Thomason progression
  • Desert Dog said:
    [quote=Renzey]Walter/Desert Dog: If Dog is betting according to the schedule laid out in the book, he'll be averaging $21 per bet with about 82% of all bets being $10. I have to say though, that 80% penetration is uncommon, and will produce better results than most "real game" scenarios. 70% is more typical. In fact, at 80% penetration, his average bet will actually be somewhat larger since so many "20+" counts will occur in that extra 10% of penetration.

    You're right. The last few hands between 30% and 20% are often the ones with the highest counts. I'm going to finish this series at 1000 hands and then for the next series I will watch for the "30% left" message and hit the button that will reshuffle then rather than wait for the automatic reshuffle at 20%.[/quote]

    Here's the 1000 hands results, using the KISS II count, six deck shoe, 80% penentration, min bet $10, max bet $100.

    Chips Bought: $ 500
    Winnings: $2930
    Ending bankroll: $3430
    Hands won: 443 (44.3%, or 49.6% excluding pushes)
    Hands lost: 451 (45.1%, or 50.4% excluding pushes)
    Pushes: 106 (10.6%)

    Up nearly $3,000 in a week, and 1000 is reasonable for the actual number of hands you can play in a week. Looks like a decent living. I'll send you all postcards.

    Next series I'll reshuffle immediately after 70% of the cards are dealt.
  • Thanks Desert Dog!
  • Jeff

    I am using an old Avery Codoza program. Have had it 8 or 9 years. I like it.

  • DAY 13 Another loss, but not so bad this time.

    (1) Total time played 29hrs.-39min.
    (2) Total bets made 3136
    (3) Total amount bet $47395
    (4) Average bet $15.18
    (5) Total amount lost $315
    (6) # of four straight wins (5) three 4 times, one 5 times, one 6 times.
    (7) # of four straight losses (7)

    Keep having more losing streaks than winning streaks. Should have a winning day some time.

  • Tuffy - That is still pretty d**n good. You are only losing one unit per hour after 30 hours........and my thanks to you for doing this. I think everyone appreciates it.

  • TUFFY: Your current status is -0.66%, but the "luck factor" still has a pretty big influence at this stage of the game (within $1100 of where you are two times out of three).
  • TUFFY: Also be aware that you should lose 4 in a row three times for every two times you win 4 in a row.
  • Renzey - I came up with the same number....-0.66, but I am going to wait until he gets to 5K to calculate the SD......I agree with the "luck" factor, but will stand by "looks decent" so far.......Not a "winner", just looks decent.

  • Renzey

    I have often wandered what one standard diviation was in blackjack. I did not realize it was as high as $1100 with my level of betting. I know it is fairly easy to figure with stocks if one uses a tight index fund asset allocation. Still it is nice to know it is that wide in blackjack.

  • Desert Dog said:
    The initial buy-in of $500 is now up to $2,600.

    Could someone please actually define "buy-in"? I know this sounds pretty dumb, but I always assumed it was the initial amount of chips you bought. However, hearing it in the above context, makes me wonder if I'm mis-understanding. Thanks.

    (BTW, I'm still having good success with the Labby combined with Mr. Thomason's quit points. I really think more folks should give this a serious try)
  • jm: The initial buy-in is just that... the amount of money that you invest in chips at the start of play... not to be coufused with "bankroll" -- the amount of money you're willing to lose for one session.
  • DAY 14 At last a winning day

    (1) Total hours played 31hrs.--51min.
    (2) Total bets made 3382
    (3) Total amount bet $51067
    (4) Average bet $15.16
    (5) Total amount LOST $148
    (6) # of four straight wins (5) five Three four straight, one five straight, one six straight.
    (7) # of four straight losses (6)

    Ahead $167 today. Cuts the overall loss down. Tomorrow will be the halfway mark. With the standard deviation at $1100 the volitity has been very low these 14 days. Will see about the next 16 days. Also the average bet has hugged the flat line at $15.
  • Tuffy - Your SD is going to change as your number of hands increases. I imagine that $1,100 Fred posted yesterday was for the hands played to date......but I didn't personally check it.

  • tuffy: When I run sims using a $20/30/40/50 progression, my long-term average bet is $31.80. I think that you will find that winning sessions show a higher average bet, and that more winning sessions will raise your average bet to about $15.90. And, even though your initial bet in the progression is always $10, you would have to bet $14 as an initial bet if you were just flat betting every time.
    Also, I'm surprized that you haven't yet hit a few 8, 9, or 10 hand winning streaks.
    Keep up the good work!
  • Fred: Could you please show the math behind your "frequency of winning and losing streaks" post? Thanks.
  • Yep, Fred.....Please give us your math. I've been chasin' my behind today and haven't had a chance to confirm your numbers, but I thought that "ratio" would be about 5:4 in lieu of the 3:2 you posted.

  • WALTER and GRIFTER: Woops! My apologies-- went too fast. Grifter is right. The ratio of 4 consecutive losses to 4 consecutive wins is just about 5-to-4. The simplest but slightly imperfect way to get this is to group your streak together, then border it at the beginning and end by a streakbreaking outcome.
    For example, the probability of a win is .435, a loss is .480 and a push is .085. Four straight wins (counting pushes as a streakbreaker) would then be;
    .565 x .435 x .435 x .435 x .435 x .565 = .0114.
    Four straight losses would be:
    .520 x .480 x .480 x .480 x .480 x .520 = .0143.
    So in Tuffy's 3382 hands, it would be normal for him to have (.0114 x 3382) 38 winning streaks of exactly four in a row and (.0143 x 3382) 48 losing streaks of exactly four in a row.
    An occasional poster on BJMath named "Lefty" gave me a much more complex, exact equation for calculating this, but once you get over a sample size of a few hundred, it virtually doesn't matter.
  • Fred -- more questions:
    1. What's the origin of ".565" and ".520"?
    2. Can you recalculate and remove the "push" from the calculation... treat the push as a "non-hand"?
    3. Can you determine average bet size for each series, based upon tuffy's $10 to $25 progression? I realize this could be difficult because the preceeding bet for the progressive player (the loss prior to starting a four-hand winning streak) could be anywhere from $10 to $25, but...
  • WALTER: 1) .565 is the probability that this hand was either a loss or a push (.480 + .085). .520 is the probability that this hand was either a win or a push (.435 + .085).
    2) You could treat a push as a non-hand by counting every result as either a .435 (win) or a .480 (loss). But in that case a sequence of;
    L, W, P, W, P, W, P, W, L would be counted as four consecutive wins. Is that how you work your progression?
  • DAY 15 Pretty evan today, but a small loss. The half way day.

    (1) Total hours played 33hrs.--53min.
    (2) Total bets made 3614
    (3) Total amount bet $54557
    (4) Average bet $15.11
    (5) Total amount LOST $161
    (6) # of four straight wins (3) two four straight, one five straight.
    (7) # of four straight losses (2)

    Pretty even day. Was ahead by a small amount most of the 2 hours until the last 5 minuets. When the shoe ended then I quit. Which I normally do after two hours when either the shoe ends or I have 4 straight losses. Not many streaks of either wins or losses today.

    I sometimes wonder about quiting on four straight losses. It did not happen today but more often than not I don't finish the shoe before having four straight losses. And after two hours that is usually when I quit: either at the end of the shoe or four straight losses, whichever comes first.

  • Fred: The answer to your question is "yes". I consider a push as a non-hand, since no money changed hands. Same is true of split hands when I win one and lose one. But if I lost one side of a split hand, and won the other hand which was also a double down, I'd count it as a win since there was a net profit.
  • DAY 16

    (1) Total time played 36hrs.---07min.
    (2) Total bets made 3847
    (3) Total amount bet $58166
    (4) Average bet $15.12
    (5) Total amount LOST $216
    (6) # of four straight wins (4) three four straight, one five straight
    (7) # of four straight losses (5)

    One point I forgot to mention. In the software I use the dealer takes all .50 cent payoff's on blackjacks as a tip. He also at times takes $1.00 as a tip. Stats do not record the total amounts, but today it was $3.00 and yesterday it was $3.50. I think about $3.00 for the 2 hours of play is about what it is. These amounts are subtracted from the wins before the amount is added to the bankroll. As I said I think about $3.00 to $3.50 is the amount each day. That seems about right for a tip, so I guess it is right to charge that amount against the total wins. The total amount LOST reflects those deductions from win's.

  • DAY 17

    (1) Total time played 38hrs.--21min.
    (2) Total bets made 4082
    (3) Total amount bet $61516
    (4) Average bet $15.07
    (5) Total amount LOST -$242
    (6) # of four straight wins (4) two four straight, two six straight
    (7) # of four straight losses (6) six

  • Day 18 Bad day.

    (1) Total hours played 40hrs.--25min.
    (2) Total number of bets 4308
    (3) Total amount bet $64766
    (4) Average bet $15.07
    (5) Total amount LOST $508
    (6) # of four straight wins (2) two Two four straight.
    (7) # of four straight losses (8) eight. Glad today is over. Would have changed tables 8 times in 2hrs. & 11Min.

  • DAY 19

    (1) Total time played 42hrs.--36min.
    (2) Total bets made 4546
    (3) Total amount bet $68386
    (4) Average bet made $15.04
    (5) Total amount LOST _$431
    (6) #of four straight wins (4) four two four straight, two five straight
    (7) #of four straight losses (3)) three

    A little better today, but still down more than the house edge calls for. Well within one standard diviation though.

  • DAY 20

    (1) Total time played 44hrs.--45min.
    (2) Total bets made 4782
    (3) Total amount bet $72016
    (4) Average bet made $15.07
    (5) Total amount LOST -$385
    (6) #of four straight wins (4) four 2 four straight, 1 five straight, 1 nine straight.
    (3) # of four straight losses (6)

    Two-thirds of the way through and the loss is almost exactly where the house edge says it should be. Lets see what the next 10 sessions will produce.

  • DAY 21 Bad day

    (1) Total hours played 46hrs.--48min.
    (2) Total number of bets made 5042
    (3) Total amount bet $75756
    (4) Average bet $15.03
    (5) Total amount LOST -$670
    (6) # of four straight wins (6) 5 four straight, 1 five straight.
    (7) # of four straight losses (10) ten (ouch)

    There's always tomorrow.

    PS: To anyone just looking in here I am playing this on my computer. Not betting real money.
  • Using KISS II count on my computer, 6 decks, heads-up, reshuffle when 30% left in shoe, (vs previously letting software default to reshuffle at 20% which is unrealistic), low bet $10, high bet $100, results after 666 hands (this wasn't deliberate, just where the latest shoe ended):

    Initial buy-in: $1000 (in two $500 installments -- needed to spend a second $500 when chips got low)
    Current bankroll: $2,425 meaning I am $1,425 up.
    Hands won 275 (41.3%)
    Hands lost 325 (48.8%)
    Pushes 66 (9.9%)

    Desert Dog
  • Dawg - What is a spread like that (1-10) giving you as an average bet with this method?
  • Grifter said:
    Dawg - What is a spread like that (1-10) giving you as an average bet with this method?

    Grifter -- Fred can probably tell us the answer. My software doesn't track it, and I'm too lazy to write down each bet. All I can say is most of the time -- probably more than 75% -- the bet is $10 because the count is 19 or lower. Just as a recap, Fred's betting schedule for six decks is: 19 or less = 1 unit, 20 = 3 units, 21 = 6 units, 22 = 9 units, 23 or more = 10 units. I'd estimate that in this current simulation, half of all the bets greater than $10 are $100 bets. There have been great shoes, shoes from hell, and a lot of really boring shoes. And a few walk aways when the count was 4 or less a certain distance in. This move is as important and increasing your bet in a high count.

    In real play I'd bet a more conservative ramp, 5 units at the high end. And then after a while quit counting altogether and order a drink.
  • Gracias, Senor Dawg - I can get pretty close with those numbers, but yes it would be better if Fred sees your post and responds.

    Yessir.....I feel the same way about counting 6D. In fact, I don't even start with a count, I just begin by drinkin'. :wink:

  • Am I correct in calculating that Tuffy / Thomason's results are (to this point) working out to be about double the expected house edge?
  • Zebra - Yep, he is close to double right now, BUT he had a very bad day today. At the end of yesterday he was very close to 'right on' with the EV.

  • Crackers,

    You were close on your call that my 3rd visit mght be a loss but it was actually my 4th visit. I walked out about $400 down but on my 5th visit I left ahead $800. How often do you go to Cherokee. I try to go once a week atleast. maybe Ill see you there. Todd.
  • tuffy/Charles: Now that you've reached the 5,000+ bets point, I'm trying to compare your simulated results to the 5,000 hands of manually-dealt play contain in my progressions book. The first data that strikes me as strange is comparison of number of four or more consecutive wins. I show that you've had 67 "strings" of 4 or more consecutive wins, and since you didn't keep track of these numbers for the first 3 sessions I've added 15 more strings to your score, making your record 91 strings of four or more consecutive wins. I compared this record to my 5,000 hands, and I had 167 strings of four or more consecutive wins... a big difference!
    Conclusions? None at this point, but the results seem to support my contention that strings of consecutive wins occur more frequently in live play than they do in simulated play. But it could also be that your player just had lousy luck for this number of sessions... without some knowledge of win/loss percentages, it's impossible to tell how a flat bettor would have fared when compared to your progressive bettor.
    Your efforts are appreciated. Perhaps another poster could run a similar study where win/loss records are known... And later on I'll post some results from studies of other sets of manually-dealt cards.
  • Walter

    Thanks for your post. It also seems to me that I am getting a lot of four string losses compared to four string wins. Maybe some evening out in the last 9 days. Hope so. I did not know until you told me what it should be. I have just been playing and recording the results, what ever they turn out to be. Todays were a little better. Will post them now.

  • DAY 22

    (1) Total time played 48hrs.--58min.
    (2) Total number of bets made 5290
    (3) Total amount bet $79628
    (4) Average bet $15.08
    (5) Total amount LOST $633
    (6) #of four straight wins (6) six. 4 four straight, 2 five straight.
    (7) #of four straight losses (4)

    A little better today.

  • Walter- I went through 5,000 of my hands, looking at 4 wins and 4 losses in a row. All were hand played, using six decks.
    Results per 1,000 hands (W/L)
    22/28 * 21/24 * 30/37 * 30/40 * 24/38
    Total Win 4 127 Lose 4 167

    I have the results of several thousand hands, if you want me to do more, just let me know.
  • "Crackers,
    You were close on your call that my 3rd visit mght be a loss but it was actually my 4th visit. I walked out about $400 down but on my 5th visit I left ahead $800. How often do you go to Cherokee. I try to go once a week atleast. maybe Ill see you there. Todd."

    Todd, my husband and I are there about once a week also. I'm the old gal with the black Harrah's jacket on. Introduce yourself. I play the BJ "machines" next to the Dealer tables about 50% of the time. Jackie :P
  • DAY 23 Good day.

    (1) Total time played 51hrs.--12min.
    (2) Total bets made 5536
    (3) Total amount bet $83412
    (4) Average bet $15.21
    (5) Total amount LOST $479
    (6) #of four straight wins (6) SIX. 3 four straight, 2 five straight, 1 six straight.
    (7) # of four straight losses (4) FOUR

    Things went well today. Wins were steady. Streaks favored the wins. The loss now pretty close to the expected house edge. Stayed within one standard deviation for the entire 23 days, although the standard deviation is narrowing now as I understand it. Seven more days. It would be ironic if the final loss (?) was to end up within 10 to 20 basis points of the expected house edge.

  • To Grifter, Dog, Walter and Midnight:
    1) A 1-to-10 spread with the KISS II will average about a 2.1 unit bet.
    2) The striking difference between Tuffy's 91 streaks of 4 or more wins and Walter's 167, both in 5000 hands jumps way out. One standard deviation for this is right around 10 such streaks within 5000 hands, so you should virtually never see 91 one time and 167 another time. We need to clarify first whether the same Win/Lose/Push criteria was used in both experiments to define what consitutes 4 or more in a row. If the methodology was the same, then we have to seriously question whether computer dealt hands and manually dealt hands perform the same. Then again, the 10,000 manually dealt hands I checked from Bob Hubby's "Blackjack Tracker" conformed very closely to mathematical probability. We've got to be very careful here. We may need the manual experiment done all over again once we're sure everybody's playing off the same page.
    3) Midnite - Were your hands dealt manually or by computer, and what was your win/lose/push criteria to define a streak?
  • I'm at the 1000 hand point of the KISS II test, reshuffling when 70% of the cards have been dealt from a six deck shoe. In my last post at the 666 hand point I was way up. Now the story is different. There was a shoe that went to a count above 23 and stayed there, so I was betting $100 hand after hand, and mostly losing, or at best pushing. Very few wins. A few losses on doubles at this highest bet. Had to increase my buy-in by another $500 to $1500, and current bankroll is $1155, so I'm down $345. This was the worst negative fluctuation I've ever seen. Glad it's just a simulation.

    Hands won, 416 or 41%
    Hands lost, 486 or 48%

    I suspect that what will happen is that after another 1000 hands I'll be back in the win column. Will keep you posted.
  • Fred- My hands are all manually dealt and I have a bunch of them... I never consider pushes.
    I don't even record them as you neither win or lose. I used six decks. I looked only for 4 wins in a row or 4 losses in a row. I would be glad to do some more when we get it straight, as to how to count them. i.e. is a W-W-DW a win of 4 in a row ?
  • Midnite/Fred - Midnte, since you don't count pushes it would seem to me this would skew the results for this particular 'test' Fred is pursuing. i.e. For every 100 hands you show, the actual hands played would be about 108.

  • Day 24 Another good day.

    (1) Total hours played 53hrs.---23min.
    (2) Total number of bets 5766
    (3) Total amount bet $87072
    (4) Average bet $15.26
    (5) Total amount LOST $315
    (6) # of four straight wins (6) six. 3 four straight, 2 seven straight, 1 nine straight.
    (7) # of four straight losses (4) four.

    I don't know why the number of four straight wins is so low. I just record them as they play out.

    In counting 4 straight wins I:
    (1) Ignore pushes
    (2) Count double downs as only 1 win if I win. & 1 loss if I lose.
    (3) On splits: If I win both I count as 1 win. If push don't count. If I win one and lose one I count it as a push & don't count. For example: If I have 2 straight wins and push on anything I count the next hand as starting with 2 straight wins going for 3.
    (4) When shoe ends I stop counting wins, whever I was, and starting with the next shoe, begin over at 0 wins.

  • Tuffy - This is Renzey's "ball game" so I don't want to interfere, but I am so interested in this thread that here are a couple of my thoughts till he gets back to you.

    You said (in italics),
    "(1) Ignore pushes" .....This is correct IMHO.

    "(2) Count double downs as only 1 win if I win. & 1 loss if I lose." ...I agree. This is correct IMHO.

    "(3) On splits: If I win both I count as 1 win. If push don't count. If I win one and lose one I count it as a push & don't count. For example: If I have 2 straight wins and push on anything I count the next hand as starting with 2 straight wins going for 3." .....I have to disagree with this and this could skew the results. I have always counted splits as two hands....Two wins counts as two wins and two losses as two losses.

    "(4) When shoe ends I stop counting wins, whever I was, and starting with the next shoe, begin over at 0 wins.".....This is a real "biggie" as far as amount of streaks and I wonder if Renzey is taking this into consideration. I honestly don't remember how many players you are using in your simulation, but for instance if it is four then you are stopping any possible streak about every 16 hands......What a difference that will make with Renzey's stats.

    Sorry, didn't mean to get so long and let me say you are doing one h*ll of job with this test.

  • Grifter

    I have stopped with some win streaks many times when the shoe ended. If I had two wins, or three, or even one when the old shoe ended that streak would end there and I would start over with 0 wins the first hand in the new shoe. It happened once today with a 3 win streak. I end the shoe with a win about as often as with a loss. I think it happens quite often, but I don't have any figures as to how often. Also I play with 3 players including myself, but only keep track of my own play. I'm not sure how much difference that makes.

    Counting splits winning twice as two wins would add some to the winning streakes. Adding 2 losing splits would also add some to the losing streaks. I just haven't been tracking any of that. Hope this helps.

  • My hands will be for 100 "decisions" and not 100 hands, as I do not count pushes. With splits it may be a few more than 100.

    (1) (2) & (3) I agree with my Pal, Grifter

    (4) I did not mark where one shoe ended and a new one begain, so I can not start over with a new shoe. This can make quite a difference. i.e. I end the shoe with three wins and then win the first hand of the next shoe. You have 1 in a row and I have 4 in a row.

    Now is a string of 8 wins (or losses) in a row, counted as 1 win or 1 loss of 4 in a row or counted as 2 wins or 2 losses ?
  • I read Walter's book to say that a streak ends when a shoe ends. Winning or losing. If you win three and the shoe ends, for the first hand of next shoe you're back to the starting bet. (Page 18.) If you lose three at the end of a shoe and you lose the first of the next shoe, you're not at the four loss quit point, you're at one loss.
  • Guys and Girls: I think we are making some progress. Here's my take on how consecutive wins/losses should be recorded.
    1. I consider a win to be a winning BET, not just a winning hand, so a double-down win would be two wins. A win, a win, and a split with a successful double-down on one side of the split would be counted as five consecutive winning bets.
    2. Consecutive wins/losses end at the end of a shoe.
    3. Pushes don't count.
    4. Blackjacks count as one win.
    5. Even though two or more winning bets might occur from one initial hand, I don't jump up the progression when this happens. An increase from $10 to $15 would occur regardless of the number of winning bets that originated from the initial dealing of a hand.
    If you think about it, we don't play this game to win hands. We play to win MONEY, which means winning bets, so that's what we should track.
    Having said this, it still seems to me that tuffy's numbers are low, since splits and doubles should only add about 10% to the number of bets placed, and won or lost. In other words, 100 hands would normally result in about 110 bets being placed, due to splits and double-downs. Midnight's results and my results are generated from manually dealt play, which seems to support my contention that streaks of winning bets occur more frequently with manually-dealt play than they do with computer generated play. :D

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!