How much does Daniel Dravoit's upgrade to KO really improve SCORE for a typical shoe game? I'm wondering if you have any data you can share.
Also, would the indices for this style of KO be any different than KO out of the box?
MJ
How much does Daniel Dravoit's upgrade to KO really improve SCORE for a typical shoe game? I'm wondering if you have any data you can share.
Also, would the indices for this style of KO be any different than KO out of the box?
MJ
Haven't looked at it since it came out. His brother actually brought me the book. I don't remember the SCORE change. I think Ken Smith had some sims in the book. Not a great deal. The indices are simpler than KO's. It's the betting that's different. CVCX can Kelly optimize Dravot style betting strategies.
"I don't think outside the box; I think of what I can do with the box." - Henri Matisse
If you have CVData, but do not have a CVCX license; you can still call CVCX to obtain optimal Kelly betting. However, this doesn't work with Dravot's KO Color of BJ. That requires a CVCX sim since CVCX keeps EV and Variance data for all reasonable penetrations at once and can group the data over multiple depths and use the groupings for optimal betting calcs.
A bit difficult to explain. For an attempt from 12 years ago, see: https://www.blackjacktheforum.com/sh...-Based-Betting
"I don't think outside the box; I think of what I can do with the box." - Henri Matisse
BennGunn,
A 10% increase in SCORE is substantial in my opinion. Some people learn 50 extra indices for a 5% increase in SCORE. Do you find it difficult to remember how much to bet at each depth? I did not realize CVData can simulate Color of KO.
MJ
Last edited by MJ1; 05-24-2022 at 09:18 AM.
Norm,
Do you think Dravoit's insurance tweak makes sense? I believe he says to ignore the dealer ace for counting purposes when making the insurance decision. I can understand why you would want to disregard it however isn't this index value based upon a simulation which factors the dealer upcard into the count? Don says index values should be used in the same manner that they were simulated.
MJ
Running count insurance data using KO tags is fairly simple:
1 deck
2 decksCode:Count tags {1,-1,-1,-1,-1,-1,-1,0,0,1} Decks: 1, IRC = -4 Insurance Data (without regard to hand comp) No subgroup (removals) are defined **** Player hand: x-x **** Cards RC TC ref 48 -2 -2.17 38 -1 -1.37 20 0 0.00 5 1 10.40 4 0 0.00 3 1 17.33 2 0 0.00 1 1 52.00
4 decksCode:Count tags {1,-1,-1,-1,-1,-1,-1,0,0,1} Decks: 2, IRC = -8 Insurance Data (without regard to hand comp) No subgroup (removals) are defined **** Player hand: x-x **** Cards RC TC ref 96 -3 -1.63 95 -2 -1.09 63 -1 -0.83 32 0 0.00 7 1 7.43 6 0 0.00 5 1 10.40 4 0 0.00 3 1 17.33 2 0 0.00 1 1 52.00
6 decksCode:Count tags {1,-1,-1,-1,-1,-1,-1,0,0,1} Decks: 4, IRC = -16 Insurance Data (without regard to hand comp) No subgroup (removals) are defined **** Player hand: x-x **** Cards RC TC ref 192 -3 -0.81 191 -2 -0.54 190 -3 -0.82 140 -2 -0.74 92 -1 -0.57 47 0 0.00 7 1 7.43 6 0 0.00 5 1 10.40 4 0 0.00 3 1 17.33 2 0 0.00 1 1 52.00
8 decksCode:Count tags {1,-1,-1,-1,-1,-1,-1,0,0,1} Decks: 6, IRC = -24 Insurance Data (without regard to hand comp) No subgroup (removals) are defined **** Player hand: x-x **** Cards RC TC ref 288 -3 -0.54 281 -4 -0.74 222 -3 -0.70 163 -2 -0.64 109 -1 -0.48 56 0 0.00 9 1 5.78 8 0 0.00 7 1 7.43 6 0 0.00 5 1 10.40 4 0 0.00 3 1 17.33 2 0 0.00 1 1 52.00
k_cCode:Count tags {1,-1,-1,-1,-1,-1,-1,0,0,1} Decks: 8, IRC = -32 Insurance Data (without regard to hand comp) No subgroup (removals) are defined **** Player hand: x-x **** Cards RC TC ref 384 -3 -0.41 381 -4 -0.55 369 -5 -0.70 310 -4 -0.67 240 -3 -0.65 178 -2 -0.58 119 -1 -0.44 61 0 0.00 9 1 5.78 8 0 0.00 7 1 7.43 6 0 0.00 5 1 10.40 4 0 0.00 3 1 17.33 2 0 0.00 1 1 52.00
Hmm, I think something is not right here.
I'm going to pass you my indices so you can check them with yours.
1) 1D, 20 cards remaining
2) 2D, 26 cards remainingCode:+----------+----------------------------+-----+-----+---------------------+ | Play | TC | RC | IRC | EV | +----------+--------------+-------------+-----+-----+---------------------+ | Ins | -2.666667 | -2/ 39 | -1 | -4 | 0.01868374178675136 | +----------+--------------+-------------+-----+-----+---------------------+
3) 4D, 52 cards remainingCode:+----------+----------------------------+-----+-----+---------------------+ | Play | TC | RC | IRC | EV | +----------+--------------+-------------+-----+-----+---------------------+ | Ins | -1.625000 | -2/ 64 | -1 | -8 | 0.01675762536870185 | +----------+--------------+-------------+-----+-----+---------------------+
Sincerely,Code:+----------+----------------------------+-----+-----+---------------------+ | Play | TC | RC | IRC | EV | +----------+--------------+-------------+-----+-----+---------------------+ | Ins | -1.118280 | -2/ 93 | -1 | -16 | 0.00704420683992346 | +----------+--------------+-------------+-----+-----+---------------------+
Cac
I am not clear on how you get TC indicies.
Let's go back to HiLo single deck for a moment. The first point where ins EV is positive is 47 cards remaining with a running count of +2 where probability of a ten with one ace specifically removed is .34043. With 48 cards remaining and a running count of +1 insurance is an even bet with probability of a ten = 1/3.
I consider TC as an afterthought. If any strategy can be based on TC that's fine though.
k_c
That's right but you have to find the minimum TC with an advantage (EV > 0). Notice that I'm not considering EV = 0 as is the case with 48 cards remaining and a RC of +1 (+1.083333).I am not clear on how you get TC indicies.
Let's go back to HiLo single deck for a moment. The first point where ins EV is positive is 47 cards remaining with a running count of +2 where probability of a ten with one ace specifically removed is .34043. With 48 cards remaining and a running count of +1 insurance is an even bet with probability of a ten = 1/3.
I consider TC as an afterthought. If any strategy can be based on TC that's fine though.
k_c
So for Hi-Lo the minimum TC with and advantage is +1 with 38 cards remaining (+1.368421).
+2/47 has a positive EV but it is not the minimum (+2.212766). If I choose the latter (+2/47) as an index, I would be missing the +1/38. Choosing the minimum I am not leaving anyone out.
I hope this clarifies a bit the reason for the choice.
Sincerely,
Cac
PS:
Code:Probability ( 1, 48) = 0.10948379351740700 | EV = 0.00000000000000000 | TC = 1.083333 Probability ( 1, 38) = 0.11848084476084628 | EV = 0.00673558467365609 | TC = 1.368421 Probability ( 1, 37) = 0.11724956168846722 | EV = 0.09273854206857113 | TC = 1.405405 Probability ( 1, 36) = 0.11615750533711657 | EV = 0.18414147877827691 | TC = 1.444444 Probability ( 1, 35) = 0.11520502807569055 | EV = 0.27847377068517254 | TC = 1.485714 Probability ( 1, 34) = 0.11439179051559317 | EV = 0.37746685848827344 | TC = 1.529412 Probability ( 1, 33) = 0.11370982242446477 | EV = 0.48216522729260802 | TC = 1.575758 Probability ( 1, 32) = 0.11315687042991192 | EV = 0.59223852463659021 | TC = 1.625000 Probability ( 1, 31) = 0.11273095115303063 | EV = 0.70891051616766010 | TC = 1.677419 Probability ( 1, 30) = 0.11242834403282574 | EV = 0.83303002664525394 | TC = 1.733333 Probability ( 1, 29) = 0.11224827600011453 | EV = 0.96498071843429400 | TC = 1.793103 Probability ( 1, 28) = 0.11219071480790882 | EV = 1.10603795422197226 | TC = 1.857143 Probability ( 1, 27) = 0.11225530170173108 | EV = 1.25729406479844208 | TC = 1.925926 Probability ( 1, 26) = 0.11244394430607697 | EV = 1.41965707340514502 | TC = 2.000000 Probability ( 1, 25) = 0.11275959305850702 | EV = 1.59482790974114774 | TC = 2.080000 Probability ( 1, 24) = 0.11320538702747235 | EV = 1.78444307341845132 | TC = 2.166667 Probability ( 1, 23) = 0.11378686901253821 | EV = 1.99010297258943591 | TC = 2.260870 Probability ( 1, 22) = 0.11451108123654910 | EV = 2.21443414443043274 | TC = 2.363636 Probability ( 1, 21) = 0.11538559148603061 | EV = 2.46009226051608909 | TC = 2.476190 Probability ( 2, 48) = 0.05474189675870350 | EV = 0.00000000000000000 | TC = 2.166667 Probability ( 2, 47) = 0.05474189675870349 | EV = 2.12765957446807707 | TC = 2.212766
Last edited by Cacarulo; 05-24-2022 at 04:33 PM.
OK.
Now the question I have is for HiLo, 8 decks. You previously pointed out that the minimum TC with an advantage there is for 185 cards remaining and RC = +11, TC = 3.09. The first time in an 8 deck shoe that insurance should be taken is with 384 cards remaining and a TC = 3.93. If RC with 384 cards remaining is +28 then TC = 3.79 but insurance is not positive EV. However, if index is 3.09 would that mean insurance should be taken for RC=+28, cards remaining = 384?
k_c
Aha, I see what your point is and I totally understand it. The "minimum TC" should be the starting point since we know that from there on down the EV is always negative. From there on up we could obtain the sum of the product between the EV and the probability and find at what point that average is maximum. I think that would be the best. It is likely that we will find some negative points along the way. Ideally, they should be the least.OK.
Now the question I have is for HiLo, 8 decks. You previously pointed out that the minimum TC with an advantage there is for 185 cards remaining and RC = +11, TC = 3.09. The first time in an 8 deck shoe that insurance should be taken is with 384 cards remaining and a TC = 3.93. If RC with 384 cards remaining is +28 then TC = 3.79 but insurance is not positive EV. However, if index is 3.09 would that mean insurance should be taken for RC=+28, cards remaining = 384?
k_c
Do you agree?
Anyway, I always check the indices by simulation with one, two and even three decimal places. I have also verified the TKO ones.
For SD the result is exact. For 8D it is +3.1. Later I will try the simulation with two decimal places. Stay online.
Sincerely,
Cac
Bookmarks