I would prefer to ask this in one of the non-theory forums because I'm interested in what people are actually doing but I think a lot of people will just glibly say they bet whatever size will make them the most money.

I ask because I had previously favored the approach Michael Acevedo suggests in his book, which is to use bet size per spot but to be careful about choosing the size.

But I recently watched a RiO video by Oxaxa(sp?) advocating two different sizes in the IP vs BB SRP that strongly favor IP but are somewhat dynamic. (E.g. AK3r).

So he would have you splitting into something like B33, B90, and checking whereas Acevedo would have you doing B67 and checking.

Probably a very familiar topic to most of you...I'm just wondering which approach is generally favored by serious pros that are not super elite high stakes types? Have people come around to believing it's worth the effort to balance multiple sizes?

Both approaches are fine.
My approach is to mainly use 1 sizing per spot (not on river) and to use 2 sizings on some specific boards like maybe that AKr, some turns where I want to split block and overbet...

I won't split things things like half pot and pot, or whenever the sizings are relatively close. Mainly block and overbet type thing.

There's frequent situations where you block mid or third pair on the turn or river and IP is supposed to raise like 15-20% and fold virtually nothing, when in reality no one is coming close to those frequencies. Blocks in general way overperform in practice vs theory, so not including them is just a flat out mistake. At baseline you should have block + whatever size makes sense for the polar component (something between 62% - 143% usually, although there are situations such as turn probes where you also need a 200% - 300% component). Really the only exception to this is when you're IP on the river, there you can just play only the polar component, but in every other situation you need to have block + polar as a baseline.

BTW balancing two sizes is nowhere near as complex as some would have you believe, it's actually very intuitive once you study some good sims.

headsup midstakes, similar to previous posters sentiment.


33/66/150 flop
turn and river i simplify to two sizings.
intuitively, the the frequency of the smaller sizings is much higher, and its pretty much board/runout dependent.

i think id be kidding myself if i said everything is at perfect equilibrium, but, it's not so complex to be unreasonable. the ob/x simplifies nicely, so does range bet small. the middle sizings are where it gets stupid imo

Interesting... for your three flop sizes do you think you're reasonably balanced? I feel like at that degree of complexity I would just wind up betting exactly according to hand strength.

board texture is driving force, not strength. and it would be complex until you understand why solver does certain plays.

the process for me was, running dif board types by group (straights/flushes/3gap etc) aggregate data frequencies, figure out which bet sizings are preferred. if its heavy mix i simplify to the easiest/lowest ev loss .
if im surprised by why a hand is played a certain way, i try to isolate the variable.
learn how tp/2p/lp /fd etc is played, in general, on dif board types.

in the end u come away with a pretty good understanding of what should be bet how often, for what sizing, and why - on dif board types, in a big picture sense. some assumptions are made, but, there are trends.



you can take it a step further by having aggregate data/ ranges up with RNG to keep frequencies in check. i know higher up players who do this. xtra monitor /game tree files open/rng

if ur trying to make money, and ur balls deep into a session where you still need the game tree and an RNG to not be xploited - u need to find better action, imo.

This is actually what you want to do in practice, because population isn’t raising enough vs the block size. This makes it unnecessary to protect your blockbet range with strong hands, in fact the strong hands themselves will be higher ev as a big bet since they’re not getting raised much when you bet small.

People think that having 2 sizes is only important when you’re playing nosebleeds, when in fact it’s the opposite: 2 sizing strats are an exploitative tool more than anything else. Against tougher opposition they lose their value because good players know they have to raise more vs the small size, in order to prevent you from betting according to the strength of your hand.

Clickr for example is playing against the best players in the world, and only uses 1 size on flop and turn.

Word

Just to clarify the bolded, weren't you talking about three sizes per flop? (Thus texture independent.)

I think I understand the basic reasons why different hands would want to belong in different sizing classes.

on kk2 im only using 33%

on qJ7 fd ill mix/lots of OB

certain boards are complex strats (dynamic/equity converged) , others r heavily favoring one size at high frequency

OVERALL across all boards, i simplify to the sizes mentioned in previous posts. not using all sizes on all boards.

if you give solver multiple options, it doesnt always elect to use them//sometimes its so infrequent or miniscule.

other boards its a blender mix fest, and if you gave it 50 sizings to choose from it would probably do each one 2% of time

Thanks for the clarification. I think I'm going to wind up doing something like what you posted.

nice dawg let us kno how it works , don't throw ur back out poundin tha fish 2 hard