Quote:
Originally Posted by Shipnickle
Good points. Especially on simplifications. Usually people simplify to range bets when the EV loss is minimal to none compared to a strategy with several sizings on a particular flop texture. But this naturally means you also have to know how to defend and barrel properly after the range bet.
Facing a tough pool: Do you think that the flop multiple bet-sizes + check game-tree is easier to play than the simplified game-tree of only one flop action of 1/3 bet-size? Let's say the solver EVs are the same.
I don't think it's a problem to simplify a solver strategy. Actually that's what we do every time we make a sim. No one runs solves with more than 5 bet sizes or whatever. Our main objective is to simplify as much as we can without losing too much EV. If we wanted to play the real solver strategy we would have more than ~3-5 sizings anyway.
I mean yeah there are some boards you can simplify to range bets where the EV loss is minimal, but from what I've seen on HH/other people and some even decent regs' videos/streams, they are range betting 1/3 on too many boards and on a lotta boards where you're not supposed to or it's lower EV than say having a bigger sizing (even if you choose one) or having 2 sizes.
E.G. a lot of regs will just auto 1/3 range any A high, any K high, any Q high, any "dry" board or any low board without taking into consideration positions/ranges/etc. It's really, really easy to exploit this for large amounts of EV vs avg/bad regs, i.e. bluff-raise more of your continuing range rather than flat.
This isn't simplifying a solver strat. It's just listening to random nonsense gto that someone on the internet spouted out and mindlessly using it, or just doing what every other average/bad reg is doing and hopping on the bandwagon.
Facing a tough pool, well I'd imagine overall it's always going to be harder to have multiple flop sizes + checks otf since combinatorically you'll just have a gazillion more game trees you may or may not be comfortable with. Whereas if you just range bet you kind of force more "binary" situations.
Don't get me wrong, I am totally not against simplifying solver strategies. Not at all. I run my initial sims with 3 flop sizings (33,70,120) then look at the output to see which sizings it likes. And try to rationalize why it prefers certain sizings, and almost always there will be a good reason for doing so which makes sense theoretically and exploitatively. Usually pio highly prefers one sizing among the three so i rerun it with only that one flop sizing. There's just no way I'm going to remember 2-4 flop sizings on board textures lol since my memorization skills suck so there's just no way im doing that.
however sometimes I think it makes sense to have 2 sizings, and i like it, and pio likes it lol. e.g. JT4ss having an overbet and 1/3 range otf. I try not to mess with too many 2 sizings otf just because it's a hassle to remember but if I think population will play like garbo vs an overbet sizing otf and our range warrants one, i'll try to incorporate it.
Anyway that was quite a rant by me, that's why i dont like posting online as much anymore tbh because it bugs me to see this type of stuff lol. Only came back for certain pg&c's.