Splitting 10's

[iCountNTrack]

The BS edge for standing with T,T v. 6 (6-deck, S17) is 70.3%. Since the index of +4 represents the crossover point where splitting first becomes preferable to standing, the advantage can be only slightly higher than that.

Don

Yep, results from CA shown below for TC = 4 and TC = 7 (an approximation since I am only looking at one composition)

[seriousplayer]

Yep, results from CA shown below for TC = 4 and TC = 7 (an approximation since I am only looking at one composition)

Thank You for the answer. The program "Blackjack Composition Dependent Combinatorial Analyzer" has a PE = 100% and Hi-lo only have a PE = 51%. I don't think it would be fair and accurate to actually use this program to generate EV for count systems like Hi-Lo?

[iCountNTrack]

Thank You for the answer. The program "Blackjack Composition Dependent Combinatorial Analyzer" has a PE = 100% and Hi-lo only have a PE = 51%. I don't think it would be fair and accurate to actually use this program to generate EV for count systems like Hi-Lo?

It has nothing to do with PE. I am just looking at one deck composition and computing the expectation value and standard deviation for each playing decision based on the rules and shoe composition using full composition dependent combinatorial analysis. However like I said this is only an approximation since I am looking at one composition only. The correct way to do this is to find all shoe compositions at all penetrations with the TC in question and compute the EV for each playing decision, and finally compute the weighted average using deck composition probability. However this problem is computationally intractable at the moment. One day computers will be fast enough.

[Gramazeka]

ENHC, S 17, DAS, 6 D, Sp 3

10,10 vs 6

TC 4 Stay= 0.7202 ; Split= 0.7104
TC 5 Stay= 0.7247 ; Split= 0.7587
TC 6 Stay= 0.7294 ; Split= 0.8284
TC 7 Stay= 0.7341 ; Split= 0.8990

Perfect index= 4.3 TC

Not infinite deck.

[Three]

The first split may draw heat and second more heat. OTOH, not resplitting tens when you can and should may give you reprieve from the heat. Laughing it off and saying something like not pressing your luck too far, or you took a shot, but you're not crazy, may help you look more like a gambler.

I have used this approach each time. Partly because of the heat issue and partly because I don't want to end up with 4 stiffs hoping for a dealer bust. It worked great at deflecting the heat. All of the sudden I went from possible AP to shot taker. The attitude in the pit totally changed and all day they were talking about the crazy gambler that got his money back by splitting TT only once on a few different rounds. Fortunately I got strong hands catching T or A on all but 1 round on which the dealer was nice enough to bust. Of course there was a Ploppy uproar on the 2 of the 3 times I shared the table with others but everyone won and they were very happy and excited at my good fortune. 3 splits in two different shoes at 2 different tables and it worked out to be a cover play by the way I handled it. You have to love the psychological aspects of the contest between the catchers and the counters in this game.

[Three]

Thank You for the answer. The program "Blackjack Composition Dependent Combinatorial Analyzer" has a PE = 100% and Hi-lo only have a PE = 51%. I don't think it would be fair and accurate to actually use this program to generate EV for count systems like Hi-Lo?

You are miss applying PE to this play. HILO has an average PE of .51% because it is so bad at most and the most frequently used index plays due to the ace being counted as high card and acts as a low card. In this matchup the ace is a lesser high card for playing the hand. The play is a strong play for HILO index. Since the ace is worth about the same as the T in terms of matchup EOR against a 5 and half that against the 6. HILO PE for this play is strong for a level 1 count. This play is as much on dealer bust probability as it is on you getting a good card.