Winning %

Hi,
If we playing basic strategy perfectly, ( but no card counting, etc.), on a long run how many hands we gonna win and lose from 100 hand average?
Don't count hands we push.

Thanks

Win about 42 and lose about 49 (see link below)

https://wizardofodds.com/games/blackjack/variance/

The above link should answer your question more thoroughly. Scroll down to "Probability of Net Win" and you should find a table that says "Summarized Net Win in Blackjack" (it is about 1/4 to 1/3 of the way down the page).

Quote: 21

Hi,
If we playing basic strategy perfectly, ( but no card counting, etc.), on a long run how many hands we gonna win and lose from 100 hand average?
Don't count hands we push.

Thanks

You will have to tell us how to count BJ? Is that one win or 1.5 wins?
What about a double down? 2 wins or 2 losses or just one?
What about split hands? Do both count as a single hand? What if you win one and lose one? Double on one and lose, but plain win on second.... etc....

Looking at a simulation I ran many years ago (CSM UK rules) it was Wins=45.182%, Losses=52.449%, Ties=8.809%. This adds up to 106.440% so counted splits as different hands. Factoring to 100% it's 42.45, 49.28 8.27 so roughly what was quoted above.

Quote: charliepatrick

Looking at a simulation I ran many years ago (CSM UK rules) it was Wins=45.182%, Losses=52.449%, Ties=8.809%. This adds up to 106.440% so counted splits as different hands. Factoring to 100% it's 42.45, 49.28 8.27 so roughly what was quoted above.

I agree that your numbers are roughly equivalent, especially considering that a split hand with both a win and a loss might count as a tie.

I am curious. 106.440% implies an increase of 6.44% dues to splits - but that seems a bit high to me, even with resplits. Your thoughts?

A quick look (UK rules, infinite deck) has 53 out of 2197 hands that begin to split. Obviously once started they might continue to re-split. Roughly I make that about 3% extra. I'd have to look at the code, as when doing it I was only interested in the EVs at various counts so tallied WINs etc.

The code also might have counted a double as 2 WINs (or LOSEs), as I would want to remember the player either won, tied or lost two units. I get that 144 out of 2197 hands hard double (ignoring the soft ones, it's getting late here) and that is 6.55%. As 6.55 is very close to 6.44 it's possible I'm counting hands (including splits) and then counting Units being paid out (or lost or tied).

I'm not planning to re-write the sims to count how many hands made a profit (or how much) but I would be interested to see whether I'm counting hands initially dealt (rather than hands eventually played after splitting etc.) as this affects the House Edge calculation.