Hi all,

I was playing around with Piosolver when I came across what for me was a stunning revelation. We are defending the BB vs a CO open at 100bb effective. Flop action is a 1/3 pot cbet and a call. On the turn, the board is QcTd9hJh. If CO continues with a pot-sized bet, the solver has BB shoving most Kx with hearts - and CO folds virtually all his K-hi straights with no heart redraw. I was amazed because until now I would have automatically called. Now I see that if the raise size is large enough, the 20% chance that we call and win nothing is far more than winning half the pot minus half the rake (assuming online; I'm pretty sure many casinos push back the stacks without raking if we're chopping).

I wanted to find the "break even" point (how large the pot has to be to make the EV of calling positive) and was hoping someone could confirm or correct my logic here. Assuming 5NL stakes and a 10% cap on the rake and the pot is raked at the very end and that CO always faces a K-hi straight (AK would have reraised preflop and an 8 wouldn't shove), we are looking for an X which is the size of the pot minus the raise (because in the EV calculation I assume we are considering calling to win half this amount). So we want to find the place where the EV of losing the raise amount is equal to the EV of winning half of X, right?

.2(500-x/2) = .8x/2 - .1x/2, or x = 250

Thanks in advance!

Actually, the rake is a flat 50, so now I'm thinking we have:

.2(500-x/2) = .8x/2 - 50, or x = 300

[QUOTE=solarglow;58083151]
AK would have reraised preflop/QUOTE]

Many people, especially at low stakes, do NOT reraise with AK, or don't do it always. So you can figure certain hands are unlikely - but that never means that is impossible. A lot of times a person will shove with the nuts, knowing that someone with the 2nd nuts can't fold. Be careful out there.

I think the solver is just a guide. I wouldn't follow it blindly, especially on later streets, and especially in the micros. I want to extract ideas and concepts, and this is one example. I would have just blindly called in this spot, but now I'm thinking about it.

Just because the solver folds doesn't mean I will in reality. Depends upon the player. Most players in the micros are nitty (either fish or regs). A raise here from them will be rare, but I think at this point it's best to fold a K-hi straight without a heart in CO's position. Now there are some spewy players who I'm never folding to with a K-hi straight, but they are rare.

This spot makes me think of any time there is a raise when there's 3 or more likely 4 to a straight on the board. To call a pot-sized raise you'd have to be good 33% of the time assuming no-chop. It's the chop that for me makes this situation special and why I wanted to do the EV calculation. It's probably best to just fold vs the nitty types because they often have the higher straight or the flush redraw.

well villain has a 20% edge on the turn action. we have a 40% share of the pot. so our expected loss = turn action * .2

our expected win is pot * .4

so if turn action is 50 we lose 10
and if the pot is 25 we win 10



so the turn action (the amount of the turn raise that is as of yet uncalled) has to be twice the pot before we want to fold

Thanks! I like how you express that we have a 40% share of the pot (1/2 of 80% for the chop).

What about rake? Right now I'm playing on GGPoker and they will rake the 200bb pot even on chop, so that's .50 or 10bb coming out of the pot. I think we have to factor that in the decision.

are you using your own defined preflop ranges or something you've confirmed is moderately close to optimal? it's really easy for blockers to have outsized effects on pure strats in solvers (but these wouldn't be achievable by most humans)

I'm using my own ranges which I based on what I thought were decent charts put out by a pro a few years back. The preflop solvers are too expensive, and I play such low stakes that I never thought it was worth it. My plan was to revisit my preflop strategy and try and make it more GTO once I climbed up a bit and the competition got tougher.

The GTO ranges are starting to trickle down as freebies, at least at 100bb stack depth in cash games. PokerCoaching has a good one.

yeah so the thing about solvers is they're very much solving for specific parameters. the ranges used could well be a decent approximation of preflop GTO but once you're three streets of range narrowing down the line, those 0.4 combo differences can end up making outsized impacts on solver outputs - IIRC pio works by iteratively calculating an MES (I'm sure it's more complicated than that but it makes sense).