Three Card Poker Edge is not -3,37%, it is -3.40799%?

This is my first post on this forum, have read here for a long time and think this is the best forum to post to get answers to my question.
I have done some calculations with python on Three Card Poker, to get optimal strategy on “edge sorting”, “hole carding”, “know the first card” etc… for this game.
But before I start that I try to calculate the edge for the game if you play it the way it`s supposed to be played. Because then I can compare if my result is the same as the correct edge (-3.372981%).
First time I run threw my code I got that the edge was -3.372981%, everything seems fine.
But then I noticed I have ranked low straight (A,2,3) same as a high straight (Q,K,A) I change my code
so (A,2,3) is the lowest straight of all. Then run the new code and find that the edge is -3.40799%.
I have like check everything like a million time to see what I am doing wrong. But can´t find it.
So I just wonder that maybe, the true edge for this game is -3.40799%? Maybe someone who has try to calculated out the edge can check their code and see if they done the same mistake as I did first and ranked (A,2,3) same as (Q,K,A)
All that I can find on internet says that the edge is (-3.372981%). And Wizard of Odds himself have write on his homepage that the edge is -3,37%
I use this payable.
straightflush pays 5 to 1
three of kind pays 4 to 1
straight pays 1 to 1

So is it any chance that the right edge for this game actually is -3.40799% and not -3.372981%. Need answer can’t sleep otherwise.

Ps. My English is like a drunk kid is writing I know..

I stand by 3.372981% and I treat A23 as the lowest straight. I'm quite sure Stanley Ko and James Grosjean get the same figure. Otherwise, I would have heard about it by now. I've had that figure on my site for at least a decade now.

Can you share your code?


Not sure how you coded it, but perhaps you changed it where a player's A23 loses against a dealer's AKQ....but you still treat a player's AKQ as a push against dealer's A23? To me, that would make sense why the original % is the same as the "known" %....since a player and dealer are equally likely to get AKQ vs A23 or A23 vs AKQ.....half the time the player wins, half the time the player loses.....or in other words, the player/dealer always push.

Quote: RS

Can you share your code?


Not sure how you coded it, but perhaps you changed it where a player's A23 loses against a dealer's AKQ....but you still treat a player's AKQ as a push against dealer's A23? To me, that would make sense why the original % is the same as the "known" %....since a player and dealer are equally likely to get AKQ vs A23 or A23 vs AKQ.....half the time the player wins, half the time the player loses.....or in other words, the player/dealer always push.

Thanks mate! Cant understand how I have miss it. I have check my code so many time and miss this, Shame on me!. Now it work fine again. So I can finnaly go on with my work.



And sorry Wizard I missbelive you, I should know better.

Quote: Wizard

I stand by 3.372981% and I treat A23 as the lowest straight. I'm quite sure Stanley Ko and James Grosjean get the same figure. Otherwise, I would have heard about it by now. I've had that figure on my site for at least a decade now.

Actually they both contacted me first and I corrected their errors. No thanks necessary.

deleted

Isn't correct basic strategy under regular (no hole card or anything) circumstances just Q64 or better to play?

ZCore13

deleted