Why and How to Use Multiple Bet Sizings
HTML-код
- Опубликовано: 28 сен 2024
- One of the most controversial (and nerdy) topics in the world of poker strategy is whether you should use more than 1 bet sizing on flops and turns.
In this video we go deep into our philosophy for why we believe using multiple sizings can benefit many players and we offer a simple heuristic for how to implement such a strategy.
Source Vid: • $100K Heads-Up SHOWDOW...
What's up guys
Nice 27% pot size range on As7s3h, monkey. Come 1c/2c we go HU for rolls.
Hey Doug! I’m really enjoying your HU challenge at The Lodge! It’s good for the poker community. I appreciate your contributions to the culture. Keep it up, my dude.
Doug Polk here...
bbrbrbrbrbrbrbrbrbrbrbrbreakkinnng comment
Thanks4 encouraging the content creator2 paywall his content… idiot
The guys at the local casino have been employing a 25 bet sizing strategy for years. They just need to put in 500,000 years more volume to balance it correctly.
Obviously there are diminishing returns when it comes to using multiple sizings and cons to it as well. But it doesn't necessarily mean you only should use 1 size. Also, I don't think anyone at the local casino is really worried about balancing.
Idk dude. Local dudes playing $2/$5 at the gardens are the hardest to beat in town 😂
I would pay a lot of money for a course by you. These videos are always impressively informative.
DONT GIVE HIM ANY IDEAS LET THE CONTENT BE FREEEEEE
He will truly appreciate it if you make your money playing poker and then give him the credit.
GTO Check bro, trust me, start to see some aggregated patterns and you start to replicate them. $50 a month I think for upload of 25+ hands or something like that. I forget.
@@skaterjon89everything this guy say in his video make no sense to me.
I don't know how many times I will comment that I love your videos. Music background, humour, and great analysis as always. Keep it up! Greetings from Las Vegas 🎉
Your videos are truly a gift to the poker community.
The biggest problem is not being faced up when using multiple sizes. Its actually pretty obvious sometimes, at least in midstakes, when players arent bluffing enough when they use the big size on the flop in combination with a small size. Or when they use the small size to probe, they are usually more capped than gto, and its really hard to balance that. Good regs will catch onto this, and WILL exploit you. Especially cause nowadays database analysis is so common. Good regs are buying millions of hands online to analyze tendencies
It's insane that such great content of poker has only 12k plays while videos of "pros” and celebrities playing nonsense hands are viral all across the Internet. Anyways, been a fan for several years and hope you can keep it up. Lots of love from China💗
By far the best poker channel around
Wow that weight loss bet really worked out for Doug, body is looking amazing.
lmao that intro, gold as always
He’s actually the intro goat - I’m just realizing that after the last few videos
glorious end 🤣.. .amazing vid... now i wanna do more complex sims because of you hahaha. keep doing vids please!
I have to disagree with the point at 10:42, it is not just Villains perception of our range that matters. Villains perception only matters when we are trying to estimate Villains ranges. Our actual range matters for balance which takes place across many hands in similar situations.
It does actually matter in the long run even though it might seem like it doesn't because it might seem like Doug won't be in that spot enough times in his life but I think that's a fallacy. Curious on others thoughts though!
Love the videos and analysis!
Very good video!
Where does Doug rank in heads up these days, top 25, top 50?
I don't think he really dedicates himself to studying to be the top currently. But I would guess he could be if he wanted to.
I think there is some confusion here when talking about heroes actual range and his perceived range (perceived by the villain). If hero knows that villain perceives his range differently than it actually is and devises his strategy based on this he is playing exploitative poker. Of course whenever you can you want to exploit...
Also not sure what "you don't know if you are exploitable" should mean. If you restrict Player 1's tree you get less EV, he is being exploited for that extra EV compared to the bigger tree, not more or less. You can also check the EV of actions with specific combos of Player 2 and compare them with a sim where Player 1 has more sizings, the EV you see is the REAL MAX EV you will get playing against a Player who is using such a static strategy (as long as Player 2 stays in the tree as well) (ignoring things like bunching effects which would impact the EV's very slightly which a postflop solver wouldn't usually consider and would need additional assumptions, so the only assumptions we make are that the cards to come and in Players 1 range are uniformly distributed).
Just 2 minor points :) I like the videos in general :)
The only time you don't have to care about villain's perception of your range is if you are playing perfect GTO, which no one is, especially if you have a very simplified game tree. The EV indicated by a solver is not the real max EV precisely because your opponent is not constrained to the same limited game tree (or any game tree calculable by current solvers). The solver is also clairvoyant to each player's strategy, but you are not. The example I gave with Doug's hand versus Hank is illustrative. If you are using a limited strategy, but your opponent is not aware of your limited strategy, his reaction to your actions in theory will diverge from your sim, which in turn will impact your strategies because you aren't aware of your opponent's range or strategy. This effect is compounded for each subsequent node as the ranges become more and more diverged based on each player's perception of the other's range which will be flawed. That being said, if you are able to play a simplified game tree very well, yes, that strategy will most likely be effective but there is no guarantee of that or of unexploitability because all of the metrics you are seeing are based on the toy game where both players have clairvoyance.
@@FindingEquilibrium Hey man, "The EV indicated by a solver is not the real max EV..." you know hero's EV in solver is actually their min EV (even in this toy game) since it's always against the counter strategy, right?
Solver uses CFRM to optimize minimax (en.wikipedia.org/wiki/Minimax). min (max -EV) is minimiziing your maximum loss..which is not maximizing your EV.
The solver is finding the strategy with the highest lower bound. It's not finding the maximum EV strategy.
Your audience is going to get very confused by this.
@@ChowChow414 The "minimax theorem" is a very confusing name. From the hero's perspective, the solver is looking for the "maximin value" - it *maximizes* the *minimum* possible EV across all of the villain's counter-strategies - i.e. it chooses the strategy whose strongest counter-strategy provides maximum EV for you (the game is zero-sum therefore that minimum EV from your POV will also be the villain's best reply). Hero's maximinEV in this situation might be positive or negative depending on the original ranges - I think taking the additive inverse of that number (the player's loss) and trying to explain the algorithm around it only makes it more confusing. It's as you said in your 3rd paragraph - we're looking for the *highest* lower bound - i.e. maximum possible EV assuming villain replies perfectly.
I think what FE meant is that this number is not guaranteed to be game-theoretically correct when we restrict villain's replies when looking for minEV (again, from hero's perspective) values in different counter-strategies. We might (in theory) overlook an out-of-tree reply to our supposed maximin strategy which would in turn make a different maximin strategy optimal (and higher EV; or lower "loss", if you insist)
@@Cowtymsmiesznego Yes, so the solver is solving the minimax objective function, there is no way around this. It can feel confusing all it wants, but that's on us to explain what is happening accurately.
The only time we are maximizing our EV with the solution is when villian is countering us..that's a single, unlikely case and not the rule, so talking about it like it's the rule is misleading.
The EV in the solver is a lower bound. Maximizing a lower bound does not imply maximization generally. Saying otherwise is implying playing GTO is maximizing your EV...it's not.
You can simplify to a certain uncommon sizing that the solver will say is 0.01% lower EV, but it will print against humans because it's so counter intuitive and they will make more mistakes...despite still being pseudo "GTO".
This is where maximizing a lower bound being analogus to maximizing breaks down. One will make more money.
Does my point make sense or am I just being pedantic?
There is more to EV than just choosing the model parameters..it's the realistic reaction as well. I get his point, but I feel he missed a big part of the decisions with simplifying.
Also idk a single high stakes reg that only practices defending vs their own simplifications. Using the multisize solutions is good for this. You still get the lower bound guarantees, but if you want to hard counter them, then you need to study THEIR simplification.
But you can just study your simplifications for offence, and the multisite for defence...you don't need to choose.
I know he is selling a product..but..idk man.
@@ChowChow414 Nah, you have a point. I just wasn't a fan of how you explained it by considering the player's "loss" - it lowkey suggests that the best you can do with GTO is break even, which is simply not true.
But otherwise it's absolutely true that the only scenario where the "NE" solution is guaranteed to (and indeed usually the only scenario where it does) maximize EV is when villain plays the best counter. However, the NE solution is also the only one to do no worse against any other counter-strategy, which is a very powerful feature. When I argue in favor of GTO I like to compare it to chess - where playing the minimax NE approximations is much more widely approved, and it's much easier to see why these moves are considered "best". But then again, if you can succesfully limit the villain's reply to some specific strict subsets of available options, there will likely be non-equilibrium strategies that guarantee higher EV. And in the extreme, if you could see your opponent's cards, you obviously wouldn't want to play "GTO" like an idiot.
TLDR: I agree with everything you're saying, I just didn't like thinking of the model in terms of the hero's "loss" - it gets confusing coz GTO will, in specific scenarios, guarantee you positive EV - e.g. when you're on the button 3-handed.
Thank you
That intro was so funny
The tattoo…👏👏👏👏
amazing video
perfect to fall asleep to
If a player cannot make adjustments on the fly from simply focusing on opponents actions from hand to hand, how are you pro card player? This type of reliance on solvers at best is a massive weakness and at worst has some nefarious undertones to say the least.
Idk why but the use to Heuristic by all GTO solver gurus, drives me nuts. There are splenty other synonyms, english is the most plentiful language in this area. Yet i hear Heuristic used at nauseum.
This whole video can be broken down to what i have been saying about this entire solver generation, For most the players playing ,especially live. You cannot solve for stupid. Listening to Doug level himself into bad play after bad play, then reinforce it once away from the action is hilarious. His live cash game results speak for themselves. He is a solver slave that seems to lack the ability to naturally find pattern recognition.
Haha, sooooo guud ^^
Phil Ivey, arguably a top 5 player of all time, has never read/studied a poker book in his life. Phil Helmuth, the greatest tournament player of all time, doesnt study this crap at all. So if this kind of material is "optimal" for your poker game what are these guys and many other greats doing? Cause they aint following GTO. GTO stuff is fun to look at but it wont make you a great poker player.
You realize that using the example of two of the best to ever play actually helps make his point right?
If you’re at the uber elite level of exploitative player like either Phil, then you can change your strategy damn near perfectly on the fly from hand to hand.
For the rest of the world, balance and study will be the optimal path.
You don’t use exceptions to the rule to show the rule isn’t the rule.
@@Solving_Live_Poker Dont get me wrong, GTO and math are part of the game, but math nerds like to think thats all there is to the game when there is a lot more to it. Study GTO all day and take that to your local 2-5 game and see how you do. You will get destroyed. Poker is a multi faceted game with many variables, so studying GTO all day, thinking its going to make you a great player is just not quite accurate.
@@DonTrump-sv1siif you memorize gto and don't adjust, sure, you'll do poorly at low stakes. But if you understand why gto takes certain lines, and can adjust your game to the suboptimal play of your opponents, you will absolutely crush your opponents.
Of course you do, and the more exceptions the weaker the rule.
@@DonTrump-sv1sitake a purely guess-work exploitative approach to your local 2/5 game with a 100 hand analysis on all players and choose the wrong side of exploitation and see how you go.
"Last hand he bet big with AA on a AQTsss board therefore when he bets small he must only have bluffs"
"This guy is old, he never bluffs, I will fold my middle set."
Damn I caught a stray
I think that means you “made it.” Be flattered, Mike!
Paraphrasing; "A prophet is without honor in his own land". FE is simply superb in his melding of range analyses, exploitative exploration(s) (note the alliterative geniusness here.. :) ), situational and elemental breakdowns of the essentially unique nature of every discrete hand, and his ability to spot everything from our own confirmation biases to a Rashomon-like perspective of each of our psychological desire(s) to be "correct" ex-post-facto, supporting our need for affirmation. What makes this so hilarious (and, it must be noted, Doug, who is brilliant in his own willingness to observe and display his own biases, yet he posts this, and other hands on Twitter, shows Doug's deep desire to actually learn, rather than seek a form for self-congratulations.
All the best to you, FE, and there's an absurd hilarity in watching your cogent analyses and the attendant viewership numbers, s, conjunct with someone like Matt Berkey, who'd likely break my arms, if they were to be interposed between he and his spine, were I to inadvertently disrupt his practice of patting his own back.
Oh well. Again, FE, multiple kudos to you (this from a guy that's played poker since 1960. Yes, 1960).
Dougs going to be 😂😂😂😂😂 when he sees this
What happened to this guy? He hasn't made a video in almost a year
I'm not sure adding more bet sizes makes your strategy more GTO. I say this because if the start a sim with say 4 sizes per a street the solver may like them all at first. As you solve closer to 0dev it will bet some sizes more and more and others less and less and may eventually say I don't actually like this size at all
Yeah, there will be many situations where you offer the solver various sizes but it will only choose one. What this video refers to is deliberately removing sizings that the solver actually uses.
Jajajaja sos un genio
I'm the best player ever. lol. This is the video I've been waiting for. It explains how closely the actions of life (your life) and poker (your actions) operate. In that, you can not only create your own reality, but you can bend and shape reality in a way that suits your ideals, character, personality and specific play style. With various bet sizing, you can create, as well as manipulate how villains respond to your unique actions. There is so much more to discuss about such a topic. We could have debate upon debate. But for now, Thank you for all the wisdom bestowed upon us all. I will definitely watch this a few times to in order to pluck out more gems. Cheers man.
Can't express how much knowledge and hilarity I get from watching your videos. Thank you for the great work!
I have a question at 22:45, why T♣️A♥️ and T♣️A♠️ become the only two combos Solver makes full pot bet while other AT combos remain small block bet with single size flop/turn. I don't see the blocker or unblocker here. Also, Solver change its choices for betting combos and sizes with multi size flop/turn. Can't really understand the reason behind. And furthermore, do you think Doug really know and remember T♣️A♥️ is the combo he should bet heavily, or did he just bet with all his AT combos?
EQ, I intuitively knew all the things you say about GTO from the first time I heard of it, but I could barely articulate 10% of what I knew. To hear you fluently and lucidly articulate my instincts is truly a wonder to behold.
You're the only person I've heard who truly understands GTO on a meta level. No-one else even comes close to your wisdom and clarity. The only other person who I'm impressed by is MJ Gonzalez, but maybe he is more strategy focused. Uri Pelig is pretty good too, but not on your level.
Either way, massive respect to you. You are in a league of your own.
Yours MMA sherdog and doug polks videos are sick thanks for improving my game big respect 🙌
Good music choices and poker = lovely
Once again - excellent video
Love u ❤
Great video as always!
Your lack of call to action for GTOx is killing your EV.
What is call to action? Marketing thing? Yes, very lacking in that area...
1:00 Blur more or not at all.
Followed by Martin Kabhrel, Prometeus Poker...
lol it's a joke, Doug is not sending texts about GG software