being here and we apologize for the lateness in asking
25   you to come forward and appear before us.
0035
 1                 MR. FERNANDEZ:  Good afternoon, Mr.
 2   Chairman and Commissioner Cox.  As you know, we have
 3   entered into an interagency agreement with the
 4   University of Texas Arlington for the conduct of a
 5   demographic study of lottery players.
 6                 With me this afternoon I have Dr. Richard
 7   Cole, who's the dean and professor of the School of
 8   Urban and Public Affairs from the University of Texas
 9   Arlington.  I believe with him he has a research
10   associate, Robert Wilkins.  I also have with us today
11   Dr. Sizemore, our -- our internal researcher.
12                 And with that, I would ask Dr. Cole to
13   present their findings.
14                 DR. COLE:  Thank you very much.  I
15   believe I can go through this quickly.  There's a lot
16   of data and information in the report and we're not
17   going to go through page by page.  But we felt we would
18   go over the highlights with you and respond to
19   questions as we can.
20                 As mentioned, my name is Richard Cole.
21   And I'm dean of the School of Urban and Public Affairs
22   at the University of Texas at Arlington.  And with me
23   today is Bob Wilkins, a research associate at the
24   school who's worked with me on this project.
25                 CHAIR CLOWE:  And do you get to bask in
0036
 1   the glory of the Rose Bowl?  Are they sharing that with
 2   you guys or not?
 3                 DR. COLE:  Absolutely.  Well, we bask in
 4   the glory for sure.
 5                 CHAIR CLOWE:  Okay.
 6                 DR. COLE:  As much as we can.
 7                 We both want to say that we very much
 8   appreciate the opportunity to have worked with the
 9   commission on this effort.  We know it's an important
10   project for the commission and for the state of Texas
11   and it's an honor of us to have been -- for us to have
12   been part of it this year.  And we thank you very much
13   for that.
14                 Before we look at the results of this
15   year's demographic study, I want to briefly review our
16   research design and discuss a few points related to
17   methodology and research design.
18                 The telephone survey was conducted during
19   the months of October and November of 2005 using a
20   method known as random-digit dialing, a process that
21   ensures that all households in Texas with listed and
22   unlisted phone numbers have an equal opportunity of
23   being included in the sample.
24                 The survey was available in both English
25   and Spanish and Spanish-speaking interviewers were
0037
 1   available for those who wanted it.  We completed
 2   interviews with 17 hundred Texas residents age 18 or
 3   above, a sample size yielding a margin of error of plus
 4   or minus 2.4 percent.
 5                 Throughout this report, we report
 6   statistical tests known as tests of statistical
 7   significance.  Let me briefly review the purpose of
 8   these tests.
 9                 Whenever we have information collected
10   from a sample as compared with information collected
11   from the entire population, there's a chance that
12   relationships we observe in the sample data may not
13   actually accurately reflect true relationships in the
14   population.
15                 You can never entirely eliminate the risk
16   of making mistakes, but tests of statistical
17   significance tell us the probability of making an error
18   in concluding that a relationship found in the sample
19   actually does represent a relationship exists in the
20   population.
21                 When a relationship found in the sample
22   is reported to be significant -- and let's say the .05
23   level, this means that in only five out of one hundred
24   samples of this size would we find a relationship this
25   large if, in fact, no relationship exists in the
0038
 1   population.
 2                 When report -- when we report a
 3   relationship to be significant at the .001 level, this
 4   means that in only one out of a thousand samples would
 5   we find a relationship this large if, in fact, no
 6   relationship exists in the population.
 7                 If a relationship is reported as not
 8   significant, this means that the relationship in the
 9   sample is too small to conclude with an acceptable
10   degree of confidence that the relationship actually
11   exists in the population.
12                 And so tests of significance can be
13   thought of as measures of confidence.  They tell us how
14   confident we can be that relationships found in our
15   sample actually do reflect relationships existing in
16   the population.
17                 When we are confident that a relationship
18   exists in the population, we say that relationship is
19   statistically significant.  When we are not confident,
20   we say the relationship is not statistically
21   significant.
22                 Another issue deals with the use of
23   unweighted data.  Our study, like all previous studies,
24   uses data in their unweighted format.  This means that
25   the data, after having been collected, are not adjusted
0039
 1   to match specific proportional demographics of the
 2   population.
 3                 Reporting our results this way permits
 4   direct comparison with previous studies and does not
 5   affect group-by-group comparisons such as comparing
 6   participation rates of men with women, Hispanics with
 7   non-Hispanics, lower income with higher income
 8   individuals, and so forth.
 9                 And for the most part, these are the type
10   of comparisons we report on in this study.  It should
11   be noted, though, that projections of sample results to
12   the population should be interpreted with appropriate
13   caution when using unweighted data.
14                 And a final methodological issue concerns
15   extreme cases.  A few respondents report
16   extraordinarily large amounts of money spent on the
17   various lottery games.  In some cases, these extreme
18   cases skew the results so much so that the reporting of
19   the average or mean amount spent on any game is
20   misleading.
21                 In order to control for this, we
22   typically report the median as opposed to the mean
23   amount spent.  The median is that number that divides
24   the distribution in half so that 50 percent of
25   respondents report spending that amount or more and 50
0040
 1   percent report spending that amount or less.  As such,
 2   the median serves as a correction for extreme cases.
 3                 So before we report the results, I would
 4   pause if there are questions on any of those issues.
 5                 COMMISSIONER COX:  I just have one
 6   question on the sample selection basis.
 7                 DR. COLE:  Yes.
 8                 COMMISSIONER COX:  You used land line
 9   telephone --
10                 DR. COLE:  Yes.
11                 COMMISSIONER COX:  -- numbers?
12                 DR. COLE:  Yes.
13                 COMMISSIONER COX:  And I assume you got
14   from the various phone companies the population of this
15   block of numbers that you used just like another
16   business.
17                 DR. COLE:  That's -- that's essentially
18   true.  And then from that block, a randomization
19   process is used to randomly select the phone numbers.
20                 COMMISSIONER COX:  Okay.  So from that,
21   you got listed and unlisted numbers.
22                 DR. COLE:  That's true.
23                 COMMISSIONER COX:  What did you do with
24   the business numbers that you had?
25                 DR. COLE:  The business numbers are
0041
 1   deleted from the sample --
 2                 COMMISSIONER COX:  So they're just
 3   residential numbers?
 4                 DR. COLE:  Right.
 5                 COMMISSIONER COX:  And do you have any
 6   concern about the fact that, as I read, a growing
 7   number of people don't even have land line phones, they
 8   have cell phone?
 9                 DR. COLE:  Yeah, it is.  Cell phones are
10   eliminated from this sample as they, by law, must be.
11   But it's -- it's an increasing concern for survey
12   research in general, the number of people in the
13   population who are using cell phones as opposed to land
14   line phones.
15                 Data that I've seen indicated that it's
16   still a reasonably small percent, maybe three, four,
17   five percent.  But it's an increasing concern.  There's
18   no -- there's no question about it.
19                 I think that most studies have
20   demonstrated that, at this point, it does not seem to
21   make a difference.  In other words, people who use cell
22   phones are not considerably different from the rest of
23   the population.  So that bias tends to wash out.
24                 But I think that anyone involved in
25   survey research knows that that's an increasing issue
0042
 1   that will have to be dealt with at some point.
 2                 COMMISSIONER COX:  And how large do you
 3   think the economic strata you're missing at the bottom
 4   of the income array because they just can't afford a
 5   telephone?
 6                 DR. COLE:  I -- I don't think it's very
 7   significant.  We -- we have looked and we'll look in
 8   just a second, a comparison of the income demographics
 9   of the sample with the population of Texas.  And it's
10   very, very close.
11                 COMMISSIONER COX:  Thank you, Mr.
12   Chairman.
13                 CHAIR CLOWE:  Thank you.
14                 DR. COLE:  Okay.  Bob -- Bob, let's go to
15   the first chart.
16                 The first chart on the board, in fact,
17   does look at the characteristics of the sample.  And I
18   will not pause long on this except to note that, in
19   many of these categories, we were very close to the
20   population of the state of Texas.
21                 The income category, for example, you see
22   the percentages that fall in our sample.  I'll just
23   quickly read from top to bottom the census portion in
24   those same categories.  Under 20 thousand, it's 24
25   percent.  Then on up, it's 14 percent, 13, 10, 9, 9,
0043
 1   and 11.  So we're very close there.
 2                 We're close on the race variable, at
 3   least in terms of white and African-American or blacks.
 4   We're very close on the gender variable.  And we're
 5   reasonably close on the age variable.
 6                 The two variables where the sample
 7   characteristics deviate from the statewide numbers
 8   significantly are -- is in the area of education, where
 9   we have -- where in the state of Texas, there are about
10   20 percent who report less than high school and we have
11   about 10 percent.
12                 At -- we're close on the other high --
13   education factors and also Hispanic origin, where we
14   report about 20 percent whereas the adult Hispanic
15   population in the state of Texas, according to the
16   census, is about 28 percent.
17                 So those two factors are off a little bit
18   from the population.  The rest are very close.  I don't
19   think that any of them are off enough to affect the
20   analysis.  But that's the results that we have.
21                 Bob, let's go then to the next chart.
22                 The next chart looks at the percent of
23   the population saying that they have played any of the
24   lottery games from the initiation of the lottery in
25   1993 until the present.  The top bar -- tool 6, of
0044
 1   course -- represents the results of our survey.  All
 2   the other bars represent the results of previous
 3   surveys.
 4                 And you can see that, in general, there's
 5   been a decline in participation over the years up until
 6   perhaps this particular year, 2005, where we show this
 7   year roughly 51 percent say they played any of the
 8   lottery games in comparison to 47 percent last year.
 9                 So for the first time in a long time, it
10   looks like the percent saying they played at least has
11   not declined.  And I think we might conclude perhaps
12   that the proportion playing has stabilized at about
13   50-percent level.
14                 Bob, let's go to the next chart.
15                 The next table looks at the demographic
16   characteristics of players and nonplayers and compares
17   them on a number of demographic characteristics.  You
18   can see those on the board.  We have income, education,
19   race, Hispanic origin, gender, age and employment
20   status.
21                 And there are some variances in all of
22   those categories reported.  However, the only three
23   that are statistically significant and, therefore, for
24   which we have confidence that these represent the
25   population, are:
0045
 1                 Gender, where males report playing more
 2   heavily than do females by a small -- reasonably small
 3   percent; education, where there's a increase in play up
 4   until the some college/no degree category, where about
 5   56.8 percent say they play, and decline after that; and
 6   finally, employment status, where those employed
 7   indicate that they play at significantly higher rates
 8   than those unemployed.
 9                 None of the other characteristics that we
10   looked at distinguished significantly between players
11   and nonplayers this year.
12                 Bob, let's go to the next chart, please.
13                 The next table, which in the report is
14   presented as table three, is a complex table with lots
15   of information, as you see.  It's so large that we have
16   to put it in two -- two parts on our -- on the board.
17                 But what we have here is, for every
18   lottery game, we've indicated the number and the sample
19   who say they play, the percent who say they play, the
20   average number of times played per week, the average
21   number of times played per month, the average number of
22   times played per year, the average spent per play,
23   average spent per month as a mean, and the average
24   spent per month presented as a median.
25                 I won't go through all of that with you
0046
 1   except to just make a couple of highlight points.
 2                 Bob, go to the next chart, please.
 3                 First of all, the most popular game again
 4   this year is indicated to be Lotto Texas, where 84
 5   percent of our sample say they play that or say they
 6   played that within the last year.  The least popular is
 7   Texas Two Step, played by only 15 percent of the
 8   population.
 9                 And in terms of frequency of play, the
10   various scratch-off games tended to be played more
11   often than others.  And I'll not report any of the
12   other details of that except to say that this parallels
13   very closely with the same results found in last --
14   last year's study, 2005.
15                 As an example, that year 80 percent said
16   they had played Lotto Texas.  We found 84 percent.  So
17   it's -- the percentages in each of those categories are
18   roughly parallel to that.
19                 Then we conclude the report with a number
20   of charts that show various bar graphs.  Each of the --
21   each of the charts shows a set of bars that indicate
22   the percent in that category of person playing lottery
23   this last year and also shows a set of bars indicating
24   how much they say they spend per month.  The percent
25   playing are in blue and the amount spent per month is
0047
 1   in red.
 2                 The first chart presents that by
 3   education.  And you can see that respondents with a
 4   college degree and graduate degrees report playing at
 5   levels somewhat lower than the rest of the population.
 6   And that difference is statistically significant.
 7                 You could also see that the amount of
 8   money spent per month on Texas lottery games tends to
 9   decline consistently with levels of education, from a
10   high for those with less than a high school education
11   of about 50 dollars a month to low for those with a
12   graduate education of about 13 dollars a month.  And
13   that really should shift statistically significant.
14                 The next set of charts looks at those
15   behaviors by income.  And while there's a lot going on
16   in those bars as you see up and down, none of the
17   relationships are statistically significant, meaning
18   that the relationships are not large enough that we can
19   report they do or do not represent relationships in the
20   population.
21                 The next chart looks at lottery play and
22   dollars spent by racial categories.  Here you can see
23   that rates of participation are highest for blacks or
24   African-Americans and lowest for those in the other
25   bar, which in Texas is primarily Asian.
0048
 1                 However, that relationship is not
 2   statistically significant.  But when we look at
 3   expenditures per month, we see that blacks or
 4   African-Americans report spending the largest amount of
 5   money per month and white respondents report spending
 6   the least.  And that difference is statistically
 7   significant.
 8                 So while the percent playing between the
 9   races does not differ statistically, the amount of
10   money spent by those playing does tend to differ in a
11   significant way.
12                 The next chart looks at -- separately, it
13   compares Hispanic players with non-Hispanic players, in
14   terms of percent playing and amount of money spent per
15   month.  And you can see that there are no significant
16   differences reported there.
17                 The next chart looks at percent play and
18   amount spent by gender.  And you can see that males do
19   significantly report playing more so than do women.
20   But the difference spent per month is only two dollars
21   and it's not statistically significant.
22                 The next figure looks at percent play and
23   dollars spent by age.  And you can see that, while
24   there are some variances reported there in the sample,
25   none of those differences are statistically
0049
 1   significant.
 2                 The final chart that we will report on
 3   today looks at play and amounts spent by employment
 4   status.  You can see again that participation rates of
 5   those employed are higher than for the unemployed or
 6   the retired.  And the difference is statistically
 7   significant.
 8                 And you can see that the reported
 9   differences spent per month by those various categories
10   are not statistically significant.
11                 So I will conclude my report with this
12   summary.  We find the rates of participation appears to
13   have leveled off this year.  For the first time in
14   several years, the percent saying they have played any
15   of the lottery games in the past year did not decline
16   from the previous survey.
17                 Point two:  In terms of those demographic
18   factors related to play, we found participation to vary
19   significantly by gender, education and employment
20   status.  Men, those with some college, and those
21   employed report playing at higher rates than others.
22                 And finally, in terms of amount of money
23   spent, we found those characteristics to be most
24   directly and significantly related to be race and
25   education.  Those with lower levels of education and
0050
 1   blacks or African-Americans reported spending more per
 2   month when they played than the other respondents.
 3                 And that concludes our report.
 4                 CHAIR CLOWE:  Very good.  Gary, do you
 5   have any comments on this?
 6                 MR. GRIEF:  Just briefly, Chairman and
 7   Commissioner Cox.  My comments in reading through the
 8   report, when I look at the rates of participation and
 9   the amount of money spent by those different
10   categories, none of that is surprising to me.  I think
11   it's fairly consistent with past reports.
12                 I think it's very important that anyone
13   who reads this report have a good understanding of what
14   the term "statistically significant" means.  And I hope
15   that, when everyone goes away from the meeting today,
16   that we all have a good understanding of that.
17                 The encouraging part of the report for me
18   was the 51 percent that reported playing the lottery
19   over the past year.  And I think that is a -- a direct
20   reflection of the success that this agency is having
21   with marketing our products towards all players.
22                 And finally, after three or four, or
23   whatever the number was, of consecutive years of
24   decline, we've seen a stabilization at least in this
25   report.  We could see that change next year.
0051
 1                 But in general, that's my response to
 2   this report.
 3                 CHAIR CLOWE:  Any questions?
 4                 COMMISSIONER COX:  I do have a couple of
 5   questions.
 6                 Dr. Cole, I'm interested in how these
 7   spending numbers that you determined per person would
 8   blow up to the population, the adult population of
 9   Texas.
10                 DR. COLE:  Yes.
11                 COMMISSIONER COX:  For instance, if we
12   look at figure six, we show how many females play.  We
13   know how many adult females there are in Texas.  We
14   know how much each -- we have a number here for what
15   each spends.
16                 And if we do that for the women and we do
17   that for the men, would we get something like three and
18   a half billion dollars?
19                 DR. COLE:  Well, you should for sure.
20   I -- I did a quick extrapolation myself.  I think, to
21   actually do it correctly as you say, you would have to
22   take literally every category of respondent because
23   every category plays at a different rate.  And when
24   they play, they spend at a different level.
25                 So I think you would have to take a very
0052
 1   large matrix and blow every single one of them up to
 2   the population and then add them up to do that.  And I
 3   did not do that.
 4                 But I did take a quick approximation,
 5   taken all adult Texans 18 years of age older and
 6   multiplying it by the percent in this study who say
 7   they spent and use them as a monthly multiplier an
 8   amount of money on average -- on average spent
 9   correcting for the very extreme cases those who
10   reported spending more than one standard deviation
11   above the average, multiplying that by 12.
12                 And I get a figure of roughly 3.8 billion
13   dollars plus or minus, you know, an error factor there.
14   I think the actual expenditure for the last 12 months
15   was around 3.6 billion.  So it's a rough approximation.
16                 Again as I say, I think to do it one
17   hundred percent accurately, you would need to do it for
18   every category of the sample.  And I would guess, if so
19   done, it would be even closer to the 3.6 million.
20                 COMMISSIONER COX:  Okay.  So you did it
21   for which one of the particular slices?
22                 DR. COLE:  I did it for the entire
23   sample, for the entire sample.
24                 COMMISSIONER COX:  For the entire sample?
25                 DR. COLE:  Yes, sir.
0053
 1                 COMMISSIONER COX:  Okay.  So there might
 2   be an over and under, if we took gender versus income
 3   or something like that?
 4                 DR. COLE:  Right.  Right, definitely.
 5                 COMMISSIONER COX:  So you know it didn't
 6   come out to five and a half or one and a half?
 7                 DR. COLE:  Pardon me?
 8                 COMMISSIONER COX:  You know it wasn't way
 9   off.  It was --
10                 DR. COLE:  Right.  It was very -- it was
11   very close.
12                 COMMISSIONER COX:  So if there's
13   underreporting or overreporting or just general
14   inaccuracy, it doesn't look like to you it's very big?
15                 DR. COLE:  Right.  That's true, sir.
16                 COMMISSIONER COX:  Thank you, Mr.
17   Chairman.
18                 CHAIR CLOWE:  Gentlemen, we thank you.
19   We appreciate the report and your concise report to us.
20                 DR. COLE:  Thank you very much.
21                 CHAIR CLOWE:  Thanks, Mike.
22   Commissioner, may I move we go into executive session
23   at this time?
24                 COMMISSIONER COX:  Do you want to try to
25   get Dr. Eubank?
0054
 1                 CHAIR CLOWE:  Sure.  Where are we with
 2   him?  What item?
 3                 COMMISSIONER COX:  I don't know which
 4   item.
 5                 MR. GRIEF:  Nine.
 6                 CHAIR CLOWE:  Say it again.
 7                 MR. GRIEF:  Nine.
 8                     AGENDA ITEM NO. IX
 9                 CHAIR CLOWE:  Then let's take up next
10   item number nine, report, possible discussion and/or
11   action on the lottery terminal functionality, including
12   the quick pick feature, for all games and/or impact to
13   players.
14                 Mr. Anger?
15                 MR. ANGER:  Good afternoon,
16   Commissioners.  For the record, my name is Michael
17   Anger.  I'm the Lottery Operations director.
18   Commissioners, I'm joined at the table today to your
19   left by Dr. Randall Eubank, the agency's statistical
20   consulting services vendor.
21                 And I'm also joined today on your right
22   by Chenghui Luo, consulting software engineer, and Ron
23   Goodrich, technical manager for Gtech Corporation.
24                 Dr. Eubank has joined us today to report
25   on his analysis of quick pick related data to terminal
0055
 1   functionality for all the Texas Lottery online games.
 2   In addition to his initial analysis, Dr. Eubank worked
 3   with representatives from Gtech, including Mr. Luo, to
 4   complete secondary review and analysis related to his
 5   initial findings.
 6                 Mr. Luo and Mr. Goodrich have been --
 7   been kind enough to make themselves available today to
 8   respond to any questions following Dr. Eubank's
 9   presentation.  And with that, I'll turn it over to
10   Dr. Eubank.
11                 DR. EUBANK:  Good afternoon.  For the
12   record, my name is Randy Eubank.  I am a professor in
13   the Department of Mathematics and Statistics at Arizona
14   State University.  It's nice to be in Austin again and
15   Hook 'Em Horns.  I can say -- as an Aggie, I can say
16   that here and get away with it and --
17                 CHAIR CLOWE:  We thank you for that.
18                 COMMISSIONER COX:  There was a shirt out
19   there that said Gig 'Em Horns.
20                 DR. EUBANK:  But not at A&M.  I can say
21   that much.
22                 COMMISSIONER COX:  It was actually
23   maroon.
24                 DR. EUBANK:  As -- as Michael indicated,
25   the reason I'm here is, over the last few months, I've
0056
 1   been working on the analysis of data that had been
 2   supplied to me by Gtech on their quick pick generation
 3   mechanisms.
 4                 And so today I want to talk -- tell you a
 5   little bit about what I've done and some of the things
 6   that have come up as part of that analysis.
 7                 Before I -- I go -- I delve into this and
 8   overkill everybody with numbers, I want to go off on a
 9   couple of tangents here and to preface some of the
10   things that come up.  And the first thing I want to
11   talk about for just a couple of minutes has to do with
12   random number generation.
13                 So when we're talking about the quick
14   pick generation mechanism or whether we're talking
15   about (inaudible) method of generating the -- the
16   megapliers, there's various -- various bells and
17   whistles that get added on top of this to make things
18   truly random.
19                 But at the heart behind all this work is
20   something called a random number generator or pseudo
21   random number generator.  And this is -- even though it
22   may be relatively new in terms of the experience of
23   some of the people in the room, this is a technology
24   that goes back -- it basically traces its way back to
25   the beginning of computers when people started putting
0057
 1   computers together.
 2                 And some of the big names from
 3   mathematics and computer science pop up here.  So the
 4   technology goes back into the 1940s with very good
 5   random number generators starting in about 1950.  So in
 6   scientific terms, this is pretty old stuff.  This has
 7   been around for a long, long time.
 8                 So there's nothing startling in terms of
 9   random -- what's going on with random number
10   generators.  And beyond that, the -- so random number
11   generators have been extensively studied in -- in the
12   public sector, at the universities, National Bureau of
13   Standards and other places of that nature.
14                 And these are a publicly-available, open
15   source.  They cost the price that I can afford, zero.
16   People can get them for nothing.  And they -- they're
17   known to work exceptionally well.  So for -- as an
18   example of that, there's something that's fairly recent
19   in this context.  It's something called Mersenne
20   Twister.
21                 And so the Mersenne Twister is -- let me
22   briefly talk about what a random number generator does.
23   You give it a starting point.  And what this algorithm
24   does is it spits out a long sequence, hopefully a long
25   sequence of numbers.  And one or two things can happen.
0058
 1                 At some point, either the numbers all
 2   become zero or, at some point, the sequence starts over
 3   again.  Now that doesn't mean that the sequence can't
 4   have the same numbers in it sometime.
 5                 What I'm saying is, at some point, if you
 6   knew where it started and you knew where it's going to
 7   start again, you can see every -- you know everything
 8   that's going to happen after that point when it starts
 9   over.
10                 Now the -- how long it takes for this
11   thing to start over is the period of the random number
12   generator.  And things in the order of two-to-the-30
13   and so on are not -- are relatively standard.  Well,
14   the Mersenne Twister has a period of -- and this is
15   public -- this is open source code.
16                 I can tell you two places that you can go
17   and download the C code for this for nothing.  You may
18   not be able to afford a Cadillac Coupe de Ville, but
19   you can have the Coupe de Ville of random number
20   generators on your own PC for zero dollars.
21                 All right?  And that's this Mersenne
22   Twister.  It has period two-to-the-20,000 roughly.  Now
23   to give you some context of how big that is,
24   two-to-the-30 is about 4.3 billion, which is what they
25   tell me Gtech is -- add or subtract about point -- a
0059
 1   trivial .2 billion and you'll get about the -- what
 2   I've heard is about what was paid for Gtech by
 3   Lottomatica.
 4                 And I -- I don't know what that is in
 5   lira.  So -- so but anyway, that gives you some sense.
 6   I mean, so this is two-to-the-30 -- 30 or 32, actually.
 7   And so two-to-the-20,000 thousand, you -- if we started
 8   that on -- on something like Texas Two Step and we used
 9   that for generating the sequences there, the sun would
10   probably burn out before we'd have to worry about it
11   starting over again.  And that's something -- I'd like
12   to be around for that, but I don't think I will.
13                 All right.  Now so the -- the -- the --
14   at issue here is that Gtech -- there are lots of really
15   good random number generators out there that are in the
16   public domain that have been tested, that have been --
17   there's published research on them.  Scientists have
18   looked at them.  They've proved theorems about them and
19   have extensively simulated them to where they really
20   understand them.  They've found some good ones and
21   they've found some bad ones.  But the good ones are out
22   there and they cost you nothing.
23                 Gtech doesn't use any of those.  And now
24   that doesn't mean that their random number generator is
25   either good or bad.  But a lot of things that I've done
0060
 1   that I'll talk about today would not -- would have been
 2   moot issues if one of these had been used because it
 3   would be something that you could point to a spot in
 4   the literature and say there's 47 hundred articles on
 5   this that have already proved that it worked.
 6                 All right.  Now back to the other thing,
 7   which the previous discussion talked about a little bit
 8   and so I maybe don't have to say so much about this, is
 9   the term "statistical significance" because I'm going
10   to -- this is going to come up a lot.
11                 So let me go back over again what -- what
12   that means.  Now the term "statistical significance,"
13   when I say something is statistically significant, it
14   means that something has happened that I don't expect
15   to happen under a presumed model.
16                 Now when I say model, I'm not talking
17   about some sort of arcane mathematical construct.
18   Models are really very simple things that lots of
19   things -- in the context of the lottery, for example,
20   in the case of Cash Five, there are something like
21   about 400 thousand ways that you can choose five
22   numbers from 37.  Those are every possible board that
23   you can get in Cash Five.  There's about 400 thousand
24   of them.
25                 So a model here would be that every one
0061
 1   of those boards has the same chance of being selected
 2   or every one of those possible boards has the same
 3   chance of being generated by the quick pick generation
 4   methods.
 5                 So if I say that something's
 6   statistically significant in the context of Cash Five,
 7   what I'm saying is I'm seeing departure from what I
 8   would have expected under this uniform generation
 9   model.  And that's -- that's what I mean.
10                 Now unusual things happen.  And my
11   standard -- there are three standard rates that -- that
12   people use for level of significance:  Five percent, 10
13   percent and one percent.  I like to think of these
14   things as -- I like to think of this from the
15   standpoint of trial by jury.
16                 The level of significance is my
17   reasonable doubt.  The model here is what's on trial
18   and I'm trying to convict the model.  And the level of
19   the significance is my reasonable doubt.
20                 Now just like in trial by jury, I can
21   never be sure -- I can never be sure that somebody is
22   guilty.  I don't have to be sure.  We only have to be
23   sure beyond a reasonable doubt.  If we had to be
24   absolutely sure, there would never -- I would --
25   wouldn't have to worry about committing crimes.
0062
 1                 So the -- the difference between this and
 2   trial by jury is that we have standards that we usually
 3   go by.  And it's the five percent, 10 percent and one
 4   percent.  My typical use here is the five-percent
 5   level, which means that I call something unusual if I
 6   would only see it in about one out of every 20 times if
 7   everything was working exactly right or five times out
 8   of a hundred if everything was working -- was working
 9   exactly right.
10                 Because unusual things can happen --
11   because unusual things can happen, that doesn't
12   necessarily mean the model is incorrect if I say that
13   something's significant.  It just means that there's
14   evidence beyond a reasonable doubt that it's not
15   correct.
16                 Sometimes guilty people get put in jail.
17   And sometimes innocent ones get put in there.  All
18   right.
19                 And the last thing has to do with
20   statistical significance versus practical significance.
21   And I think this is something that I probably can't
22   emphasize enough.  The two terms can coincide but they
23   may not if -- an example of that, let's -- let's
24   suppose that I have a new drug that is able to cure one
25   more person out of a million than the drug that's
0063
 1   currently in place.
 2                 If you give me enough data on it, at some
 3   point I will eventually be able to show that there's a
 4   statistically significant difference between the two
 5   drugs.  And the question is whether or not that's
 6   practically significant.
 7                 If you're that one person out of a
 8   million, you'd probably say that it was.  But if you
 9   were on the board of directors at Pfizer, you wouldn't
10   change your entire marketing mechanism to do something
11   which, in your opinion, in the global structure
12   probably is not worth the effort.
13                 So they would -- somebody would call
14   that -- at Pfizer, they would say that that's of no
15   practical significance.  All right.  Now that's the
16   setup.  So now I get to spend the next few minutes on
17   the punch line.
18                 The -- this all began, as you know, on
19   June 16th when Gtech reported that they had detected an
20   abnormality in the quick pick generation mechanism for
21   Pick 3 so the numbers one, two, three and four were
22   being picked more often than they should be, around 11
23   percent, and the other numbers were being picked in --
24   with a smaller chance than they should be.
25                 So they -- they corrected the code and
0064
 1   tested the new -- the new implementation of it and
 2   indicated that everything had -- was fine at that
 3   point.  Now shortly after that was when I got a call
 4   from Michael, as I recall.
 5                 And I was asked to take -- to undertake
 6   the following project.  I had -- there was three parts
 7   to this.  One was to -- using data provided by Gtech, I
 8   was to analyze live quick pick transactions taken place
 9   immediately before the code fix, before the problem was
10   detected, live data after the problem was detected --
11   this is all for Pick 3 -- and also do a complete
12   analysis of the quick pick generations for all the
13   online games, including Pick 3.
14                 The data that I received for this had --
15   came in three forms.  There was live data that comes
16   from whatever the standard terminal is that Gtech uses
17   in most of the -- in most of the stores.  But there are
18   two other types of terminals.
19                 And I know less than -- I'll say what
20   little bit I know.  I won't go into the details of
21   this.  There are two other types of terminals that
22   are -- that are used.  They're not used very frequently
23   or they're in locations that are of -- that don't get a
24   lot of business.
25                 They're called GTD and SST.  And because
0065
 1   they're used so seldom or used not as much, it would
 2   take a long time to generate the amount of data that I
 3   need to do the analysis.  So I decided on the basis of
 4   that that those terminals, the data I'll get from them
 5   would be simulated.
 6                 So they would have a terminal.  They'd
 7   set it up and it would go off and generate the -- the
 8   quick picks.  And that was the data that I looked at.
 9                 So the -- the methodology -- one of the
10   reasons I was asked to do this is to give sort of an
11   independent view of the whole thing from a different
12   approach.  And that's -- that was my perspective on
13   this.  I didn't -- to my knowledge, I didn't repeat any
14   of the tests that Gtech would have used for this.
15                 I came at it from the -- from the point
16   of view of a blank slate.  And the basic problems that
17   arise here, in terms of testing or uniformity in two
18   cases, you've either got numbers that are picked from a
19   group of numbers with replacement or without
20   replacement.
21                 In the context of Cash Five, you're
22   selecting five numbers from 37.  You can't get the
23   number 1-1-1-1-1.  Okay?  That can't come up because
24   you do it without replacement.  So that's a case of
25   sampling without replacement.
0066
 1                 Something like Pick 3, for example, you
 2   can get a triple:  0-0-0 could come up.  And so that's
 3   a case where you're sampling with replacement.
 4                 The -- and then there are sort of mixes
 5   of these.  With something where you have a bonus ball,
 6   you can think of that as basically being a mix where
 7   you have without replacement for the -- the primary
 8   board.  And then the bonus ball is something that you
 9   could think of as being sampled without replacement or
10   with replacement, rather.
11                 So to find out the things that were sort
12   of the good things to do, I went to a database,
13   statistical database, called the current index to
14   statistics.  And this is a huge database, went back
15   many years, looks -- all the standard journal -- a lot
16   of nonstandard journals.
17                 And I knew some of this literature was
18   out there already.  But to make sure that I was
19   thorough, I went through the -- this large database,
20   found some interesting things, did my search engine
21   work through there and looked at more statistical
22   articles than most people would ever want to look at
23   and came up with a collection of methodology.
24                 The details on that you can find in the
25   reports.  So you can -- you can trace those down if
0067
 1   you're interested.  I won't go into any great detail
 2   other than that and other than to say that I think I
 3   was thorough in finding the good -- the good things.
 4                 So now let me talk a little bit about the
 5   data analysis itself.  So the -- I won't give the
 6   details on how much data I had.  As I -- as I go any
 7   further along, I won't spend much time on that.
 8                 But -- but for the Pick 3 case, let me
 9   talk about that for just a couple of seconds.  So I had
10   about 140 thousand live quick picks prior, for Pick 3
11   prior to the -- the code problem and about 120 thousand
12   after the code fix.  And then there was a hundred
13   thousand simulated from the GBT, 150 thousand from the
14   SST terminals.
15                 My analysis didn't show anything of
16   interest on the simulated data, so I won't talk about
17   that.  But I will talk for a couple minutes about the
18   -- the two live data sets.
19                 Now one of the issues here is just to see
20   if the -- to talk about what was actually going on
21   prior to the code fix.  And what I want to show you
22   here is, this is a corresponding confidence interval.
23   Simultaneous corresponding confidence intervals
24   actually correspond -- they were constructed from the
25   -- the live data set prior to the code fix that show --
0068
 1   these are things which, with 95-percent simultaneously,
 2   have a probability of capturing whatever the true
 3   probability of getting a 0-1-2-3 and so on.
 4                 Now you can see -- if you look here, you
 5   can see.  Here is this .93 or .093.  And here's about
 6   .11.  These are what the true probabilities should have
 7   been if the code problem was as it had been -- as it --
 8   as it had been diagnosed.
 9                 These are confidence levels constructed
10   from the sample.  As long as the red dots are inside
11   the blue lines, that's -- that's consistent with what
12   they believe the code -- the problem in the code was.
13                 And so looking at this, one would say
14   that they actually hit the nail right on the head and
15   did -- they -- they did know exactly what the issue
16   was.
17                 After the code was fixed, everything
18   should have occurred with probability .1.  And that's
19   what these true probabilities are.  And here are the
20   corresponding confidence intervals, of course, that --
21   that resulted from the live data after the code fix.
22                 Now this looks strange because I've
23   changed the scale.  But the scale here is only about,
24   what, three one-thousands.  So this is -- this is
25   taking a microscope and looking at it.  And there's
0069
 1   nothing going on, in terms of all this -- looks like
 2   bad dental work.
 3                 Now something more interesting --
 4   something more interesting here was -- what I just
 5   showed you was a univariate marginal distribution.  I
 6   looked at how many times I saw 0-1-2-3 and so on.  But
 7   there's another way to look at this data because these
 8   -- they really come to us in threes.
 9                 Each board has three things on it and
10   there are one thousand possibilities starting from
11   0-0-0, 0-0-1, 0-0-2 and so all the way -- and so on,
12   all the way up to 9-9-9.  And that's what I'm going to
13   show you next.
14                 So I looked at every -- I had enough data
15   where I could look at every one thousand of these
16   things.  And that's what this plot is right here.  This
17   is a -- a -- basically a histogram.  I'm showing you
18   the frequency of times that I saw numbers everywhere
19   from 0-0-0 up to -- to 9-9-9.
20                 And this is -- for those of you -- I'm
21   sorry.  For those of you who do crossword puzzles, I'm
22   going to -- this is a new word that you'll want to
23   remember.  This is in lexicographic order starting with
24   0-0-0, 0-0-1, 0-0-2, all the way up to 9-9-9.
25                 Now this red line here is the -- sort of
0070
 1   the no-fly zone.  It's the -- it's an -- it's a
 2   simultaneous upper bound.  And if everything is working
 3   as it should -- if everything is working as it
 4   should -- there's people that probably have to -- are
 5   going to have hearing problems now after -- after the
 6   amp amplified my voice, which doesn't need much
 7   amplification.
 8                 This is the no-fly zone right here.
 9   Things shouldn't come up above here.  Everything should
10   be -- all these spikes or all these -- these bars and
11   histograms should all be -- and you can see that
12   there's several of them up above it.
13                 And if you look, it's easy to diagnose.
14   This is 0-0-0, 1-1-1, 2-2-2, 3-3-3 and so on.  All the
15   triples were getting picked with higher probability
16   than they should have.
17                 So -- and this is -- this is for the
18   post, after the fix, live data.  The same things showed
19   up in the prefix live -- live data.
20                 I wrote a report on this.  This
21   information was passed on to Gtech.  They did some
22   investigation.  Their conclusion was that the reason
23   this had happened was because there was a terminal
24   functionality that would allow players/attendants to
25   reject a quick pick and that that was what had caused
0071
 1   the problem.
 2                 There were -- there were places where the
 3   players were just sitting there rejecting quick picks
 4   until they got them something that they liked.  And I
 5   guess they liked triples.
 6                 Well, in any case, the -- the terminal
 7   quick pick functionality was -- was turned off.  A new
 8   set of data was collected.  And this is basically the
 9   same picture I showed you a minute ago except now I'm
10   showing you both upper and low bounds for this.  And
11   instead of the bars, I'm just putting some circles
12   here.
13                 CHAIR CLOWE:  Dr. Eubank, let me ask
14   you --
15                 DR. EUBANK:  Sure.
16                 CHAIR CLOWE:  -- to define more fully
17   what you mean when you say players were allowed to
18   reject.
19                 DR. EUBANK:  They --
20                 MR. ANGER:  Chairman, essentially with
21   the quick pick game -- and I'll ask representatives
22   from Gtech to -- Gtech Texas to step to the table and
23   correct me if my explanation is inaccurate.
24                 But with the Pick 3 game, the quick pick
25   generation actually would take place on the display
0072
 1   screen.  So the number would generate and the clerk
 2   could see the number that came up when they hit the
 3   quick pick button before completing and sending the
 4   rest of the transaction through.
 5                 CHAIR CLOWE:  Before it's printed?
 6                 MR. ANGER:  Before it printed.  And that
 7   was unique to just that game.  None of the other games
 8   could you see the -- the random number that was
 9   generated, random number combination that was generated
10   prior to the generation of the ticket.
11                 That took place off screen.  And so when
12   the ticket generated, that's the first time you would
13   see it.
14                 CHAIR CLOWE:  So did the players then
15   look at the screen with the attendant and the number
16   would come up on the screen and they would say "No, I
17   don't want that" and then there would be another quick
18   pick display?
19                 MR. ANGER:  That is my understanding.
20   Gtech -- we presented this information to Gtech.  They
21   went out and spoke to stores out in the field and
22   communicated with their service representatives out in
23   the field.  And that is the explanation that came back.
24                 So Gtech suppressed the functionality
25   that would allow that quick pick number to be generated
0073
 1   and shown on the screen prior to the transaction.  And
 2   when they did and they provided new data to Dr. Eubank,
 3   this anomaly that he showed on a prior screen that
 4   triples came through at a higher rate disappeared and
 5   everything -- and I'll defer to Dr. Eubank that
 6   everything returned to a normal distribution.
 7                 CHAIR CLOWE:  And I've never heard of
 8   that before.  And I assume -- I believe I heard you say
 9   that's the only game where a preview, if I could call
10   it by that title, appeared before the ticket was
11   printed and a player might be allowed to reject the
12   quick pick and go on and make additional and other
13   choices.
14                 MR. ANGER:  That is correct.
15                 CHAIR CLOWE:  And that's been removed?
16                 MR. ANGER:  It has.
17                 CHAIR CLOWE:  Okay.
18                 DR. EUBANK:  So as Michael said, after
19   the -- with the data that -- that we received after
20   this functionality had been removed, you don't see --
21   so this is basically the same sort of thing.  We have
22   -- I have now an upper and lower no-fly zone because I
23   want to look for things in either direction.
24                 And you don't see the triples again.  Now
25   this is, I think, the number 4-1-6.  And that -- this
0074
 1   is simultaneously.  Every one of these one thousand
 2   things should be in here.  So this is something that
 3   shouldn't have happened either.
 4                 But it's completely different from what
 5   we saw a second ago.  So I -- I'm not going to talk
 6   about that in any depth.  I don't have any explanation
 7   for it other than to say that this is a piece of
 8   evidence and sort of move to the -- to the next one.
 9                 Now so for Cash Five, nothing that I did
10   revealed anything of substance, any sort of source of
11   thing that's out of the -- sort of the norm for what we
12   would expect from a statistical perspective.
13                 For the Texas Two Step data, there was no
14   problem with the simulated data.  But for the live
15   data, there were tests that I did on bivariate marginal
16   distributions that showed up having a P-value of .1.
17                 Now what is a P-value?  A P-value --
18   we'll talk about this in a little bit, but let me go
19   back and say it again.  It's basically the observed
20   level of significance.  So when I say that something
21   has a P-value of .01, I'm saying that there's about a
22   one-percent chance that something like this would
23   happen if everything was working exactly like you would
24   expect it to.
25                 I -- my cutoff is five percent.  And so
0075
 1   anything below five percent I call statistically
 2   significant.  So this is statistically significant in
 3   the sense of what I think of as evidence beyond a
 4   reasonable doubt.
 5                 Now what is the bivariate marginal
 6   distribution?  Let me talk about that for just a
 7   second.  This is on the primary boards for Texas Two
 8   Step.  So that's four numbers taken from 35 without
 9   replacement.
10                 The univariate marginal distributions
11   mean I look at the number -- I look at the number 1, 2,
12   3, 4, all the way to 35.  And I see how much -- how
13   often they come up.  Bivariate says I look at 1 in
14   combination with 2, 3, 4, how many times on a board did
15   I see 1 with 2, 1 with 3, 1 with 4, all the way to 1
16   with 35, 2 with 3, 2 with 4, all the way up to 2 with
17   35 and so on.
18                 There's about 600 of these combinations
19   that you look at.  And you can test under the model
20   that everything's being generated with -- with -- all
21   the boards appear with uniform frequency.  All these
22   pairs have to appear with uniform frequency, as well.
23                 So there was indication that that was not
24   being -- that those boards were not -- that -- the
25   quick picks generated for the primary boards for Texas
0076
 1   Two Step game were not being generated with this
 2   uniform frequency, at least in terms of the way the --
 3   the data showed on the bivariate distributions,
 4   bivariate margins.
 5                 So for the -- the Lotto Texas game, none
 6   of the live data sets had any issues.  But for the --
 7   both the simulated data, some unusual things happened.
 8   For the data from the GBT terminal, there were issues
 9   that arose with both the univariate and bivariate
10   marginal distributions for the primary boards, the
11   P-values that fell below my five-percent cutoff, about
12   two percent, two and a half percent.
13                 And then for the SST terminal data, there
14   were the -- when you looked at the bonus numbers that
15   appear on the generated boards, those showed up as
16   being statistically significantly different from
17   uniformity at something along the one-percent level.
18                 All right.  And finally with Mega
19   Millions, nothing untold showed up.  So let me
20   summarize and then try to say a few other words that
21   maybe puts this in a -- in a proper context.
22                 First of all, what -- if you look back at
23   everything that happened, we've certainly seen some
24   unusual things that we wouldn't have expected to see.
25   And it's not only unusual in a statistical sense but
0077
 1   it's also, I have to admit, unusual in my experience
 2   working with random number generation.
 3                 Certainly when I signed on to -- to do
 4   this project, I had no idea that I would see some of
 5   the things that I saw.  And there have been some -- a
 6   number of hours of sleep lost in my life as a result of
 7   that.
 8                 But to say this -- to put this all in
 9   sort of a global perspective, you have to actually talk
10   about what unusual means here.  And unusual certainly
11   depends on the level of significance that you use.
12                 Gtech, in terms of the analysis that they
13   do, they -- their level of significance that they use
14   is about -- and depends on how you say that.  But to
15   get a -- a correct comparison to the sort of way that I
16   did my test and the way they do things, their level of
17   significance was about -- is about a half a percent.
18                 And so is it that I'm just being too
19   picky with my five percent?  And I think the answer is
20   here that neither one of us is -- we're really doing
21   different things.  So one can't say that half a percent
22   is something that you could use all the time or
23   necessarily that five percent is something they should
24   use or that half a percent is something I -- I should
25   use.
0078
 1                 So the way I think about significance
 2   here, the only way I think you should even think about
 3   significance, levels of significance, is in terms of
 4   repeated sampling.  So the idea is that, if I could do
 5   this thing again and again and again, then if I'm using
 6   a five-percent level of significance, about five
 7   percent of the time I'm going to see something unusual
 8   that's going to happen.
 9                 Now in the real world, most of the time
10   we only get one sample.  And I would claim, to at least
11   a certain extent, that I only had one sample here.  I
12   had -- and I'm going to waffle on that in a minute.
13   But for right now, I claim that, in terms of SST for --
14   the SST terminal data for Lotto Texas, I only had one
15   of those.
16                 I didn't have 200 thousand of those.  I
17   only had one.  And if that's the case, then you have to
18   make a call.  So I've got something that's significant
19   at the five-percent level.  What's going on?  Is
20   this -- is the -- is the defendant guilty?  Is there
21   evidence beyond a reasonable doubt?
22                 I don't necessarily mean that I'm sure
23   that they're -- they're guilty, but I'm sure beyond a
24   reasonable doubt or not.  In trial by jury, you can't
25   say well, okay, let's let him go and see if he does the
0079
 1   crime again.
 2                 I mean, no -- nobody's going to do that.
 3   And that's the way data analysis is.  You don't get
 4   another run at it.  So as a rule, you only see it one
 5   time.  And from that perspective, something like the
 6   five-percent level is a reasonable thing to do.
 7                 Now the unique thing about random number
 8   generation is that it's -- it's the one case that I
 9   know of that's completely different from the -- what
10   mostly happens in the real world.  It's one case and
11   it, of course, is a real-world scenario, too.
12                 It's the one case that I know of that's
13   different in the sense that you can generate data --
14   and by the way, I'm from Texas A&M.  I have to say
15   this.  You can generate that until the cows come home.
16   And nobody's going to stop you.
17                 You own your random number generator, you
18   can just keep on doing it.  I wrote a random number
19   generation routine.  I took one -- one of these
20   off-the-shelf things, got it out of the textbook, wrote
21   it up in a couple hours, had data going, and I
22   generated data set after data set after data set that
23   would have been -- you could view as being the primary
24   boards for Texas Two Step.
25                 And by doing that again and again and
0080
 1   again, I can prove to any level of significance that
 2   you want whether my random number generator is working
 3   or not.  That's -- Monte Carlo simulation is the only
 4   place that I know of that you can just keep on doing it
 5   and, if you want a certain level of accuracy, you can
 6   get there.  Just let that computer run long enough.
 7                 So the -- the problem here is that, with
 8   the way things work for Gtech, there are some technical
 9   difficulties with doing that.  So you can say, "Why
10   doesn't Gtech just do that and just prove to us beyond
11   any doubt that their random number generator's working
12   the right way?"
13                 There's technical difficulties with doing
14   so.  So I -- I view this as -- it's just -- it's --
15   it's -- it's a problem, in a -- in a certain sense,
16   that it makes life a little bit more difficult than it
17   should have been in that there -- there's sort of a
18   double-whammy here.
19                 There's not a random number generator
20   that's being used that's one of these things that's
21   been studied extensively in the public sector with --
22   with lots of literature and backup on it.  And on the
23   other hand, because of the way that the mechanism is --
24   has been created for generating these quick picks, you
25   can't actually say, well, I'm just going to show you
0081
 1   that it works.
 2                 So instead, we're -- we're left with a
 3   situation like the one we've got where I looked at some
 4   data.  I found some stuff that's strange.  And, you
 5   know, does it mean there's something not right or -- or
 6   what?
 7                 Now -- now I'll waffle on what I said a
 8   minute ago.  So I said that I only had one sample.
 9   Well, that's true but it's also not true.  I actually
10   had 17 different data sets that I looked at.  And I
11   looked at a whole lot of statistical tests.
12                 So I found a number that were
13   significant.  But is it because sometimes strange
14   things happen?  Could it be that I -- I looked at a lot
15   of tests.  It came up that we saw several of these
16   things that -- that were significant.
17                 And so we've seen some evidence.  And the
18   question is whether or not -- is the evidence
19   sufficient to convict?  Now so how does a person go
20   about doing that?
21                 Well, it turns out that this happened at
22   a opportune time because there had been a good bit of
23   interest in something lately that has to do with
24   statistical genetics and connections to micro ray
25   study.
0082
 1                 There's something that -- that's gotten a
 2   lot of interest in the literature lately called working
 3   with false discovery protection.  So the idea is you're
 4   doing a lot of tests and you want to be sure at some
 5   level, which is roughly like the level of significance,
 6   of not -- you want to ensure yourself that you don't
 7   get too many false discoveries.
 8                 There's a number of procedures that are
 9   out there for doing that.  And I instituted one of
10   these.  It's something that was developed by Benjamini
11   and Hoshberg in the Journal of the Royal Statistical
12   Society in the last four or five years.
13                 Now what this does is, it's a procedure
14   that you use.  And you say -- you put all your tests
15   together.  And it gives a lower bound for the P-values
16   for these tests.  And if the -- the P-values fall below
17   these lower bounds, then you say I've seen something
18   and the defendant is guilty beyond a reasonable doubt.
19                 Now I ran this at the -- the five-percent
20   level in terms of -- I mean, I controlled my false
21   discovery at the five-percent level.  And so I'll --
22   I'll cut to the chase here and I'll show you the -- the
23   results of this.
24                 Now what I'm showing here, the blue
25   line -- wow, that showed up great.  Didn't it?  Okay.
0083
 1   There they -- let me just -- so watch my -- my corner
 2   here.  There's the blue line.
 3                 And there is the red line down here.  The
 4   red line is the Benjamini and Hoshberg lower bound.
 5   Now you can't really see it down here, but the blue
 6   line is just barely above the red line.
 7                 If it had fallen below the red line, then
 8   I would have said that, with false discovery controlled
 9   at five percent, that we've actually seen something.
10   So on the basis of this, if we -- if we put it all
11   together -- and let me just say that there -- that
12   Benjamini and Hoshberg is one of the standards here.
13                 There's criticisms of it.  It's very
14   conservative.  So it's the -- it's -- it's setting --
15   in some ways, it sets its reasonable doubt, in terms of
16   the way it does things, a little bit -- makes things a
17   little bit more stringent, a little bit harder to get
18   across that bound.
19                 But it's one of the standards in the
20   area.  So there -- it's certainly a reasonable
21   technique to have used here.  But using that, one
22   cannot say with certainty, at least not at this -- this
23   -- with a five-percent level of protection for false
24   discovery, you cannot say with certainty that what we
25   saw here wasn't just a result of the fact that we did a
0084
 1   lot of statistical tests.
 2                 So I think we found a lot of evidence.
 3   So if I had to summarize this, we found a lot of
 4   evidence but maybe just not enough evidence to come up
 5   with a guilty verdict.  So my verdict on this is not
 6   guilty or I -- I haven't found enough statistical
 7   evidence to say that this quick pick mechanism is not
 8   working exactly like it should.
 9                 The -- and so maybe the last thing that
10   should probably be brought up here has to do with --
11   we've talked about statistical significance.  At the
12   very beginning, I talked about practical significance.
13                 So let's suppose that -- that we found
14   some other approach and we did -- that -- that this was
15   statistically significant.  Would there be any
16   practical significance here?
17                 And my feeling is that there would be no
18   practical significance.  Whatever's happening here, if
19   anything is happening, is extremely subtle.  And any
20   implications it might have, either positive or
21   negative, for the lottery or the lottery's players I
22   think are of no practical consequence and probably not
23   something that could be quantified in any way.
24                 COMMISSIONER COX:  I'm going to step out
25   there and ask one question.
0085
 1                 DR. EUBANK:  Okay.  Skewer me.
 2                 COMMISSIONER COX:  Actually, two
 3   questions.  How many hours did all this take?
 4                 DR. EUBANK:  I was 35 years old when I
 5   started this.
 6                 COMMISSIONER COX:  Wow.  You indicated
 7   early on that, if I understood it correctly, there are
 8   lots of public domain, well-tested random number tables
 9   out there.
10                 DR. EUBANK:  Yep.
11                 COMMISSIONER COX:  And that Gtech had not
12   chosen to use one of those and, had they chosen to use
13   one of those, your work would have been less difficult.
14                 DR. EUBANK:  Certainly.
15                 COMMISSIONER COX:  Is that an ongoing
16   problem that we would want to ask Gtech to think about
17   reconsidering the random number table that they use or
18   is it a one-time thing that's behind us now?
19                 DR. EUBANK:  So my -- my feeling is that
20   it -- there's no reason to ever put yourself in a
21   position where you can be questioned.  So if you ask me
22   to create a random number generation mechanism, I would
23   go out and get one that -- that was -- was out in
24   the -- in the literature and I'd use it because nobody
25   could ever cause it -- nobody could ever bring what I
0086
 1   was using into doubt.
 2                 All right?  The -- the other side of this
 3   has to do with this -- there's modern technology there.
 4   I don't know -- I'm speaking out of school because I
 5   don't know the -- all the details.  But I do know that
 6   the -- that the number generation mechanism that's
 7   being used by Gtech is something that's been around for
 8   quite some time.
 9                 There are modern techniques out here.
10   And I -- my feeling is that -- you know, as a person
11   who lived in Dallas and used to drive on Central
12   Expressway, the longer you wait to fix -- to bring
13   something up into today's age, the more -- the harder
14   it's going to be when you finally get around to doing
15   it.
16                 So if -- if they -- if they would ask me
17   to -- should you use one of these new random number
18   generations, I -- the -- the price on them is just
19   right.
20                 Now in terms of implementing it, I know
21   that it's not cheap.  But it's never -- that's
22   something that has to be done, in terms of mechanics
23   and putting all this in the terminals.  And I know that
24   there's cost involved here.
25                 But I think in the long run -- you know,
0087
 1   this is just my opinion that a layman who knows nothing
 2   about the business side of things, my opinion is
 3   there's no reason to ever put yourself in doubt.
 4                 There's wonderful things that are out
 5   there that nobody could ever ask -- ever could --
 6   nobody could ever doubt you.  And if they had told --
 7   if they said we use the -- this Mersenne Twister, for
 8   example, I'm not sure that this whole analysis would
 9   have ever made much sense.
10                 COMMISSIONER COX:  Thank you, Dr. Eubank.
11                 CHAIR CLOWE:  Well, I can see you
12   gentlemen from Gtech are just salivating at the
13   opportunity to comment on this.  Who wants to be first?
14                 MR. GOODRICH:  For the record, my name is
15   Ron Goodrich.  I'm a software technology leader in --
16   in the -- living and working here in Austin at the
17   Gtech Austin Technology Center.
18                 And I did bring a little comment I wanted
19   to read from -- from the Gtech -- from the Gtech group.
20   Actually, that -- that would be the guys that worked at
21   corporate with Dr. Eubank and some of the folks in
22   Austin that worked in collecting the six hundred tons
23   of data that was -- was analyzed.
24                 First of all, they all wanted to thank
25   Dr. Eubank for the -- for working with -- with us and
0088
 1   with Gtech in examining this issue.  It has been --
 2   it's been a lot of work.  And particularly Tom Mormon
 3   and my associate, Chenghui here, spent a lot of hours
 4   working on it.
 5                 And we do agree that there's been a
 6   substantial exchange of data and discussion regarding
 7   the random number generation.  And it was through,
 8   actually, a test proposed by Dr. Eubank that we did
 9   discover that there -- you know, we agree that there is
10   some level of -- that although the selections are
11   random, there's some slight statistical deviations in a
12   couple of the games.
13                 And, you know, let's not forget some of
14   the other games did -- did just come right through
15   clean.  At this point, we want to emphasize that --
16   that this issue does not have any practical effects on
17   the citizens of Texas and -- who do elect to quick pick
18   their numbers.
19                 However, in agreement with Dr. Eubank and
20   an abundance of caution, Gtech -- and because it would
21   not be wise to -- well, let me back up.  Gtech has
22   begun review of our random number generator routines.
23                 And in an abundance of caution, we -- we
24   do not want to -- to rush something out into the field
25   that could end up just coming back on us with 10 times
0089
 1   the concern.  So we don't want to rush anything to the
 2   field.
 3                 However, work -- we wanted to make sure
 4   that you know that work has begun analyzing the best
 5   approach to -- and we're pleased to -- we will be
 6   reporting our -- our findings back to this commission
 7   in the future.
 8                 So that's -- that's where I leave it.
 9                 DR. EUBANK:  Could I just make one
10   comment, actually two comments?
11                 First, I want to second the statement
12   about the practical significance of this.  I've already
13   said that, but I want to emphasize that I -- I agree
14   with that.
15                 And the second part of it is that Gtech
16   has been really very forthcoming with all this.  It's
17   been a very -- it's been a -- there's been a lot of
18   mutually shared information.  And anything I've asked
19   for they've provided for me.
20                 CHAIR CLOWE:  Dr. Eubanks, could I
21   summarize what you told us in that my understanding is
22   that you have conducted extensive review, almost
23   exhausting.
24                 DR. EUBANK:  Yeah.
25                 CHAIR CLOWE:  And that you found some
0090
 1   things that you felt were of some level of significance
 2   that you asked Gtech to look at and they have agreed to
 3   do that and are in that process.
 4                 But in your opinion, there has been no
 5   anomaly or mistake that has impacted the players of
 6   Texas where they have played these games and taken a
 7   random generator or quick pick function.
 8                 DR. EUBANK:  That's correct.
 9                 CHAIR CLOWE:  I think the thrust of my
10   concern was, you know, have the players of Texas been
11   treated fairly.  And I think your answer is yes, but
12   there may be some issue here that has some level of
13   less-than perfection.  And we'd really like it to be as
14   close to perfection as possible.
15                 DR. EUBANK:  Yes, sir.  And that's --
16   that's -- that's -- if I could have known those words,
17   I could have cut 30 minutes out of it.
18                 CHAIR CLOWE:  You have struck me silent.
19   Thank you, gentlemen.
20                 Now?
21                 COMMISSIONER COX:  Yes, sir.
22                    AGENDA ITEM NO. XVII
23                 CHAIR CLOWE:  The Texas Lottery
24   Commission I move will go into executive session to
25   deliberate the appointment, employment, elevation (sic)
0091
 1   and/or duties of the executive director, acting
 2   executive director and/or deputy executive director, to
 3   deliberate the duties and evaluation of the Charitable
 4   Bingo Operations director and internal audit director
 5   and to deliberate the duties of the general counsel
 6   pursuant to Section 551.074 of the Texas Government
 7   Code, to receive legal advice regarding pending or
 8   contemplated litigation and receive legal advice
 9   pursuant to Section 551.071(1)(A) or (B) of the Texas
10   Government Code and/or to receive legal advice pursuant
11   to Section 551.071(2) of the Texas Government Code
12   including, but not limited to, Gametech International,
13   Inc. versus Greg Abbott, Attorney General of Texas et
14   al., Gametech International et al. versus Greg Abbott
15   et al., Charles Isbell versus Atkins et al., Red Men et
16   al. versus Gary Grief, Texas Lottery Commission versus
17   Joel Bowen and Associates, Inc., dba JB & Associates,
18   employment law, personnel law, procurement and contract
19   law, evidentiary and procedure law and general
20   government law, lottery operator contract, Mega
21   Millions game and/or contract.
22                 Is there a second?
23                 COMMISSIONER COX:  Second.
24                 CHAIR CLOWE:  Before we take the vote,
25   counselor, if we take up the Mega Millions game or --
0092
 1   and/or contract, do we need to ask for an exemption to
 2   deal with that contract?
 3                 MS. KIPLIN:  No.  The notice is solely to
 4   receive legal advice regarding.
 5                 CHAIR CLOWE:  So there's no deliberation
 6   of that point?
 7                 MS. KIPLIN:  No deliberation.
 8                 CHAIR CLOWE:  So this is within the
 9   notice?
10                 MS. KIPLIN:  Yes.
11                 CHAIR CLOWE:  And properly made as a
12   motion?
13                 MS. KIPLIN:  Yes.  It would be to receive
14   legal advice on that game and/or --
15                 CHAIR CLOWE:  Very good.  Very good.
16   Thank you.
17                 All in favor, please say "aye."
18                 COMMISSIONER COX:  Aye.
19                 CHAIR CLOWE:  Aye.
20                 The vote is two-zero.
21                 Texas Lottery Commission will go into
22   executive session.  The time is 1:35 p.m.  Today is
23   January 13th, 2006.
24                 (Executive Session from
25                  1:34 p.m. to 3:08 p.m.)
0093
 1                   AGENDA ITEM NO. XVIII
 2                 CHAIR CLOWE:  The Texas Lottery
 3   Commission is out of executive session.  The time is
 4   3:08 p.m.
 5                 Is there any action to be taken as a
 6   result of executive session?
 7                     AGENDA ITEM NO. XV
 8                 CHAIR CLOWE:  If not, let's move on to
 9   item number 15, consideration of, possible discussion
10   and/or action on external and internal audits and/or
11   reviews relating to the Texas Lottery Commission and/or
12   the Internal Audit Department's activities.
13                 Ms. Melvin?
14                 MS. MELVIN:  Good afternoon, Chairman,
15   Commissioner.  For the record, my name is Catherine
16   Melvin.  I'm the director of internal audit.
17                 I have one item of update regarding
18   internal audit.  I'm very pleased to bring to you the
19   results of a recently-completed quality assurance
20   review of the internal audit function.  The quality
21   assurance or peer review is required by professional
22   standards and by the Texas Internal Auditing Act.
23                 Our quality assurance review was
24   performed by Mr. Richard Tarr, T-A-R-R.  Mr. Tarr is a
25   highly-qualified, independent reviewer.  And if I may,
0094
 1   I'd like to read his overall opinion.
 2                 Based upon the work conducted, I concur
 3   with the conclusions made by the internal auditor that
 4   the internal audit activity of the Texas Lottery
 5   Commission fully complies with the Institute of
 6   Internal Auditors International Standards for the
 7   Professional Practice of Internal Auditing, the U.S.
 8   Government Accountability Office Government Auditing
 9   Standards and the Texas Internal Auditing Act.
10                 This opinion, which is highest of the
11   three possible ratings, means that policies, procedures
12   and practices are in place to implement the standards
13   and requirements necessary for ensuring the
14   independence, objectivity and proficiency of the
15   internal audit function.
16                 Okay.  And with that, I'd like to move on
17   to --
18                 COMMISSIONER COX:  Could I say something?
19                 CHAIR CLOWE:  Sure.
20                 COMMISSIONER COX:  Congratulations,
21   Catherine, to you and your staff.  That's a wonderful
22   report.
23                 MS. MELVIN:  Thank you, sir.  We're very
24   pleased.
25                 Regarding external audit, the State
0095
 1   Auditor's Office continues their work on two audits
 2   that are ongoing right now, the human resources audit
 3   and the required security audit.
 4                 In addition, the state auditors have sent
 5   notice of their intent to audit Lotto Texas.  We
 6   received that notice yesterday.  And so I would like to
 7   read their audit objectives regarding this audit as
 8   stated in their letter.
 9                 And in meeting with the auditors
10   yesterday, they indicated that it's always possible
11   that the scope may change.  But as of right now, these
12   are their stated objectives.
13                 The audit objectives are to determine
14   whether selected Lotto Texas activities are conducted
15   in accordance with commission rules, policies and
16   procedures.  This includes determining whether all
17   instances in which advertised jackpot amounts exceeded
18   estimated sales amounts have been identified, prize
19   payments were calculated and paid in accordance with
20   commission rules, policies and procedures, and that
21   amounts transferred to the Foundation School Fund were
22   calculated accurately and in accordance with Texas
23   Government Code 466.355.
24                 So we look forward to welcoming more
25   auditors to our office and certainly working with them
0096
 1   through that process.  Any questions about that?
 2                 COMMISSIONER COX:  Catherine, you did an
 3   audit of jackpots back to way back.
 4                 MS. MELVIN:  Yes, sir.
 5                 COMMISSIONER COX:  And you did an audit
 6   of transfers to the -- to the Foundation School Fund.
 7   Do you have any thoughts about why the state auditor
 8   might be looking at something that you've recently
 9   looked at and they have your report on?
10                 MS. MELVIN:  They -- I met with them.
11   And they certainly are going through my working papers
12   and will rely or will review my work and determine the
13   extent of reliance that they may place on that work.
14                 Certainly the State Auditor's Office has
15   the sole independence and authority to select whatever
16   scope they choose.  I believe that they made clear in
17   the beginning of the year that they would do some
18   additional work at the Lottery Commission, including an
19   audit of Lotto Texas, both jackpot prize payments and
20   the advertised amounts.
21                 COMMISSIONER COX:  So you have no
22   indication that this has anything to do with concern
23   about the adequacy of the work that you did?
24                 MS. MELVIN:  Oh, no, sir.  No, sir.  But
25   we -- you know, we'll make all of our working papers
0097
 1   available.  We're certainly happy to work with them and
 2   share with them the methodology we used.
 3                 COMMISSIONER COX:  Good.
 4                 MS. MELVIN:  Okay.  Finally, the State
 5   Auditor's Office recently completed an audit of the
 6   agency's financial statements for fiscal year 2005.
 7   I'd like to invite Mr. Michael Apperley from the State
 8   Auditor's Office to join me.
 9                 MR. APPERLEY:  Good afternoon, Chairman
10   Clowe, Commissioner Cox.  My name is Mike Apperly.  I'm
11   an audit manager for the Texas State Audit Consults.
12   I'm the contact manager for the Texas Lottery
13   Commission.
14                 So any questions you have on audits at
15   the Lottery, you can funnel those through me.  And
16   Catherine has my contact information.  I can get that
17   to you if you like.
18                 Also, I was the audit manager for the
19   annual financial report opinion that was recently
20   conducted.  This is the first year that the State
21   Auditor's Office has actually opined on the financial
22   statements.
23                 In the past, the Lottery Commission has
24   used public firms.  And the new state auditor decided
25   that he would like a little more control or input into
0098
 1   the audit process.  So we decided that we would this
 2   year opine on the financial statements ourselves.
 3                 The results of that audit basically were:
 4   We looked at the financial statements.  We looked at
 5   the internal control surrounding financial reporting.
 6   And we -- and we found there were no reportable
 7   conditions in the risk control or in the control
 8   systems over financial reporting.
 9                 We also found the financial statements
10   were free from material misstatements and basically
11   represented the financial position as of August 31st,
12   2005 and the financial activity that occurred in the
13   year leading up to August 31st, 2005.
14                 If you have any questions, I'd be glad to
15   answer those.
16                 COMMISSIONER COX:  No, sir.
17                 CHAIR CLOWE:  Mr. Apperly, we're very
18   happy that you're here and that you're designated as
19   our contact with the State Auditor's Office.
20                 And as I told you in private when
21   Catherine introduced you to me, we want to be very
22   responsive to any requests that you have, any
23   information or whatever you might need in the way of
24   your work.
25                 This agency, from time to time, is
0099
 1   criticized for many of the activities that it's
 2   involved in.  It is a controversial agency because of
 3   the work that we do.  And we're respectful of that
 4   criticism.
 5                 But at the same time, we feel that you
 6   are a great asset to this agency by way of your
 7   independent audit, so many things that you can look
 8   into and opine on.
 9                 So I want you to know that we're very
10   happy that you're here and that we would like to work
11   closely with