Hi guys,

I've been working on my push-fold game with Preflop+

I'm doing well with the open-shove ranges. They are not too hard to remember, I found a few mnemonics.
So whenever I make a mistake, it's usually a 0.1BB mistake, sometimes 0.2BB.

It's different with calls.

Sometimes let's say A7o is a fold, but A8o may be a +0.8BB ev call.
Same goes with PPs, sometimes 66 is a fold, and 77 is a really profitable call.
So it's easy to do big mistakes.

Is there a way to memorize the exact ranges ?
Or should I just run thousands of hands in preflop+ untill I get a reliable gut feeling of what is right ?

The easiest memorization strategy is to create a color-coded chart.

Here's an example of a HJ shove 10bb effective, followed by the calling ranges:



Instead of trying to memorize every calling range individually, you can simply combine the positions into one color-coded chart. I prefer google sheets, but any spreadsheet application can work.



The charts make it easy, but you'll want to incorporate training exercises to really lock it into your mind. For example, here's a training drill where we play against an early position shove 10bb deep. Regardless of what software you use, some combination of color-coded charts backed with regular training drills is IMO the fastest way to memorize shortstack play.



Good luck!

I would say the best way to memorize push-ranges is to write down which hands to push with for 10BBs, 9BBs etc. and which hands to call with for 10BBs, 9BBs etc.. As you play more and more hands, you will find that you are checking the ranges rarer and rarer as you are starting to memorize them. Marginal hands (say you have 8,5BBs and it is a push for 8BBs, but not 9BBs) can be checked after the tournament.

Keep in mind that the ranges are only optimal if your opponent is push/folding according to them as well.

I am talking about HU here.

For non-HU it is a matter of analyzing hands you were unsure of after the tournament, and then you will gradually become better and better at push/folding.

As with HU, ranges based on opponents playing perfect are not optimal against players who don´t play perfect.

I wouldn´t worry so much about playing unexploitably. You should be doing what is optimal against the other players, and they are rarely playing perfect, even in push/fold.

If you are going to use a program to improve push/fold play, use one where you can set the ranges you think the other players would play. And if you are playing single table tournaments, or have been at the final table, use a program which lets you set the payout-structure. It is more work, but it will give you the optimal range (if push/fold is optimal in the situation).

Also keep in mind that programs aren´t perfect. They don´t take all factors into account. Some take more factors into account than others, but none are perfect.

Layered charts might be a nice feature to introduce Tom ?

We're talking MTT's, right?

I think Preflop+ doesn't consider Card Removal. Ranges that consider card removal can differ quite a bit in some spots!

Yeah. Calling shoves and calling shoves after I opened I do make the most blunders and there is plenty of room for improvement. This often leaves me rather clueless.
I made some researches, but I didnt find any quality input.

Is there some in depth article online to read? I struggled to find something adequate.
I am thinking of buying "Modern Poker Theory..." by Michael Acevedo. But as far as I have looked through, open shoves are not covered?!

Another way to help with information processing is to not try to memorize the entire chart. Try to memorize the worst hands that go into your range then the better hands automatically get added in. Worst Pair, worst Axs, worst Axo, worst Kxs, worst suited/offsuit broadways, worst suited connector. Now you are memorizing 5-6 hands instead of your entire range.

This is the correct answer.

And you need to actually sit down with a paper and pen and just learn the worst combinations by heart. And repeat it every other day.

It's quite a slow, boring and tedious process. But once they stuck it will save you a lot of mental effort every hand.

I would try to understand rather than memorize. Also, most players are shoving way too tight, so you should call tighter than software recommends.

Memorizing and calling software ranges is a big leak. In general, you should call much tighter. However, it is also totally dependent on reads, particularly the player, but also the situation and how the table is playing. It depend on your stack size and the stacks behind you. It isn't great to call and fold to cold raise or commit your stack and someone behind you wakes up with a big hand. There can also be ICM issues for you. So I don't see the point in memorizing a table. You need a general understanding of what the ranges should be and then adjust that.

I think this is amazingly on point (great post).

Fortunately most people who are solver oriented think that their problems will be solved by AI and poker experts evaluations. As a system designer and programmer (bachelor of science in Computer Science) I would say that AI is only as good as the person who is developing the software and there will always be bugs. How serious they are only comes up over time (like with the self driving cars). The other thing to consider is what it is based on. My guess is it is billions of hands online that went to showdown (so you can know what the loser's had) but it could also be the winner's hands when they don't get to showdown. The problem with this is that it is all generalizations of all types of players. So while the suggested hand ranges for each situation are accurate they may be biased in the wrong direction depending on who you are playing in the moment.

So I don't memorize software ranges. But I appreciate their existence because it helps me understand the ranges of youngish GTO/Solvers at the table...

Hmmm, while I don't necessarily disagree with the general point (that solvers are not the end all be all), I don't think this is how the solvers "solve" the ranges they come up with. It is my understanding that they aren't looking at a massive database of hands, but rather solving the optimal response for each hand under the assumption that the opponent will also make the optimal response. So it's a purely theoretical exercise without the need for a database (other than the rules of the game). Which brings us to the solvers' main deficiency ... our opponents are not playing optimally; and we can improve our returns by deviating (i.e., playing exploitively). The biggest thing playing by the "solver book" gets you, is that your opponents cannot exploit you by plauing "perfectly".

And I should note that it is a good thing our opponents are not playing perfectly, because if they did, then we would have to also play perfect GTO poker ... and finally we could declare with confidence that poker is dead.

I don't think that's particularly accurate or even fair.

The best use for solvers in postflop/flop situations is to understand why it chooses the hands it chooses for certain lines and to think about the concepts behind those choices and how to apply them more generally in your game.

Looking at the ranges is useful, but trying to memorize them is ridiculous and counterproductive.

I find it kind of useless as indicated by the chart above the usual call candidates are 22+ Ax+ and suited broadways, and the marginal hands depends so much on the actual shoving range and the meta game.

The only take away is to aware you have to call tighter in early position because you are not closing the action.

I do recommend studying the charts. It is just trying to memorize them for all the different push sizes and positions of both players is useless. You do want to know how you adjust by stack sizes and positions and what the best hands to call with all.

You should generally call tighter than recommended, because people usually push too tight. With larger stacks many opponents are not push/folding their whole ranges. It is also easier to shove a 20xBB stack over a 10xBB push than to call or raise with a 50xBB stack and other similar stacks behind you.

It is also useless to memorize ICM ranges, GTO, etc. You need to study the software results and understand the concepts and how to apply them.