Reverse Counting

Well I all ready know everyone in here will call me stupid, I was experimenting on Wizard of Odds counting practice game which I believe is fairly realistic and here are the results. Game is DD 70% pen. H17, DAS , I was using Hi-Lo and just keeping a running count and here is the strange part betting in a reverse manner, -3 to anything positive 1 unit bet, -4 bet 2 units,-5 =2.5, -6 =3, -7=3.5, -8=4, -9=4.5 and -10 or more =5 units, for what ever strange reason it has produced a 100 unit profit, why is this? You are probably wondering why try this, well my thinking is seems alot of the time in good counts the dealer gets more BJs and 20s when we have alot bet, in a reverse count such as this those 12-16 hands are alot easier to make with more small cards remaining, go ahead call me crazy. It may loose that 100 units and 1000s more I will for sure give it more testing before using it for real.

In a bad count you can still win, it's just less likely. You'll need to supply everyone with # of hands played before they can tell you how likely your results are.

Its not stupid at all to experiment on no-cost practice games.

Your experience is odd, but a 100 unit profit is certainly possible. Remember that the count represents small incremental changes in probability around what is essentially a 50/50 proposition (or 49.4/50.6 or whatever.)

The following factors increase your advantage when the deck is rich in tens:
- more BJs - player wins 3:2 with his BJs and loses 1:1 with dealer BJs
- dealer must always hits stiff hands (12-16) and thus will go bust more often, whereas player sometimes stands on stiff hands (when dealer's up-card is 2-6)
- doubling on 11s and Tens (and 9s) is more profitable when the remaining deck is ten-rich
- player can split pairs and most pair splitting in basic strategy is more profitable when deck is ten-rich. AA and 88 lead the list of pairs for which splitting is more profitable when the deck is ten-rich

As others stated... Basically it's a combination of a small sampling size and variance. Just like how players shouldn't expect to win a ton more in big positive counts, you also shouldn't expect to lose 'a ton more' in big negative counts as well.

Keep doing that for 50k hands, and I'll make a bet with you that you come up negative by that point (and more than likely far before that) =).

Quote: gordonm888

AA and 88 lead the list of pairs for which splitting is more profitable when the deck is ten-rich

While your statement is very true regarding AA, with 88 it's a mixed story. Splitting 8s against a 10 (and to some extent a 9) becomes less valuable as the count goes up (e.g. the odds of you ending up with two 18s against a 20 is higher) and these plays are considered to have "reverse" indices (where the best play is to split BELOW and stand or surrender ABOVE a certain count).

More generally the splits done because they're profitable (Aces, most ones done against weak dealer upcards) generally get better as the count goes up, but the "defensive" splits (ones done because you put yourself in a better starting position by splitting, like 8s against strong upcards) can either get better or worse with the count depending on the specific pair and upcard.

5000 hands +110 units just as an update

7000 hands +115 units strange? BTW at -4 and more I don t double vs 10