You should reach sd a lot w AA. I doubt anyone could play this profitable vs AA w a correct strategy from AA.

Stacks always reset to 100 bbs each hand.

Interesting game ! I'd think the AA was a favorite but I'm not sure.

One remark : when AA flops top set and it's the nuts (like in the 97 hand), the optimal play is to go all-in and always win the whole pot.

Disagree. The optimal play is to bet enough that it is incorrect for ATC to draw and hope he does call.

Well in that case you're talking about exploitive play. The GTO play is to shove (if you could have a strategy with EV > money in the pot, your opponent would just counter that by always folding). Also it's not easy to figure out how much of a price is enough to make him calling unprofitable, given the huge value/bluff opportunity he'll have on later streets.

As the hand above shows, the calculations aren't easy to do in-game.

Betting 2x pot makes it more likely your opponent will make a mistake than openshoving. You're making a error assuming optimal play. Sometimes it's hard to figure out what optimal is. Betting 2-3x the pot with top set will never do worse than openshoving, and it has the potential to do better.

this is wat i was thinking, seems bad to bet 3/4 pot wen u flop the nuts on a wet board

Good post, but that part isn't true. We start each hand with a huge equity advantage, and it's not completely clear that Villain can overcome this with betting actions.

Seriously though how hard would it be to deisng a program to find GTO play, I doubt it would be that hard.

would snap accept this

CheckTheRiver, you're right. Obv. I think we do overcome it but you're technically correct. ATC starts each hand at -.7BB or whatever a random hand is against AA with respect to the 2BB in the pot. After that point though every street increases ATC's equity.

Sklanksy - FTOP - if you would make your decision if you knew his hole cards, etc etc.

I'm assuming the bet booked?

Booked. As soon as mers is online, I'll escrow as well.

Booked. Escrow with mers ok with you?

That's all the action I'm taking btw, since we're not going to be able to do anywhere near 1k hands.

The only problem with the escrows is that TNixon and I need to figure out what we're actually doing here. 50 hands took about an hour and a half because we're such noobs and every decision seems so complicated, plus I think maybe we were a little too friendly in chat yesterday I am pretty busy these next couple of days but of course will see this bet through to a reasonable conclusion within a reasonable timeframe.

TNixon, let's figure out a finish line of the bet, then I'll get these escrows done, then we'll play.

skates' last point about what the game is about is accurate. It's also why I'm pretty confident in my current strategy of being so passive - I think AA wins by checking down most pots and forcing ATC to bluff so much that picking off a decent percentage leads to victory. Bloating the pot makes it much easier for ATC to get that -.7 edge. I definitely feel like I had an edge in the first session, and I hope I can goad ATC into some unbalanced bluffs out of frustration from not being able to contest enough pots. I think this game is close in EV between two good players (though still feel like AA has the edge) but it's a very easy game to prop bet your slightly less smart buddy on, let him take the side he feels is obviously better (see this thread), and let him play it poorly.

We've decided on a 300 hand finish line for this bet. I don't think it'll at all settle which side has the edge (just as 300 hands of real cash isn't enough, plus once someone gets a lead the other player will have to play much more reckless and it could easily spiral out of control), but I think it'll be enough to get a gist of it. We're finding the game has a lot to teach about barelling and will definitely give some thoughts afterward on how to apply the game theory here to your standard grind.

I've PMed the two people who have bets on me against TNixon for escrow. Make sure you understand what the new terms are. I'll be busy a lot this week as I head back to school on Friday but we're going to try to get in another 50 hands or so tomorrow night.

Another big thanks to sejje who has selflessly played dealer and helped out fixing what little bugs there have been in TNixon's super cool program. Let me flop more aces, I'll tip better than he will.

I'll tip better than mers if you quit telling him when to fold!

(I soulread him by folding flop on a KK4 (showed K9) and 567 (straight). I probably should have called though, because afterward he said "****, I must not be bluffing enough". I shouldn't have let him figure that out. This next session is going to be crazy).

unfortunately, sejje told me you had air on the three triple barrels I mistakenly called down. I resolve to ignore his randomized input from now on.

He was just trying not to make it super-obvious. :/

Actually it has to be pretty close.

If you look at what happens in a 3-street static (i.e. turn&river don't change anything) NL game where pot size = 2 and remaining stacks = 99, then the claivoyant player's actual share of the pot is ~4.45*(his equity on flop).

(That's ((2x+1)/(x+1))^3, where x solves (2x+1)^3=100, i.e. on each street the clairvoyant player bets x*pot or gives up)

The AA game is certainly different from that, because of the cases where
1. ATC can't realize his equity because AA has nut top set
2. ATC sucks out only later in the hand and has fewer streets to vbet (though I think a lot of the 5+ outers would be used as semi-bluffs earlier)
3.AA sucks out too on turn/river (not often),
but still I think the number doesn't change that much.

(fwiw ATC vs AA = 15% equity, 15*4.45 = 66%)

This should be published

What an amazing execution of what began as a seemingly throwaway OP.

The main protagonists in this thread should consider taking their show on the road (that is, a wider field of view) and maybe this time next year, 'AA vs ATC' will be a hit casino diversion.

Or yeah you guys could make a theory paper out of it.

I *totally* did not catch this the first time around.

LOL

Sejje Thunderdome. I like.

After thinking about it quite a bit (but without actually trying to prove it by doing any math), I agree with you, and that's why I now think it probably was an easy call on the turn.

Definitely going to be writing up a post-mortem when we're done. Hopefully mersenneary will do the same from his point of view. I don't want to talk *too* much about things I've discovered so far, because I have money on the line obviously. $300 to mers, and $100 each to lank4ever and ICMoney.

Where does this come from?