Hi all,

I'm diving into the IP strategy in SRP. Specifically I am looking at CO, BTN vs BB.

My initial understanding was that we can range bet small (33%) on all boards that favor our hand range. So most boards other than super wet boards or 3 middling cards.

I am now realizing that is wrong. Looking at betting frequencies on boards with an A we don't bet more than 60% of the time even though this favors our range.



So a couple questions:
What is the theory behind range betting?
What boards can be range bet from the previously described positions

This is a big question, but I can give a couple pointers.

The first thing is to consider the entirety of the range matchup. Having an equity advantage is not the only thing to think about. On a board like K76s, for example, opener will have equity advantage and has combos like AA, KK, AK that bb defender can't have. But oop will still have plenty of 2pair hands, sets, pair + draw hands, and strong draws. This is definitely not an appropriate spot to rangebet.

On a board like K72r, majority of a bb defend range is complete trash. There are fewer 2p combos and basically no strong draws or pair + draw hands. To rangebet, we want the bb to have a bunch of weak hands that we can target with a small bet. When bb has plenty of strong hands, in general rangebetting isn't as attractive of a strategy. When we rangebet, that means we're cbetting a ton of trash too, and cbetting a ton of trash into a range with a bunch of strong hands doesn't really make sense. Generally, this means less connected and rainbow boards will have higher frequency small betting, and more connected and suited boards will be less inclined to bet high frequency.
--

The second thing to consider are the incentives for our betting range, especially our marginal value hands. On a board like K72r, if we have a hand like Q7s, 88, or even 33, when we bet and villain folds a hand like JTo they're folding a pretty substantial equity share. This is a lot of incentive to bet weaker value hands. However, on a board like KQ2r, hands like J6s or 98o or whatever junk have extremely low equity against made hands. If we have QTo and villain folds 87s on KQ2, that's not really very important. So despite a very substantial equity and top-end advantage on 2hi boards, solver will often play a strategy involving big bets and checks -- in fact, on some flops you can even play an overbet or check strategy, because your midstrength hands really don't benefit much from betting small. In these cases, the strategy shifts towards sizing up to push your top-end advantage and betting more polar.

Theoretically speaking, then, when it comes to rangebetting there are a few key things to look for:

-Having an equity advantage throughout the ranges
-Having incentive to bet midstrength hands for value + protection/denial
-BB having a bunch of trash hands such that a small bet can meaningfully force folds and deny equity

Hope this helps!

Thank you!

Would a good default strategy be to take the top 10% of my value hands, 50 combos from the BTN (so sets, 2p, and the top 3 TP hands), the best 10% bluff combos, and the worst 10% of air combos and always bet these (I read that flop value to bluff ratio should be 1:2, not sure if that is good info or not)? Default size to 2/3 but increase up to 125% as the board becomes more dynamic and the nut advantage remains with me. Then the other 70% of my range will either bet small or check depending on the board texture.

I realize this would be exploitable if I never had 2p+ when I check. Not sure if that is an issue at low stakes, but I could increase my betting range to the top 15% but mix in a check with all combos 1/3 of the time.

A strategy like that bot only leaves your checks exploitable, but also leaves your small betting range exploitable. In general, you should probably stick to using one betsize per flop unless you know what you're doing and are playing a mixed strategy (at minimum, ensuring each betsize has some quality bluff combos and strong value hands).

Also, generally speaking your worst trash hands wouldnt really be used as part of a large betsize. Complete trash hands are more appropriate as small bets or checks, in general. I certainly wouldn't advise taking your 10% weakest hands on every board and using them as large size bluffs.

I find it a little confusing that on two-tone flops, there is less range-betting than on rainbow flops when two-tone boards have more draws that you would deny the equity of. Is the problem that when Villain has a lot of decent equity (backdoor-)draws, it's easier for them to construct a sensible defending range with some equity and that with small bets we fail to price those hands out or am I missing the point?

Well done! Yes, on tt boards there is a big block of autodefends between fds and offsuit bdfd hands. The region a small bet puts in a difficult spot is a bit smaller and relatively weaker, and we have less incentive to rangebet against a range that has more equity hands. With that said, there is still plenty of small betting on tt boards in theory, so it's not like this is the only factor to consider, but we do tend to bet less often and for slightly larger sizes on tt boards.

Interesting thread! I'm curious, range betting and betting small are two different things, right? I'm just playing around, but I've noticed the solver often likes range betting (betting with the entire range or almost all of it) out of position in a 3bet pot when the aggressor's range is merged. It also likes betting small in other situations, but not with the entire range.

For example, I've been playing around with a CO vs BB scenario where the flop is QT9r. Equities run close on the flop and both ranges could have the straights. The solver seems to like the 33% cbet here using a small but strong portion of its range, and the BB only check raises 7% or so. I say that because often the OOP player check-raises with a high frequency vs a small cbet on the flop.

I don't mean to change the topic from range betting. Just trying to make sense of this given your great explanation of the difference between a flop like K72r and K76tt.

Generally, you're right that a small betsize doesn't necessarily imply rangebetting and this is a pretty common misconception. This is a bit beyond the scope of the thread, but there are a few important things going on here. On connected boards (both straighty and monotone), one thing is there are some invulnerable nut combos around where a lot of hands are completely dead and have 0 equity. While a lot of bluff combos are low equity, usually we have outs to overpairs, draw completers, backdoors, etc. But on connected boards, we can hit and still be completely dead at a high frequency, esp if we bet and barrel and condense villain's range. This is pretty bad for all of our hands.

Another thing is that a ton of hands in villain's range have very high equity even against our value region -- on a monotone board, for example, any single suit offsuit combo still has very high equity against almost every value hand. Usually, if we have a hand like a set or 2p, we get a lot of money in against hands that are doing extremely poorly against us and have high incentive to build a pot with these hands. But on connected boards, this really isn't the case and a ton of runouts make us regret putting money in even with decent hands otf, plus, sometimes even with value we're already just dead on the flop.

Basically, the presence of some extremely strong made value hands + future runout considerations disincentivizes us from putting a lot of money in with most hands. We have to be careful with the hands we bet, and we don't want to go too large (which just condenses v's continuing range more and more around strong combos). This actually extends to both players - you often see pretty low x/r frequencies and cautious probe strategies after flop checks through as well.

Thanks for this great response. You've explained a lot. I suppose if the turn is a brick on these straighty/monotone boards (like a 2 that doesn't make a 4-to-anything) we can do some overbetting, but the solver uses a strong range containing semi-bluffs that can make a strong hand on the river. Most of the betting though uses 3/4 sizing (for the sizes I used as inputs). Anyway, thanks again.

Now that I know I should range bet on boards like your K72r, I'll also be doing this in position. I'm used to range betting OOP, and that's almost ways in 3bet+ pots. I think I make a lot of mistakes on later streets in this situation. In 3bet pots the SPR is pretty low even after a 1/3 pot bet on the flop, so usually an overbet is actually a shove. A lesser bet can leave us with awkward stack sizes on the river.