I like 40bi rule for nl 6m, 60bi rule for nl hu, 60 bi for plo 6m, and 75 bi for plo hu.

So, if you start with $100 to play nl 6m with, you go $100/50bi= nl$2 stakes. When your bankroll falls below $50, you drop down to nl $1, to have 50 bi again. When your bankroll grows to $250, you move up to nl $5 where you now have 50 bi.

1. I think both are fine choices. There is better information for NLHE but PLO games are softer, so it depends on the person.
2. I am slowly moonwalking my way out of poker and doing random other things that will one day displace poker.

In terms of ratios of (difficulty to learn):(contribution to EV), the best place to start studying are the early streets. So the first thing you do is memorize your preflop ranges at most sequences. Then, you get an idea for your flop sizings/betting volume at most situations you see a flop. Then you develop some heuristics for grouping turn sequences together and getting an idea for a reasonable strat there (although by this point hopefully you know what you and your opponent's ranges are with precision due to doing the preflop and flop work!). Finally, spend time looking at a variety of river situations that come up to build intuitions there, but again, if you know basic GTO and what your range looks like going into the river you can basically just guess and do alright.

I think this question is confusing. My experience playing poker is that most tables have a mixture of stronger and weaker players at the table. Pro poker is then an iterative process where we play a bunch and study a bunch everyday, and try to play hands better vs both player types. It would be a mistake to seek out exclusively top regulars to play; you'll be able to battle them with some recs at the table too in the course of a normal session.

YJ,

Imo, $1/$2 isn't really a livable stake even if the games are soft (though I'm sure there are exceptions, but I'm thinking of the Vegas $1/$2 games I played 10 years ago). I think the move here is to realize you aren't a pro until you get to $2/$5, and even then it's quite marginal, with a semi comfortable living not really being achievable until $5/$10 or $10/$20 level.

That being said, if you're going to spend 60 hours/week on poker it's reasonable to expect you can move up fairly quickly. Live poker isn't my specialty, but I think it's more about finding the softest games in Vegas and reliably playing in them when they're good. That's the best way to outrun the variance. I prefer learning to play online because you get many more hands in vs on average tougher players and get to use all the analytical tools that help make you better.

Sorry I can't give a more supportive answer, but this is my perspective and maybe it will save you pain later.

I don't have a good answer to this, other than that it's certainly possible. My experience was learning to play when quite young and getting totally obsessed, which is the most common way people go pro. I think if you were efficient you could move up on 4 hours/day, with it being like 50/50 play:study. It will be slow though because of the lower volume of hours playing.

I don't think I have much expertise about this. I've seen that this varies a lot based on the person, though I agree becoming isolated is a massive problem for poker players mental health. My experience has been that I like having a home-base and a fairly traditional lifestyle, other than the poker part.

I was a bit of an anomaly in that I was using the best analytical tools available the whole time, but the tools have gotten better. So, as you said, for NLHE/PLO it's Pio/monker, and then for mixed games it's stuff like odds oracle and your knowledge of MoP. When I was taking poker seriously I would study 2+ hours most days and also play a lot, so I kept getting better everyday.

I addressed this above, but the "hunt and peck" strategy of looking at Pio results isn't the best. It's important to know ranges with precision at least until after the flop, and then you can guess on the turn and river.

So, putting that into practice, a good step would be to download all the free pio preflop outputs available and then look at how flops play at every high freq terminal preflop node and make a giant spreadsheet you can quiz yourself with.

This sounds like a good approach. I was thinking some form of looking at aggregated reports, but the spreadsheet + test yourself strategy is a good structure that seems less overwhelming/more effective. Time to get to work!(/stop being lazy :P)

Thanks for the response!

I don't really think of it this way. Pros provide liquidity to games where the edge is sufficiently large. "Edge" can be broken down into a ratio of winrate in bb/100 versus variance in standard dev of bb/100. So, we might say 6m LHE is a low edge game until we consider that the std dev might be like 25BB and the edge .5BB, meaning the player can play relatively higher stakes on a smaller bankroll while still maintaining a solid hourly.

See flagrant violation of caveat #2, please narrow it down :)

Ok, how about just 4,8 and 9.

I know this isn't the healthiest approach, but I mostly substituted a "good routine" for obsession. I was really into poker and wanted to play, study, and think about it all the time, so that's what I did. With a strong intention in place, it's more of a logistical problem in terms of choosing moderately efficient ways to spend the energy you're already going to spend.

I was careful to be practical nearer to the beginning of my pro career because I needed to be sure I could manage to make a comfortable living (for a broke 21 year old, that wasn't much). For me that was when I could beat ~$1/$2 online because I could make say 4-10bb/100 on 500 hands/hour, which was like $50/hour and more than I thought any reasonable person in my circumstances needed at that point. If you're older or starting a family, and especially when considering opportunity cost of not having a real career it's obv a lot different equation, but that's another story.

I've noticed that the best players seem to really enjoy working on poker, so they spend a lot of focused hours without what feels like heavy effort.

I got super lucky in poker and don't have any real regrets.

I mean obv with massive hindsight bias I could think of 1000 things I could have done better but to me that's kind of like going back in time and knowing what the winning lottery ticket number is, and seems like kind of a tilting reflection.

I will rephrase to top 5 poker related media.

1. PioSolver
2. MonkerSolver
3. PT4/HEM
4. Mathematics of Poker
5. Forums/videos/tableninja

I've tried to be really clear that the best players nearly all play a ton and study a ton with solvers, with the very best spending disproportionately large time studying solvers. I find people don't want to study solvers efficiently because it's boring or difficult, and I get that -- but I think it's kind of a gnarly substitution bias for me to just list a few training videos I kind of liked and let people infer they're almost as important as solver work when they're definitely not.

It's also worth noting that PioSolver at least has a free version, and there's a lot of free or cheap software out there that's excellent.

Look how often total air bets compared to global bet freq of 50-60%. In other words, group hands together conceptually. It's probably not very important to bet a certain variety of air 78% versus 40% so long as the range has the relevant properties. Don't believe me though, go test it in Pio!

nice question!

JD, see caveat #2. You are not doing yourself any favors by numbering your first question and then trying to trick me by posting the next 3 as their own posts lol

Good call NH

I've done worse than expected against players like Trueteller, Jungleman, BerriSweet, Doug Polk, Michael Thuritz, Phil Galfond, Samoleus, EM2, David Oppenheim, Fedor, probably a bunch more that aren't springing to mind right away.

Well, it's optimal to check or bet full range on many boards, so yes.

To give a better answer, I do think it usually performs well to leverage the simplicity of a 100% cb strat on a board that might play 90% bet / 10% X in Pio, and make our opponent hit his call/raise/fold frequencies perfectly in order to exploit us.

No idea. I always expect it to because I know there are people making a business out of distributing live assist software in exchange for a share of profits. But for some reason online poker remains beatable, maybe the sites are doing a better job than I give them credit for, or maybe humans are better intentioned on the internet than I give them credit for.

I have been quite lazy about using monker to study multiway postflop play. There are so many different situations that I've mostly confined myself to looking at the basic patterns of betsizing and action frequencies for common configurations of players.

@1- I tend to take the macro perspective. I find there are at least two types of studiers, the "spreadsheeters," and the "gamers". The spreadsheeter tries to approach projects systematically and documents everything they do-- so for example they decide to study 3bp BB vs BU and make a list of 12 textures they feel are representative and look at properties of each node on each texture in a spreadsheet, then study and quiz themselves on that content. A gamer (like me) is more the type to want a really quick feedback loop between content and evaluation to stay engaged; so I might build a sim, look at the outputs, and then immediately try to guess how the range is constructed and see if I'm right.
I think looking frequently at what you call "micro perspective" is mostly a waste of time in both cases, if you get it wrong it's only exploitable for a very small amount and you can verify this.
@2- Yes, particularly on paired boards where there's tons of mixing, or deep 3bp.

Hot take: So I know everyone is aging and getting really into work/life balance and happiness and stuff (I am too). But I think it's smart to disambiguate things that make us happier from things that improve performance. So, imo, my best performance in poker was a semester period in university where I vividly remember my roommate brought a giant (like 10+ pound) flat of rainbow cookies to our dorm room. I was so into poker that I just didn't leave the room the entire semester except to go to the most essential of classes, and just ate nothing but these cookies that were right next to me, and listened to GZA's Liquid Swords on repeat. My life was saturated with poker and I got really good really fast.

Is that sustainable? No, obviously not. But some pretty high level of obsession with breaks for essential human-ing is possible and I did that for many years. During the years of max obsession I improved relatively faster than my opponents.

So, basically, I just think there's a relatively linear relationship between time spent and skill acquired.

I also think people really want to be both great at poker and be happy, balanced, reasonable people who are kind to themselves and others. I'm a big fan of trying to do this, but I think the balance stuff is good for its own sake and actually makes you a bit worse at poker, and that's OK. Poker obviously isn't very important when seen from even a remotely high level.

All that being said, in the very long run, some level of balance is required. For example, I noticed that travelling around the world to play made me unhappy after awhile, so I made a home in one place. Poker is extremely stressful and it's common for players to develop mental illness that makes playing well difficult or impossible. Since in poker, a keen mind is your livelihood, it's a really bad idea to mess yourself up too bad.

See my post above about using study time efficiently. Start with preflop/flop, and group hands together conceptually.

Classified. Though I'm not playing nearly as much as I'd like, because I'm spoiled and don't like playing below $50/$100

No. I think poker has a pretty linear and slow input/output relationship between time/skill.

I just don't know. I've been predicting poker's demise for half a decade now and it seems to keep shambling on.

My heart wants sites to make really big changes that result in a more even playing field for pros and recs. For example, I'd love for the sites to make up a new game every month or couple of months, and force everyone to compete in the same pool. I think this would make the analytical tools of the pros less effective and foster a climate of learning new stuff that would be creative and exciting.

I also understand that this will never happen for a lot of good reasons. I think the status quo will just kinda continue and poker will be a niche thing for many years to come.

The obvious big thing is whether the US, Chinese, and Indian markets will open up and choose to compete on international player pools like Stars/Party. That would be a huge deal but seems unlikely to happen soon.

Sorry I don't, I haven't played those games in many years and have little idea how they play.

Well, mixed games are my favorite. I enjoy playing the big bet mix in vegas or the big mix game in vegas, especially shorthanded or hu.

As for individual games, probably PLO, badugi, and pot limit triple draw?

I think you are making this too complicated. If the site takes 5% of the pot when you see a flop, then you need an edge such that entering the pot is +EV. The necessary edge can come from the quality of your hand's EV at the node, or from mistakes made by your opponent.

My guess is that there are fairly few situations where Fedor thinks playing any two cards is +EV.

k23_BG Are you sure Fedor wasn't saying he would play any two IP vs weaker opponents in tournaments? It's an important difference.

this would be in a tournament where you DO NOT pay rake per hand to enter the pot, like you do in cash games. also there are antes etc in tournaments, which make steals with a wide range more profitable.

I don't think of it this way. Sizing choice on turn usually has more to do with the amount of protection needed in the gamestate. For any betsize b of sufficient size, there's some hand or hand class that nears indifference, so saying "what do we want to make indifferent" feels like circular reasoning to me. You mostly want to work with the EV in each range and the amount of EV generated by protection in each range.

Probably tabasco?

Poker is a terrible profession if you don't love playing/studying poker.

I don't think using multiple preflop sizings contributes almost anything to EV. I think that poker is getting more solved and more of the edge will come from sharp exploits and implementing GTO at a higher level.

But if you play in a pool where everyone studies with ranges for, say, 2.2x opens, there might be some merit to studying and implementing a strategy with 3.5x open, for example, to get people into these diferent nodes they're not used to ? This is what Domnik seems to be doing at MTTs at the moment. I don't see anyone else doing it, but I definitly have a hard time playing vs him because of it.

I think your take is reasonable Felipe. I guess if you're specializing in one form of poker (e.g., nlhe 6m non ante) there might be time to learn the standard stats along with some weirdo strat that's a bit lower EV but confuses your opponents a lot.

It's important to note how heterogeneous winrates are in live games. In a typical live game with stacks in the 200bb region, with say 7 regs of varying quality and one fairly weak recreational, winrates have to sum to 0 with rake. So, suppose the rec is losing 14bb/100 or something, the mean winrate for the regs is 2bb/100, maybe 6bb/100 for the best couple of regs and at say 35 hands/hour that's pretty thin. On the other hand, it's not that uncommon for extremely weak players to play live, and they often lose more like 50-150bb/100 in EV. During those times, winrates for the regulars can be very high, in the 15+ bb/100 range. In the times when the game is great, winrate vs the weaker players rises more or less linearly with stacksize, so you want to be as deep as possible. I think a lot of the EV in live poker comes from reliably getting in these great games. The other (much larger) portion of the time, the regs play in marginal $$/hour games so that if a weaker player wants to join, games are available, and I mean people want to get some hours in anyways.

The other big confounder as you mentioned is that variance over reasonable amounts of chronological time is massive in live poker. It's common for live pros to consider themselves to have gone through the swings and to be seasoned after grinding a game/stake for a couple of years. But a cursory look at a variance simulator will show that their results (good or bad) are mostly attributable to variance. I think reasonable estimates at mean winrate is around 8bb/100 for live (fairly generous) and standard dev skews high because of deep stacks, prevalence of straddling, and over aggro play, call this 120bb/100 for 6m nlhe. We can suppose 35 hands/hour and a pro playing an avg of 40 hrs / week or 35x40x52=~73k hands/year. With those inputs we can see (https://www.primedope.com/poker-variance-calculator/) a standard dev in WR is ~4.5bb/100, or approx half of expected winrate. Variance in reality is much higher because stakes are heterogeneous and so is game quality and standard deviation, and I think it's very difficult to get in so many hours of live play per year. Not to mention all of this stuff compounds, so running great early in one's career produces non linearly better results later in one's career.

I don't know how big of an issue RTA is in the big no limit games on stars. I do know of bot rings on smaller sites that are highly sophisticated and professional, and I also know secondhand that 3rd parties attempt to distribute RTA to regulars and take a share of profits. It's also easily observable that the EV to be had for using RTA is high. So, I think it's prudent to be worried about RTA at all levels in online poker.

The other major issue I see is that Stars ToS (which is probably the strictest anti cheating standard a community can reasonably adopt in online poker) allows fairly elaborate spreadsheets to be shown on screen. So even if my opponents are fully within the rules, I can be reasonably confident that they have massive sets of spreadsheets that give them reliable guidance on how to play perfect preflop, near perfect on the flop, and provide solid guidance for the turn and river. Being oldschool, I often play vanilla poker and just fire up my laptop and play, and I feel this puts me at a large disadvantage. So I often stick to PLO and mixed games, where my bigger knowledge advantage makes me competitive even though I'm not pushing these organizational edges.

Hey Felipe, I'm not an expert in MTT play so I can't really comment helpfully. If I looked at sets of solver outputs I could probably dig around and find the effect, but it's not immediately obvious to me.

This is the BB strat vs HJ cbet @ 40bb, against 1/3 cbet (done at 100% frequency at equilibrium, with just this sizing option) at K74r. You can see K7 and 74 are pure raises, while K4 is almost pure calling.

This is the HJ response to the check-raise. (which was 6.5bb vs 2bb cbet)

Felipe, I don't see anything from a couple minutes of looking at the screenshots that easily explains why EV for XR is higher with T2P and B2P. Sorry I can't be of more help. You could think of it as a linear range of T2P+ to XR, and then B2P-KT mix as polar XR, and then 74 specifically plays pure XR because it unblocks the TP combos for IP which are mixing 3b and needs extra protection.

I don't know what that is sorry

Hi Cory,

So, I imagine you're talking mostly about PLO, either 3bp after cb or SRP after flop XX? I find the PLO 3bp spots the trickiest, there's some boards where mixing 25/60/100 all add a fair bit of EV. I think this is just one of those spots at the more mastery level of a given game where lots of repetition and digging is needed in order to implement the complexity effectively. It's worth noting that the 2+ sizings aren't easy for an opponent either -- they have the respond to both sizings and also to exploit they must figure out which range is unbalanced strong or weak and how to exploit that, which isn't an easy thing in PLO.

For example Berri has many sizings and ranges that were fairly unbalanced, but it was still pretty hard to execute a good counterstrat that beat him for more than the added sizings were contributing to EV. Otb red baron is another good example of someone who uses many sizings which prob aren't perfectly balanced but who is hard to exploit consistently.

I'm not sure. A lot! You can prob figure this out using monker?

Meh, pass. I am a big fan of solvers.

Prob PLO? It's big bet and you have to incorporate the side card information effectively. I find I make tons of mistakes even after many thousands of hours.

Good question, I find many strong players struggle on turns following a depolar flop XR plan. So, it's important to understand where EV comes from a depolar XR strat is EV maximizing-- what Pio is saying is that protection is so important that facing a polarized range on turn in a massive pot with medium strength hands is worth it.

How to make that node work? The biggest mistake that can be made is playing a polar turn bet following the depolar flop XR, and playing a checking range made entirely of weak or medium hands (if SPR is deep), the result will be IP hammering money in against the checking range. So, what you need to do is play some blocks on turns where protection is still important, such that flop depolar value XR can still protection bet for value, and then on more polar turns you need to protect the checking range carefully. Once you do this, you'll accomplish the goal of XR a hand like Q9 and getting it to X down fairly frequently (or for one more bet) against a hand like KK. That's a big victory for OOP as you've protected on flop, gotten called by worse enough on flop, and then managed to not lose a huge amount from a high freq hand class like overpair. Another important piece is making sure you mix in enough hero calls for stacks such that your XR|XC|X range is defended by more than a few slowplays or river improves, there are enough natural bluffs for IP on many board runouts that your EV can get ravaged by frequent bluffing.

Awesome that makes a lot of sense and I have a better understanding of that situation now! Thanks! you are a inspiration to us all and also a inspiration to my nickname xD ^^!! Happy holidays to you!

I'm a big fan of playing fairly solid for a few reasons,

(1) Playing solid scales amazingly. Every moment of studying done to play a better GTO strategy increases your expected WR at higher limits if/when you eventually get there.

(2) In learning to understand how/why solid play works, you'll naturally look at the EVs of different hands in different nodes, and understand how EV is caused. The result is that when villains are making mistakes you'll know mechanically how their EV loss is caused, and what the safest exploits are to get that EV.

(3) Especially in NLHE, there are so many mixed strategies that simply calibrating your mixes one way or the other to subtly exploit an opponent's likely miscalibrations is enough to generate a large winrate. It's also a nicely devious way to generate EV, it's really hard from looking at hand histories for a villain to see how you're exploiting them because you aren't making very many dominated strategies that pop out and allow them to make inferences.

(4) One exception is the river. By the time river play happens there's so much information out there that I think it's leaving too much money on the table not to fold some pure calls and vice versa. If you know exactly how ranges evolve from preflop --> river and understand the strat on a certain runout, it's often easy to see that no matter your opponent's psychology, they have such an imbalanced distribution by the river that big leaks are inevitable. For example, suppose that the flop is Ts9h2s BU vs BB and the action goes X/75/C| Qs X/100/C | and the river comes Js and the action comes X/75. There are so many players who miss the low equity bluffs on previous streets that even if everytime they arrive on the river with a bluff they fire, they still won't have more than a negligible bluff freq. That's a dramatic example, but I often give myself leeway to exploit an opponent by 0-3 or even 4% potshare on the river if I'm fairly confident.

:)

You're probably doing fine. Spend the money on PT4/HEM so you can keep a database of results, and look at pokerdope sims so you can know if your producing a real winrate or if your results are mostly caused by variance. Try to get your hands on some solid preflop outputs from a solver because that's just memorization that contributes massively to winrate. Try to interact on forums and in person so you can talk poker and be put into contact with the right information. I mean, being on run it once is pretty strong though.

It's really never too early to be making spreadsheets about betting volume / sizings to play on each flop, and then making some heuristics for turn/river.

If you're really trying to save money you can just get results from Stars and then make a homebrew spreadsheet I think using google doc? But if you're making >$10/hour then it's prob just worth it to buy some form of tracking software, maybe there are still free ones around? There used to be an open source one.

So I agree with your premises, but I don't think the conclusion (GTO doesn't have relevance for live) follows. I do think it has less of an impact on earn than for online though, by a lot. Also, to move up to high stakes live vs tougher regs, GTO work is near necessary.

So, if your opponents are totally in La La land, I agree you don't want to play like Pio. The magnitude and high sample size of your winrate is sufficient proof that your opponents are very weak and you're massacring them. So, use nodelocking features in Pio to simulate how they're playing, and then see what a maximally exploitative strat looks like. It might seem obvious to jam AJo for 40bb vs a raiser on 35% of hands and 2 callers on 60% of hands, but where is the line drawn? Is it ATo? A9o? A2o? Etc. You can use analytical tools to add precision to your strategy regardless.

When I was playing it was never called this, but I did try and look at the tendencies of aggregated samples of regulars when I was playing lots of midstakes.

I'm having trouble understanding the question. I don't think the number of combos in their range should make it harder to play. Think of PLO as frequencies of 2 and 3 card combinations, and then subdivide those combos into mixes based on the blocker/equity effects of the 3-4th or 4th card.

DN's open pre is quite bad. My call pre is even worse. Postflop is awful by me. I would get it into my head to play way too loose in PLO vs DN for no good reason.

Turn looks like a close call. River IP is vbetting roughly JT9> or thereabouts and so positive blockers for OOP are things that block the highest freq of fullhouses --> Tx, Qx, and then 7x/8x are fine too.

This seems like an odd thing to put time into given that correctly calibrating the mixes on paired boards would come after playing reasonable well PF-->river across every sort of board texture and preflop action, which as far as I know no one in the world is capable of. I think if you're betting approx the right freq and mixing through the various hand classes it's fine.

I haven't looked at a ton of the Pluribus/Liberatus hands, but I think I can answer this fairly confidently anyways.

It's important to understand the problem the AI researchers are working on, and how it differs from what Pio+humans are doing. The AI guys need a piece of software that outputs a non catastrophic result no matter what sequence of preflop+postflop action occurs; regardless of weird betsizes or action sequences or whatever. So they build out an agent that has systems in place to tweak its base strategy to whatever it encounters. Given the prior goal, the agent often has outputs for many betsizes at a given node, and therefore it can play a postflop strategy with a large number of sizings-- particularly on the turn+river where it uses real time solves to precisely solve a 2 street tree with many sizings available.

In contrast, humans look at Pio outputs and tweak/generalize from them with their monkey brains. One feature/bug of monkey brains is that we like to use sizings that we're used to, and so we end up using a relatively small number of the possible sizing choices.

This is all particularly true preflop; the AI might not care whether it opens 2x or 2.6x but humans are hard pressed to memorize the game trees of 6 different preflop sizings.

Maybe Pluribus hands analysis coming in future vids :3? There was plenty cool, fancy and interesting hands from AI

I think the red line is a red herring !

Happy Holidays :)

I think the EVs in these spots are quite close. I don't think there's a hard and fast rule about which MP combos to continue in these spots.

Hey Ben, hope it's not to late to make a follow up question:

It's about my approach to analyzing the solver's output: thinking back to point "1" and "2" that I made, I usually try to figure out concepts and things that the solver acomplishes with his options. Like for example, when, let's say, double barreling a certain "bluff", we look at how villain's folds may increase the equity (or equity realization) of our hand (by "cleaning outs" for example).
I spend a lot of time in my studying looking for things like that, that I can kind of generalize and bring in to my in-game play, using logical thinking, rather than trying to memorize and replicate the overall strat the solver is doing.

The question is: do you think thats a good way to approach it, or is there a big problem with "humanizing" things too much ? (aka finding wrong justifications for why X or Y happens with our previously mentioned monkey brains).

I think that's a good approach

Have you controlled for betsize? If villains are overfolding a given line, the response is to bet that line more frequently. The fold frequencies described are so large that it's likely you would bet near 100% in these lines to max exploit, but perhaps something so drastic is overdoing it.

yes the betsize filter i use is from 55% to 75% turn and river that fold numbers are vs that sizes

Hi, I think I can answer this question for you. It depends exactly what you mean by exploit. The quick answer is no one is exploiting or being exploited in any of these as at least one person is playing GTO (by GTO I am meaning the Nash solution) and most definitions of exploit are one player moving away from GTO to take advantage of another's imperfect strategy. However I think a more useful question is who is losing EV.

In 1) What we can be sure of is that our EV will be at least as good as if he was opening correctly. As the definition of a nash equilibrium is that neither player can unilaterally improve their equity. In reality he will probably do worse than if he opened the GTO range, so we probably gain some EV with no changes to our strat. We have probably not maximised our EV against his strategy. To do that we would exploit him by increasing our calling and 3betting range. (Obviously when you do this you risk being counter exploited).

2) Again no one is exploiting by a normal definition. But you will probably lose some EV wrt to opening GTO, so you could call that exploiting yourself (but I wouldn't). But you certainly will do no better than if you just opened GTO.

3) Again villain does no better against you than if he was 3betting correctly (and probably does worse). You could do better still by exploiting him by widening your opening and reducing your 4betting and 3bet calling ranges.

Some plays lose money at equilibrium while others break even. The 3 you've described lose money at equilibrium because math. In contrast, throwing rock 100% of the time in RPS breaks even vs equilibrium strategy but is exploitable for the full game value.

I discussed this earlier, but cutting out a constant of your calling freq on each street is pretty reasonable, say if GTO is call [100-50] ranked by EV, then call 5% less, or [100-55] ranked in EV.

In PLO you learn how to balance playing a given hand class (e.g., top pair) to different lines by using your sidecards, and it helps know why hands are mixing in NLHE and what properties matter.

I would use Pio preflop or monker solver

I'm not Ben, but what I found useful is try to look for patterns. In your sim even if every hand is mixed there's still pretty clear patterns, for ex. TP's,OP's, 2p+, 2nd p+draw mostly bet (75ish%), complete air mostly check, FD's mostly bet, gutters mostly bet, most of the betting range seems pretty intuitive, just need to keep in mind that we have to check some fraction of those intuitive bets to not get exploited. Great thing is that there's boards who share very similar qualities, so you don't have to memorise patterns for every single board, for ex. on board like 865ddh patterns will remain very similar to your example. Ps. I wouldn't try to implement two different sizings OTF, I don't think it will gain much ev and implementing them close to accurate is almost impossible to human. Rerun sim couple times w 1 size and check witch sizings yields higher ev, or maybe ev's are very similar so check witch one easier to implement, or maybe ag witch one is harder to play against

I'd start by choosing a single sizing in between the 132 and 66, then look at results again.

Generous of you to do this at all, Sauce.

Happy Holidays!

There's a lot of ways to do this and it kinda depends on the spot. A potshare threshold works well in PLO where calling hands' EV is pretty linear and even frequencies. In NLHE you might need to look more precisely at a given spot or series of spots and see what a given adjustment does to your range. A good heuristic is to continue around 1-A with any hand that you think is close-- that'll keep you in the pot sometimes with hands you think are close folds and what they turn over might surprise you.

Since bluffs preflop often have 20-40% equity against calling ranges, you can't use 1-A type math to describe how to solve the tree.

He had a solver and knew how to use it

1. You're going to want to limp a lot because it's effectively an ante game. There isn't much info on this because ante games aren't popular at the moment. Maybe you can run a weird pio sim on it, idk.

2. Fold pre. Bet smaller flop. Rest of hand is fine with some parts of your range. Your line is fine ish though so long as it's balanced.

The biggest thing is how much protection is required on a given board texture. More protection = larger sizing. Determining protection is relatively easy in PLO, all textures without a straight, flush, or board pair are high protection. NLHE it depends a bit more on how preflop ranges interact, but in general, lower cards (such that overcards have 6+ outs) and larger numbers of straight/flushdraws leads to higher protection. The explanation is pretty simple, if more parts of villain's range can improve to beat hero's valuebets, then hero wins EV by folding these hands out. Hero then increases betting volume (and sizing) relative to a static game, and villain counters by continuing more and increasing aggression. Knowing how villain's range looks and how much EV is generated by folding out hands will help you mostly pick sizing/volume choices that closely track game EV.

Ben, shouldn't the need for protection make us want to chose a smaller betsize abstraction so that we can bet wider for value/higher frequency?

Pio preflop and monkersolver are the best choices here

I don't think it matters that much. You should be looking for more +EV hourly regardless of what you're playing, and finding the balance there is part of figuring things out as a professional.

Mixed games are very bad online, mostly because of lack of volume. Live they're pretty good, especially in Vegas, especially during the summer, and especially at the very highest stakes. It seems like NLHE is still the biggest game in town at the moment, with PLO being a reasonable choice as well.