Did this video change your approach to poker tournaments? Let us know down below! 👉Check out the article for more details about this experiment: blog.gtowizard.com/when-does-icm-become-significant-in-mtts
Hi, I feel like this video drastically misrepresents the impact of ICM. I greatly support both you guys and HRC, but the parameters used (average stack 7BB, push/fold only) are drastically different from real conditions, and while you state these conditions in your video, the remainder of the video seems to present the results without disclaimer and towards the end summarise that ICM should be taken account for at 50% or 37.5% of the field, based on the results of this experiment. I understand that the computational complexity of swimming these strategies in more complex conditions (or actual tournament conditions) is unfeasible without potentially pairing up with google, for example. However, this sim is so far different from tournament conditions that I believe its results can not be relevant. I am neither arguing that ICM should or should not be considered at these field thresholds, rather than your experiment is a poor way of determining that threshold, and this video does not communicate well enough that the real implications of ICM could be completely different to the recommendations in the video.
We appreciate your feedback. Good science always starts with abstracted models. This isn't a well-studied area of poker science, and as you know, data analysis is always a matter of interpretation. However, we believe this is very strong evidence that switching to ICM sooner rather than later is a stronger strategy. At 20:02 we demonstrated how the GTO strategy changes as the tournament progresses. These sims account for deeper spots, postflop play and more complex strategies. ICM strategies differ drastically from cEV even halfway through an event. The summary at the end of the video does state that the "sweet spot" might be quite different when accounting for more complex models.
@@GTOWizard honestly a much stronge argument could be made with just a normal solver (although the solve might need to much calculation? Maybe solve at a smaller accuracy would be enough for the claim). Lock the cEV solution in a multiple ICM sims (at different stages) and let the solver calculate the range $EV of the cEV solution. the farther it differs from the GTO ICM solutions $EV the worse it is. If the Range $EV differs drastically at 50% field left and not so much before, it would be very strong albeit hard to interpret empiric evidence for your claim. I fully agree with @Dylan that your conclusion was drawn too hastly from the experiments. The experiments do indeed show the importance but the exact thresholds can't be easily extracted. But as you said, there was more evidence in how differently the solver plays, but the video definitely suggests different logical links
@@qwertz12345654321 The principal goal of these experiments was to establish whether or not ICM adjusted ranges are actually profitable in the early/middle stages of large field tournaments, compared to using chipEV ranges. Some were arguing that ICM is inherently flawed for large fields and that the resulting strategies might be too risk averse and perform worse than regular chipEV strats. We could not find any evidence that ICM adjusted strats would perform worse than chipEV at any point of the tournament. The experiment you proposed could be used to identify when exactly the strategies start to differ, but not which of the two alternatives is actually more profitable.
I think this is something we all kind of knew, but still punted off with chip-ev strategy regardless. Very interesting and awesome to have it in solid data that preserving tournament life EARLY pays off aswell!
I simply cannot find the right words to express my gratitude or this lesson! Many THX to the GTO Wizard team for the content! This is worth a billion times more than any other paid ICM course on the market! 🙏❤❤❤❤❤❤
Glad someone finally did this. Great work. Ever since i read Dara Okearney's book ive been thinking that implementing ICM would make sense from the get go. Yes, the deviations are less extreme early on, but if you save a chip here and save a chip there throughout the entire tournament, that's gonna add up in the long run.
The second half of your comment makes absolutely no sense. "saving a chip here and save a chip there" is literally the opposite of what ICM adjustments do
The primary issue is that GTO is always incorrect in the direction of being tight. This is true by a small amount, to be sure. The main reason being that it doesn't consider the future advantage of having a big stack, or the fact that your chips won't have the same value if you keep making defensive -CEV, +$EV decisions. Furthermore, even early in tournaments, a 53-47 flip is already good enough to be very +EV, a fact which has been corroborated with simulations. So yes, folding a 51-49 will gain some marginal value early in the tournament. The problem is this is too marginal to identify in practice and over-adapting to ICM and playing too passive is a far more common error than flaming out, so this is mostly scared nits justifying their emotional preference for playing tight.
Explains why I kept having mid-tournament leaks. I would crush it early, build up a massive stack and then bleed it off during the 50%+ remaining portion and have to limp into the money. Ironically, me playing short stacks really well was because I would tighten up into something similar to an ICM model. So now I just need to do that mid-game even when I have 30+bb instead of
HOLY SHIT! I've been playing poker and i just came to this conclusion today when i finished 19th out of a 300 person field! this makes so much sense. now i know im not crazy!
Just wanted to comment. I was risking too much later in mtts and always felt like I was getting knocked out close to bubble. I started approaching icm like you said at 50% and have noticed a huge difference in cashes and higher placements. Glad I found this channel
i'm curious to see if altering your strategy based on stack size rather than players remaining makes an impact. is icm better when you're deep stacked while chip ev is better short stacked or vice versa? and does it make sense to alter your play based on stack size? i have my suspicions, but it'd be cool to see it actually modeled.
amazing! Thanks for this video. Deeply inside I was considering the ICM factor much sooner then bubble starts but all my friends have been telling me I was wrong...
I wonder if the tournament being push / fold is impacting the cEV strategy more than the ICM one hence the difference in result? While I can't offer a way to actually model it, I would postulate that when bets are no all -in or fold, the cEV, while busting more, will have a heavier distribution of top results that will reverse at least some of the losses it shows in your graphs.
This video is awesome, thanks!! Some questions I had (just jotting thoughts, not expecting an answer lol): we went over icm from 25% down to final table, but what happens in between? how much icm is there right after the bubble bursts? is it more or less than at 50% of players left? at what point after the bubble bursts does icm become significant again, and is it the same kind of icm as pre-bubble (does it just work liberally)? Fill in the blank: in a e.g 1000-max tourney with 15% paid, there is the same amount of icm at _% left (post bubble) as 25% left. (For a given field size) How does large field satellite strategy change with a small, average, bigstack with 2x paid left? 1.5x? 1.2x? stone bubble?
Generally, ICM pressure decreases a bit after the bubble bursts, then rises again as you approach the final table. Although it never decreases down to chipEV levels. We've recently added ICM solutions for Bubble spots, as well as a tool to display the bubble factors/risk premiums. Now you can explore these questions yourself!
It would be really interesting to see how the players do when they are unaware of the opponents strategy or if they just did not adjust at all. This would probably be more in line with actual results and I’d imagine mr ChipEV would do much better
Interesting video! It would ve cool to know more about the new, nonrecursive, ICM algorithm. Also it would be nice to see the experiment run with ICM vs ICM +FGS to see how FGS changes things, maybe in the context of a short handed SNG to save calculation costs.
Very interesting! Great content! Would love to see how this changes postflop play from the simulations on gtowiz today which is Chipev! :) You are good at explaining!
Worth noting that you use symmetrical models to compare chip ev which are going to be much tighter than the CEV chart. So anyone on this video you don't have to defend 33% bb vs bu. There are models where we call 20% and models where we call 80% there are so many variables to consider
Thank you for a great content! It would be interesting to see how those "good strategies" (ICM 50%+) perform without those "bad strategies" (chipEv heavy) in play. I wonder if acumulating chips early uncontested (without spazzy chipEv strategies) would have merit and could outperform pure ICM strategy.
I wonder how much longer tournaments would last if everyone was playing the tighter, more defensive, and slower accumulation of chips strategy that ICM suggests? And how that will affect future tournament structures? Are we going to see tournament directors going to shorter levels and higher blind jumps to account for the fact they are taking longer, which would reduce the skill level?
@@2CardConfidence there is less ICM but I think if you ran these exact sims with bounties involved we would probably see the same results just to a lesser extent
ICM is definitely best strategy whenever you are short stacked (or stacks in general are shorts). You need to push only when it's reasonable likely you won't bust out of the tournament. Still, I would not bury chipEV just yet, at the 10 bbs chipEV, while still losing, seemed to fare a bit better. But in reality most of the tournaments start of with way more chips, something like 100-200 bbs generally, so early stages might differ drastically from here. Nevertheless it seems like at least in the midstages one might really need to consider heavily leaning into ICM and trying to get into the money first while avoiding risks whenever feasible.
Is there an opportunity to test the results of a model that switches multiple times from chip ev to icm? For example we take 1000 participants tourney, 200 in the money. It plays chip ev until there is 500-600 people left, than switches to icm until you are in the money. As soon as you took the money, you go back to chip ev and as there are around 60-70 people left - back to icm. I guess it will help to reduce the amount of money lost before bubble and maximize the the amount of top 3 places
I hope I am not too late to comment on this. I have a big issue with this video: I play pokertsars in Europe and the payout structure is completely different from the one tested. Specifically the 1st place gets +/-15% and the 2nd +/-8%. This changes for sure a lot on the calculations
6:18 shows a chart for how often each strategy achieves each placement. cEV plays risky and busts more, but tends to get 1st/2nd more. ICM tends to make the money more often but doesn't place 1st as often. Top-heavy payout structures feature less ICM pressure and play closer to chip EV. Flatter payout structures have more ICM pressure and higher ICM pressure.
@@GTOWizard ok, I got that. My question was not clear, sorry: how much is the model reactive to the structure? Does a 15%/8% structure completely change the balance?
Fantastic. Eye-opening! I wonder, do the "play tighter later" and "blockers are better than drawing hands" ideas hold up in fixed-limit tournaments? Not that there are many such things, but I play HORSE tourneys once in a while, and those are always fixed-limit. Thanks!
I'm confused about the half-stack and full stack experiment. 2 questions: a) What does half-stack exactly mean here? Every player left in the field has 1/2 starting stack? Whats the purpose of using a half stack? b) ICM predicting placement accurately, What does this exactly mean? Shouldn't we be losing money when we are playing chip ev? Im just not understanding what are we trying to do here. Can someone explain this experiment?
The purpose of this experiment is to see how well ICM predicts placement probabilities when players are playing chip EV (unaware of ICM). Details of the experiment are outlined in the article, under experiment #5 blog.gtowizard.com/when-does-icm-become-significant-in-mtts/#example5 a) The reason we use two player groups, with half and full stacks, is because if everyone had the same starting stack then everyone would have equal tournament equity and be equally likely to place in each position. So we split it into two player groups, one with half-stack and one with full-stacks, with the big blind set dynamically to 1/7 of the average stack. b) ICM is a formula that predicts how often each stack will place 1st, 2nd, 3rd and so on. We wanted to see if it accurately predicted placements for each player group, despite the fact that no player in the tournament was aware of ICM. To answer you question about losing money - cEV loses money when other players are playing a better (ICM) strategy. In this case, EVERYONE is playing cEV, so no one has the edge.
So your experiment is based on starting with cEV always except if you play 100% ICM. Have you repeated the experiment with starting with ICM and then switching to cEV? It seems based off your data that playing pure ICM until final 3 or 4, then switching to cEV is the best strategy. Thoughts?
You say in 7:30 the probability of placing first is higher for someone who plays cev than for someone who plays icm. But are you taking into account that when you are chipleader in a high icm situation the one Who plays icm wins many more chips, since he can apply a lot more pressure?
Great point. There are ICM scenarios where the chip leader can open wider and play more aggressively than chipEV. These are accounted for in our simulations; every player knows and adapts to every other player's strategy. But ICM plays more conservatively on average. ICM predicts the probability of placing 1st is just your chip portion of the tournament. So a chip EV strategy, which maximizes chips at all costs, will place 1st more often - although the cost is reducing ROI%.
Nice experiments and very well explained. But let's go further : let's say the 95%Cev profil know the high ICM profiles strategies but they can't adjust to it. In your simulations, all profiles were adjusting to each others. But in reality it is impossible to make. What ppl use to do then, is to play their own strategy until they know how the should deviate VS this or this opponent. 95%Cev profil will have a better gain in reality than in your simulations as they assume other profiles will not adjust, or at least not enough, to their strategies. So, how do we simulate that ?
This is actually a critical flaw. There is no equilibrium when you combine GTO and ICM. This is a methodological issue that DTO/GTO Wizard and everyone else will skim over, because their financial interest is to sell the product of an actionable solution, so they don't want to provide evidence which indicates that there is no such solution. The allure of GTO is that it is 100% safe. You can't lose if you execute it well. Once you combine ICM in, this safety net vanishes, because my ignorance or apathy towards ICM can lose you money if you strictly apply GTO Wizard outputs (and this will happen a lot in real life).
GTO is not 100% safe in multiway spots either, but the flaw is more significant in ICM spots. The only situation where GTO is guaranteed not to lose is HU poker.
When is GTO Wizard going to get ICM postflop solutions? I'd love drills that extend beyond preflop. And also, since you've just demonstrated it is important the entire tournament, let's get preflop sims for the early stages of tournaments as well. And maybe for different common payout structures. And how about a tool that, without having to run 20 similar sims (say BTN vs BB chip ev, early stage icm, middle stage icm, late stage icm, final table icm, and all the different stack depths), favorite them, and then scroll through each one to analyze and compare, instead of all that, some kind of tool or visualization that shows how those ranges shift in each situation, which would allow us to develop some kind of heuristic or intuition on how much to tighten up in the same spot as the tournament progresses, or as your bubble factor increases.
We just launched an update that includes new preflop ICM sims for earlier tournament stages! blog.gtowizard.com/pko-solutions-and-new-dashboard/ We've hired several experts to create content about how ICM impacts postflop strategies! We plan to launch postflop ICM solutions in the first half of 2023, hopefully around March or April.
As a cash game player, this is completely confusing for me. I was always under the impression that you can't be a nit in MTTs cause the blinds and antes will eat you alive. I was also under the impression that as we approach the bubble if we have a stack we should be a "bubble bully". So basically...play tighter in the middle - late stages and much tighter on the final table?
As this video shows, playing a traditional MTT style of going for the win does indeed net you that 1st place finish more frequently. If you are chasing bracelets and don't care about your actual ROI (basically if you are DNegs or Hellmuth, where you make a lot of money from being a "famous poker player" via sponsorships or invitations to cushy private cash games), then keep on playing for the win. But if you actually want to earn the most money, you need to take ICM seriously and just deal with the fact winning the whole thing is less important than having a positive long run ROI. Would you rather be rich or famous? I'll take money over fame all day, though there obviously are advantages to fame, where you could chalk up your occasional wins and negative ROI as marketing expense.
Nit does not equal ICM. ICM is in theory trying to gain the most with the chips you have with the least possible risk. Nit is just avoiding risks as much as possible and missing most of the profitable spots that even one with ICM strategy will play.
I feel like that the ICM model is better at predicting the placement probabilities when everyone is playing chipEV than the probabilities when everyone is playing ICM solutions. Have you also ran the sims where everyone is playing ICM ranges?
Here's a question i have: GTO is unexploitable, but is ICM unexploitable? Could a savvy exploitative clairvoyant player who knows their opponent is playing ICM go looser and more aggressive than ICM (or ChipEV perhaps) to exploit the tighter calling range of the ICM player? How would these sims work out if you allowed postflop play (the sims here were push-fold) and threw in a player who was playing ICM against chipEV players, but had a looser exploitative then give up strategy against the ICM players? Let's say it's a spin and go against Miss ICM and Mr. GTO Chip EV. If Joe Opposite played ICM vs Mr. GTO ChipEV and exploitative vs Miss ICM, how would he fare? Presumably vs Mr GTO ChipEV, he's gonna save chips (or gain payout money) by using the tighter ICM model, but against Miss ICM, he could gain chips by getting more bluffs through and overfolding when Miss ICM occasionally plays back. The sim GTO Wizard did does say that every player knows every other player's strategy, but it was Push Fold. In a real game, there are many spots where Mr. Opposite could just overfold when Miss ICM plays back, a classic exploit against someone who is too tight. And ICM is too tight in terms of ChipEV (but not actual payout dollars), hence the ChipEV players scoring more wins but also more losses. If you mitigated the ChipEV risk by overfolding when an ICM player plays back, you could potentially overperform the toy game models GTO Wizard did in this video.
It's the same thing. The point of GTO or ICM ranges is that they are unexploitable. If you open BU and BB plays a perfect strategy if you open wider you make your hands that were +EV in a GTO scenario -EV and just make BB's range more +EV even if he does not adjust.
All players in this simulation are unexploitable - they just value different things. ChipEV strategies try to maximize chips. ICM strategies try to maximize dollars. Those two goals don't always align. All players in these simulations play perfectly to achieve one of those two goals. We suspect more complex strategies that allow postflop play may demonstrate that ICM becomes significant even earlier than shown in our push/fold models. But this is unexplored territory!
Question. If you where to play the WSOP Main Event only one single time in your life. Would you got for a chip ev strategy or icm strategy? I understand that icm strategy is more EV.
Congratulations, this is a truly far-reaching work ! GTO Wizard is about to become the new gold standard for modern poker study (: - I guess we would expect the same kind of behavior in even smaller fields (like 50 players for example) ? - And what about Sit&Gos ??! In a 9max Sit&Go, chipsEV should not be used at all if we refer to the "three times the number of ITM players" rule...! Or do you think that chipsEV could still be used for the first half-of a Sit&Go ?? Happy grinding to all wizards 🧙♂
SnGs are very high ICM right at the start and because of that you play very thight. You insta profit if anyone bust without risking a chip. At the bubble you can play in some spots simluar to a Saty bubble.
Thank you so much! We tested fields ranging from 90 players to 1000 players with 15% field paid, and saw the same type of behavior in all cases. I think it's safe to assume that it would scale for a 50 player MTT as well. ICM pressure is much higher at the start of an SnG compared to MTTs. For a 9max SnG, we suggest using ICM immediately.
Q: Do these sims account for ICM postflop? If not, would the mid stacks more sensitive to risk premia play tighter, shifting more jams/raises to passive continues or folds? How would this affect the other corresponding ranges?
In this experiment, the tournament simulations were based on push/fold ranges, so no postflop play. Our ICM MTT solutions however, account for postflop play. Check out this article to learn more about how ICM impacts postflop strategy! blog.gtowizard.com/how-icm-impacts-postflop-strategy/
Should we take into account field size when the LR isn't closed yet? Lots of online tourneys will be at a specific stage much longer because of players rebuying/buying in late, we then are at that specific stage longer and bubble will come later than if LR was closed
We tested various field sizes in this video, from 90 players to 1000 players. Late reg has the effect of artificially increasing field size (since there are effectively more entries)
Very good question. You don't know how many are in the field OR the prize pool yet. I suppose you could estimate both of those things to get a rough idea, and also hands ICM says you can play are unlikely to be too different in those early stages. Different from chipEV yes, but not all that different in terms of early stage ICM strategy from tournament to tournament, even if in one you have 300 people who all registered at the very start and in another you have 500 people, a third of which registered late.
This really goes against what most oldschool players and coaches have been saying for years. Really shows how much the understanding improved since the TAG days
You have to ask why the most successful players in MTTs tend to be hyper-aggressive ones in the face of this sort of simulation. There are systemic flaws in the underlying logic of ICM which only the top echelon of pros are aware of. Specifically, the equilibrium or solutions presented in GTO Wizard or DTO or similar products are not accurate or stable in the way CEV simulations are. If you want to follow ICM, but I ignore it, this can cost you a lot of money. This effect is immediately visible in the thought experiment where you shove 25 BB into me from SB with 80% of your hands because I am supposed to call only QQ+ etc because of pay-jumps. If I call TT here, I lose a little equity and you lose an enormous amount. This is just an example to prove that the concept itself is not 100% correct. At a minimum, the 20bb solution here will have to be different versus a person who cares about ICM versus one who doesn't and it's quite complicated when you consider that in real life it's not binary and you won't know your opponents level of awareness/care about ICM. In other words, the idea that you can apply this solution and you will be absolutely fine is false, as the solution assumes opponents care about ICM as much as the player does, but it can be more or less. One more point is that the nature of ICM is a systemic flaw in MTT poker design which incentivizes extremely tight play which isn't very interesting to watch, and sites are already migrating towards PKOs where the exact opposite logic applies.
@@strategygames1026 It's easy to ignore the hyper aggressive tournament players who bust out early. The hyper aggressive tournament crusher is a combination of exploitive play and survivorship bias.
You can "Star" solutions in the solution selector, filter by starred solutions, then press up and down to switch between them. For this video however, I simply opened many GTOW tabs and used the ctrl+tab hotkey in chrome to cycle through the solutions.
So if the buyin is 1000 bucks, you are making 30 bucks as the ICM and -130 as the cev strategy. The first one will play what - 10 hours on average and the second one 2 hours? If you value your hour at let's say 30 bucks / hour, the first one is losing 270 bucks on average. The ICM guy is losing 190 bucks on average. Yeah, I think it's pretty clear which strategy is better...
I agree icm starts to matter when 40% of the field is left more or less. But dont you think it's much more significant when you have more starting stacks? It makes sense folding more if you have around 4 or 5 starting stacks, because that stack may be enough to cash the tournament. However if you only have 1 or 2 starting stacks you shouldnt be tighter because of icm since that stack isnt worth preserving? Isnt there a relation between chips and icm?
If you watched the whole video, you would notice that this question has been answered. Short answer: yes it matters. For long answer: watch the whole video
@@mariodiaz3976 The lower your stack depth is the more ICM matters to you. Every chip is extremely precious when you have few of them and you really need to pick your spots. You can afford to wait much longer than people generally think, do not go shoving / calling with 15 bbs just because you are short. The bigger your stack is, the more freedom you have, even according to ICM. If you have big stack in final table, ICM even expands your range of play since others should be more careful!
It's screwed online, but live is alive and well. And honestly, most players dont have the time or brainpower to absorb an entire ChipEV strategy, let alone adding a whole new constantly shifting strategy like ICM. Like, the ranges change based on how many are in the field and also based on both absolute stack sizes and how big your stack is relative to the rest of the table. How many ranges can you memorize? And that's just preflop. GTOWizard doesn't even have a single postflop ICM spot yet, though i'm pretty sure HRC can calculate those spots.
There will be a shift away from NLH, which has already been deeply analyzed & studied, to mixed games, which can be constantly evolved. Poker fundamentals and adaptability will outperform chart memorization.
I'm an on/off mtt player for nearly 15 years and this is how I always found success. It goes hand in hand with the old timey mtt pros preached about increasing aggression when approaching bubble to win more tournaments at the cost of variance.
Does latereg have ANYTHING to do with this? I seem to unconciously switch to "icm" when latereg ends. Keeping a healthy stack and taking less marginal sports because i feel like Im working towards the bubble since that point.
Late reg has the effect of artificially increasing field size since there are more entries. The impact of ICM on these spots depends at what point latereg ends.
It does make a huge differnt it is push fold u can have a get chip stargety without shoving right before the money with this sim all the icm aren’t going to go all in at all near the bubble on the sim so of course the chip stargety is going bust before the money
It depends on the stack and field size. We've recently added ICM sims for earlier MTT stages, as well as a features that displays the bubble factors. blog.gtowizard.com/pko-solutions-and-new-dashboard/
All premium members have access to our weekly coaching! You can also view previous coaching seminars on our website via the coaching page. As of today, we have 85 hours of coaching content!
Can you add in ICM to start switching to cEV at various breakover moments? I’m interest in ICM until the bubble and getting more risky once bubble bursts.
Did this video change your approach to poker tournaments? Let us know down below!
👉Check out the article for more details about this experiment: blog.gtowizard.com/when-does-icm-become-significant-in-mtts
Hi,
I feel like this video drastically misrepresents the impact of ICM. I greatly support both you guys and HRC, but the parameters used (average stack 7BB, push/fold only) are drastically different from real conditions, and while you state these conditions in your video, the remainder of the video seems to present the results without disclaimer and towards the end summarise that ICM should be taken account for at 50% or 37.5% of the field, based on the results of this experiment.
I understand that the computational complexity of swimming these strategies in more complex conditions (or actual tournament conditions) is unfeasible without potentially pairing up with google, for example. However, this sim is so far different from tournament conditions that I believe its results can not be relevant.
I am neither arguing that ICM should or should not be considered at these field thresholds, rather than your experiment is a poor way of determining that threshold, and this video does not communicate well enough that the real implications of ICM could be completely different to the recommendations in the video.
We appreciate your feedback. Good science always starts with abstracted models. This isn't a well-studied area of poker science, and as you know, data analysis is always a matter of interpretation.
However, we believe this is very strong evidence that switching to ICM sooner rather than later is a stronger strategy.
At 20:02 we demonstrated how the GTO strategy changes as the tournament progresses. These sims account for deeper spots, postflop play and more complex strategies. ICM strategies differ drastically from cEV even halfway through an event.
The summary at the end of the video does state that the "sweet spot" might be quite different when accounting for more complex models.
@@GTOWizard honestly a much stronge argument could be made with just a normal solver (although the solve might need to much calculation? Maybe solve at a smaller accuracy would be enough for the claim). Lock the cEV solution in a multiple ICM sims (at different stages) and let the solver calculate the range $EV of the cEV solution. the farther it differs from the GTO ICM solutions $EV the worse it is.
If the Range $EV differs drastically at 50% field left and not so much before, it would be very strong albeit hard to interpret empiric evidence for your claim.
I fully agree with @Dylan that your conclusion was drawn too hastly from the experiments. The experiments do indeed show the importance but the exact thresholds can't be easily extracted. But as you said, there was more evidence in how differently the solver plays, but the video definitely suggests different logical links
@@qwertz12345654321 The principal goal of these experiments was to establish whether or not ICM adjusted ranges are actually profitable in the early/middle stages of large field tournaments, compared to using chipEV ranges. Some were arguing that ICM is inherently flawed for large fields and that the resulting strategies might be too risk averse and perform worse than regular chipEV strats. We could not find any evidence that ICM adjusted strats would perform worse than chipEV at any point of the tournament.
The experiment you proposed could be used to identify when exactly the strategies start to differ, but not which of the two alternatives is actually more profitable.
this channel is pure gold
It s just people ruining the game for their interests ;)
if you lose you can learn, to win you need to teach, this channel isn`t afraid to put it out there, great content.
I think this is something we all kind of knew, but still punted off with chip-ev strategy regardless. Very interesting and awesome to have it in solid data that preserving tournament life EARLY pays off aswell!
I simply cannot find the right words to express my gratitude or this lesson! Many THX to the GTO Wizard team for the content! This is worth a billion times more than any other paid ICM course on the market! 🙏❤❤❤❤❤❤
the quality of your videos is so darn high, its honestly inspiring to study better and search more intensivly for new insights. big thanks man
Glad someone finally did this. Great work. Ever since i read Dara Okearney's book ive been thinking that implementing ICM would make sense from the get go. Yes, the deviations are less extreme early on, but if you save a chip here and save a chip there throughout the entire tournament, that's gonna add up in the long run.
OMC's had this figured out decades ago...
The second half of your comment makes absolutely no sense. "saving a chip here and save a chip there" is literally the opposite of what ICM adjustments do
The primary issue is that GTO is always incorrect in the direction of being tight. This is true by a small amount, to be sure. The main reason being that it doesn't consider the future advantage of having a big stack, or the fact that your chips won't have the same value if you keep making defensive -CEV, +$EV decisions.
Furthermore, even early in tournaments, a 53-47 flip is already good enough to be very +EV, a fact which has been corroborated with simulations. So yes, folding a 51-49 will gain some marginal value early in the tournament. The problem is this is too marginal to identify in practice and over-adapting to ICM and playing too passive is a far more common error than flaming out, so this is mostly scared nits justifying their emotional preference for playing tight.
Explains why I kept having mid-tournament leaks. I would crush it early, build up a massive stack and then bleed it off during the 50%+ remaining portion and have to limp into the money.
Ironically, me playing short stacks really well was because I would tighten up into something similar to an ICM model. So now I just need to do that mid-game even when I have 30+bb instead of
I wish I could crush it early. I only seem to get slightly above average if I early reg.
HOLY SHIT! I've been playing poker and i just came to this conclusion today when i finished 19th out of a 300 person field! this makes so much sense. now i know im not crazy!
Just wanted to comment. I was risking too much later in mtts and always felt like I was getting knocked out close to bubble. I started approaching icm like you said at 50% and have noticed a huge difference in cashes and higher placements. Glad I found this channel
How is this free content…
i'm curious to see if altering your strategy based on stack size rather than players remaining makes an impact. is icm better when you're deep stacked while chip ev is better short stacked or vice versa? and does it make sense to alter your play based on stack size? i have my suspicions, but it'd be cool to see it actually modeled.
BEST VIDEO of 2023
muchas gracias
im not sure if this makes me want to play more tournaments, or less, because I don't want to fold that much lol
amazing! Thanks for this video. Deeply inside I was considering the ICM factor much sooner then bubble starts but all my friends have been telling me I was wrong...
yOu hAvE tO pLaY fOr tHe wIn
I wonder if the tournament being push / fold is impacting the cEV strategy more than the ICM one hence the difference in result? While I can't offer a way to actually model it, I would postulate that when bets are no all -in or fold, the cEV, while busting more, will have a heavier distribution of top results that will reverse at least some of the losses it shows in your graphs.
That sounds logical. Someone has to win those flips several times in a row.
This video is awesome, thanks!!
Some questions I had (just jotting thoughts, not expecting an answer lol): we went over icm from 25% down to final table, but what happens in between? how much icm is there right after the bubble bursts? is it more or less than at 50% of players left? at what point after the bubble bursts does icm become significant again, and is it the same kind of icm as pre-bubble (does it just work liberally)? Fill in the blank: in a e.g 1000-max tourney with 15% paid, there is the same amount of icm at _% left (post bubble) as 25% left.
(For a given field size) How does large field satellite strategy change with a small, average, bigstack with 2x paid left? 1.5x? 1.2x? stone bubble?
Generally, ICM pressure decreases a bit after the bubble bursts, then rises again as you approach the final table. Although it never decreases down to chipEV levels.
We've recently added ICM solutions for Bubble spots, as well as a tool to display the bubble factors/risk premiums. Now you can explore these questions yourself!
It would be really interesting to see how the players do when they are unaware of the opponents strategy or if they just did not adjust at all. This would probably be more in line with actual results and I’d imagine mr ChipEV would do much better
That's an interesting idea. Idk if chipEV would perform better or worse in those conditions.
Interesting video! It would ve cool to know more about the new, nonrecursive, ICM algorithm. Also it would be nice to see the experiment run with ICM vs ICM +FGS to see how FGS changes things, maybe in the context of a short handed SNG to save calculation costs.
Nice explanation! What I don't understand is why not play 100% ICM from the get go, since that showed the best results?
Very interesting! Great content! Would love to see how this changes postflop play from the simulations on gtowiz today which is Chipev! :) You are good at explaining!
Worth noting that you use symmetrical models to compare chip ev which are going to be much tighter than the CEV chart. So anyone on this video you don't have to defend 33% bb vs bu. There are models where we call 20% and models where we call 80% there are so many variables to consider
Thank you for a great content! It would be interesting to see how those "good strategies" (ICM 50%+) perform without those "bad strategies" (chipEv heavy) in play. I wonder if acumulating chips early uncontested (without spazzy chipEv strategies) would have merit and could outperform pure ICM strategy.
Interesting idea
I wonder how much longer tournaments would last if everyone was playing the tighter, more defensive, and slower accumulation of chips strategy that ICM suggests? And how that will affect future tournament structures? Are we going to see tournament directors going to shorter levels and higher blind jumps to account for the fact they are taking longer, which would reduce the skill level?
@@lemmyorleans bounty tournaments already counter that as they incentivize you to play looser
@@2CardConfidence there is less ICM but I think if you ran these exact sims with bounties involved we would probably see the same results just to a lesser extent
I guess I will start to pay a bit more attention to pay jumps and average stack after reaching the bubble than I used to 🙂
ICM is definitely best strategy whenever you are short stacked (or stacks in general are shorts). You need to push only when it's reasonable likely you won't bust out of the tournament. Still, I would not bury chipEV just yet, at the 10 bbs chipEV, while still losing, seemed to fare a bit better. But in reality most of the tournaments start of with way more chips, something like 100-200 bbs generally, so early stages might differ drastically from here. Nevertheless it seems like at least in the midstages one might really need to consider heavily leaning into ICM and trying to get into the money first while avoiding risks whenever feasible.
21:09 I love how K6s makes a very small showing as you approach the final table.
Tombos, you are the man
Fantastic work as always
Is there an opportunity to test the results of a model that switches multiple times from chip ev to icm?
For example we take 1000 participants tourney, 200 in the money. It plays chip ev until there is 500-600 people left, than switches to icm until you are in the money. As soon as you took the money, you go back to chip ev and as there are around 60-70 people left - back to icm. I guess it will help to reduce the amount of money lost before bubble and maximize the the amount of top 3 places
I hope I am not too late to comment on this. I have a big issue with this video: I play pokertsars in Europe and the payout structure is completely different from the one tested. Specifically the 1st place gets +/-15% and the 2nd +/-8%. This changes for sure a lot on the calculations
6:18 shows a chart for how often each strategy achieves each placement. cEV plays risky and busts more, but tends to get 1st/2nd more. ICM tends to make the money more often but doesn't place 1st as often.
Top-heavy payout structures feature less ICM pressure and play closer to chip EV. Flatter payout structures have more ICM pressure and higher ICM pressure.
@@GTOWizard ok, I got that. My question was not clear, sorry: how much is the model reactive to the structure? Does a 15%/8% structure completely change the balance?
Insane amount of value!
Fantastic. Eye-opening! I wonder, do the "play tighter later" and "blockers are better than drawing hands" ideas hold up in fixed-limit tournaments? Not that there are many such things, but I play HORSE tourneys once in a while, and those are always fixed-limit. Thanks!
I'm confused about the half-stack and full stack experiment. 2 questions:
a) What does half-stack exactly mean here? Every player left in the field has 1/2 starting stack? Whats the purpose of using a half stack?
b) ICM predicting placement accurately, What does this exactly mean? Shouldn't we be losing money when we are playing chip ev?
Im just not understanding what are we trying to do here. Can someone explain this experiment?
The purpose of this experiment is to see how well ICM predicts placement probabilities when players are playing chip EV (unaware of ICM). Details of the experiment are outlined in the article, under experiment #5 blog.gtowizard.com/when-does-icm-become-significant-in-mtts/#example5
a) The reason we use two player groups, with half and full stacks, is because if everyone had the same starting stack then everyone would have equal tournament equity and be equally likely to place in each position. So we split it into two player groups, one with half-stack and one with full-stacks, with the big blind set dynamically to 1/7 of the average stack.
b) ICM is a formula that predicts how often each stack will place 1st, 2nd, 3rd and so on. We wanted to see if it accurately predicted placements for each player group, despite the fact that no player in the tournament was aware of ICM. To answer you question about losing money - cEV loses money when other players are playing a better (ICM) strategy. In this case, EVERYONE is playing cEV, so no one has the edge.
I'd love to see the data for cEV Pure + smaller bet sizing. Could be a simpler strategy for cash game players who dabble in tournaments occasionally.
So your experiment is based on starting with cEV always except if you play 100% ICM.
Have you repeated the experiment with starting with ICM and then switching to cEV?
It seems based off your data that playing pure ICM until final 3 or 4, then switching to cEV is the best strategy. Thoughts?
Thanks Again Tombos!
You say in 7:30 the probability of placing first is higher for someone who plays cev than for someone who plays icm. But are you taking into account that when you are chipleader in a high icm situation the one Who plays icm wins many more chips, since he can apply a lot more pressure?
Great point. There are ICM scenarios where the chip leader can open wider and play more aggressively than chipEV. These are accounted for in our simulations; every player knows and adapts to every other player's strategy. But ICM plays more conservatively on average.
ICM predicts the probability of placing 1st is just your chip portion of the tournament. So a chip EV strategy, which maximizes chips at all costs, will place 1st more often - although the cost is reducing ROI%.
Useful content again! Thanks!
Nice experiments and very well explained.
But let's go further : let's say the 95%Cev profil know the high ICM profiles strategies but they can't adjust to it. In your simulations, all profiles were adjusting to each others. But in reality it is impossible to make. What ppl use to do then, is to play their own strategy until they know how the should deviate VS this or this opponent. 95%Cev profil will have a better gain in reality than in your simulations as they assume other profiles will not adjust, or at least not enough, to their strategies. So, how do we simulate that ?
This is actually a critical flaw. There is no equilibrium when you combine GTO and ICM. This is a methodological issue that DTO/GTO Wizard and everyone else will skim over, because their financial interest is to sell the product of an actionable solution, so they don't want to provide evidence which indicates that there is no such solution.
The allure of GTO is that it is 100% safe. You can't lose if you execute it well. Once you combine ICM in, this safety net vanishes, because my ignorance or apathy towards ICM can lose you money if you strictly apply GTO Wizard outputs (and this will happen a lot in real life).
GTO is not 100% safe in multiway spots either, but the flaw is more significant in ICM spots. The only situation where GTO is guaranteed not to lose is HU poker.
Wtg Helmuth Malcher, at least there is one Hellmuth that understands poker!
When is GTO Wizard going to get ICM postflop solutions? I'd love drills that extend beyond preflop. And also, since you've just demonstrated it is important the entire tournament, let's get preflop sims for the early stages of tournaments as well. And maybe for different common payout structures. And how about a tool that, without having to run 20 similar sims (say BTN vs BB chip ev, early stage icm, middle stage icm, late stage icm, final table icm, and all the different stack depths), favorite them, and then scroll through each one to analyze and compare, instead of all that, some kind of tool or visualization that shows how those ranges shift in each situation, which would allow us to develop some kind of heuristic or intuition on how much to tighten up in the same spot as the tournament progresses, or as your bubble factor increases.
We just launched an update that includes new preflop ICM sims for earlier tournament stages! blog.gtowizard.com/pko-solutions-and-new-dashboard/
We've hired several experts to create content about how ICM impacts postflop strategies! We plan to launch postflop ICM solutions in the first half of 2023, hopefully around March or April.
@@GTOWizard Now it's the end of May already...
As a cash game player, this is completely confusing for me. I was always under the impression that you can't be a nit in MTTs cause the blinds and antes will eat you alive. I was also under the impression that as we approach the bubble if we have a stack we should be a "bubble bully". So basically...play tighter in the middle - late stages and much tighter on the final table?
As this video shows, playing a traditional MTT style of going for the win does indeed net you that 1st place finish more frequently. If you are chasing bracelets and don't care about your actual ROI (basically if you are DNegs or Hellmuth, where you make a lot of money from being a "famous poker player" via sponsorships or invitations to cushy private cash games), then keep on playing for the win. But if you actually want to earn the most money, you need to take ICM seriously and just deal with the fact winning the whole thing is less important than having a positive long run ROI. Would you rather be rich or famous? I'll take money over fame all day, though there obviously are advantages to fame, where you could chalk up your occasional wins and negative ROI as marketing expense.
We recommend watching this video to see how ICM impacts your strategy.
ruclips.net/video/xNqyMY1EWK8/видео.html
Nit does not equal ICM. ICM is in theory trying to gain the most with the chips you have with the least possible risk. Nit is just avoiding risks as much as possible and missing most of the profitable spots that even one with ICM strategy will play.
I feel like that the ICM model is better at predicting the placement probabilities when everyone is playing chipEV than the probabilities when everyone is playing ICM solutions. Have you also ran the sims where everyone is playing ICM ranges?
Or the deviation when only like 30% of the players play 80% perfect ranges?
Here's a question i have: GTO is unexploitable, but is ICM unexploitable? Could a savvy exploitative clairvoyant player who knows their opponent is playing ICM go looser and more aggressive than ICM (or ChipEV perhaps) to exploit the tighter calling range of the ICM player? How would these sims work out if you allowed postflop play (the sims here were push-fold) and threw in a player who was playing ICM against chipEV players, but had a looser exploitative then give up strategy against the ICM players? Let's say it's a spin and go against Miss ICM and Mr. GTO Chip EV. If Joe Opposite played ICM vs Mr. GTO ChipEV and exploitative vs Miss ICM, how would he fare? Presumably vs Mr GTO ChipEV, he's gonna save chips (or gain payout money) by using the tighter ICM model, but against Miss ICM, he could gain chips by getting more bluffs through and overfolding when Miss ICM occasionally plays back. The sim GTO Wizard did does say that every player knows every other player's strategy, but it was Push Fold. In a real game, there are many spots where Mr. Opposite could just overfold when Miss ICM plays back, a classic exploit against someone who is too tight. And ICM is too tight in terms of ChipEV (but not actual payout dollars), hence the ChipEV players scoring more wins but also more losses. If you mitigated the ChipEV risk by overfolding when an ICM player plays back, you could potentially overperform the toy game models GTO Wizard did in this video.
It's the same thing. The point of GTO or ICM ranges is that they are unexploitable. If you open BU and BB plays a perfect strategy if you open wider you make your hands that were +EV in a GTO scenario -EV and just make BB's range more +EV even if he does not adjust.
All players in this simulation are unexploitable - they just value different things. ChipEV strategies try to maximize chips. ICM strategies try to maximize dollars. Those two goals don't always align. All players in these simulations play perfectly to achieve one of those two goals.
We suspect more complex strategies that allow postflop play may demonstrate that ICM becomes significant even earlier than shown in our push/fold models. But this is unexplored territory!
how does this apply to very small MTT's (2 tables) or is there too much noise in the data?
I don't even play poker anymore but I 💕 poker maffs
Question. If you where to play the WSOP Main Event only one single time in your life. Would you got for a chip ev strategy or icm strategy? I understand that icm strategy is more EV.
Congratulations, this is a truly far-reaching work ! GTO Wizard is about to become the new gold standard for modern poker study (:
- I guess we would expect the same kind of behavior in even smaller fields (like 50 players for example) ?
- And what about Sit&Gos ??! In a 9max Sit&Go, chipsEV should not be used at all if we refer to the "three times the number of ITM players" rule...! Or do you think that chipsEV could still be used for the first half-of a Sit&Go ??
Happy grinding to all wizards 🧙♂
SnGs are very high ICM right at the start and because of that you play very thight. You insta profit if anyone bust without risking a chip. At the bubble you can play in some spots simluar to a Saty bubble.
Thank you so much!
We tested fields ranging from 90 players to 1000 players with 15% field paid, and saw the same type of behavior in all cases. I think it's safe to assume that it would scale for a 50 player MTT as well.
ICM pressure is much higher at the start of an SnG compared to MTTs. For a 9max SnG, we suggest using ICM immediately.
@@GTOWizard Thanks 🙏
Q: Do these sims account for ICM postflop? If not, would the mid stacks more sensitive to risk premia play tighter, shifting more jams/raises to passive continues or folds? How would this affect the other corresponding ranges?
In this experiment, the tournament simulations were based on push/fold ranges, so no postflop play. Our ICM MTT solutions however, account for postflop play.
Check out this article to learn more about how ICM impacts postflop strategy! blog.gtowizard.com/how-icm-impacts-postflop-strategy/
So what does it mean to adapt ICM solutions early on the tournament? I can see that ICM is superior to CHIP ev but quite clear on the application
Check out our new ICM solutions to learn how to adapt your strategy! blog.gtowizard.com/pko-solutions-and-new-dashboard/
Should we take into account field size when the LR isn't closed yet? Lots of online tourneys will be at a specific stage much longer because of players rebuying/buying in late, we then are at that specific stage longer and bubble will come later than if LR was closed
We tested various field sizes in this video, from 90 players to 1000 players. Late reg has the effect of artificially increasing field size (since there are effectively more entries)
This is an amazing. Well done.
Hope I’m the only one paying attention
How top heavy were the payouts in the experiment? I wonder if it changes anything if 1st pays 300+ BIs. The 100% ICM aka the Allen Kessler strat.
The payout structure was shown for each experiment.
But how does this change for tournaments where late registration is still open with 50% of the field still remaining, for example?
Very good question. You don't know how many are in the field OR the prize pool yet. I suppose you could estimate both of those things to get a rough idea, and also hands ICM says you can play are unlikely to be too different in those early stages. Different from chipEV yes, but not all that different in terms of early stage ICM strategy from tournament to tournament, even if in one you have 300 people who all registered at the very start and in another you have 500 people, a third of which registered late.
This really goes against what most oldschool players and coaches have been saying for years. Really shows how much the understanding improved since the TAG days
You have to ask why the most successful players in MTTs tend to be hyper-aggressive ones in the face of this sort of simulation.
There are systemic flaws in the underlying logic of ICM which only the top echelon of pros are aware of.
Specifically, the equilibrium or solutions presented in GTO Wizard or DTO or similar products are not accurate or stable in the way CEV simulations are.
If you want to follow ICM, but I ignore it, this can cost you a lot of money. This effect is immediately visible in the thought experiment where you shove 25 BB into me from SB with 80% of your hands because I am supposed to call only QQ+ etc because of pay-jumps. If I call TT here, I lose a little equity and you lose an enormous amount.
This is just an example to prove that the concept itself is not 100% correct. At a minimum, the 20bb solution here will have to be different versus a person who cares about ICM versus one who doesn't and it's quite complicated when you consider that in real life it's not binary and you won't know your opponents level of awareness/care about ICM. In other words, the idea that you can apply this solution and you will be absolutely fine is false, as the solution assumes opponents care about ICM as much as the player does, but it can be more or less.
One more point is that the nature of ICM is a systemic flaw in MTT poker design which incentivizes extremely tight play which isn't very interesting to watch, and sites are already migrating towards PKOs where the exact opposite logic applies.
@@strategygames1026 It's easy to ignore the hyper aggressive tournament players who bust out early. The hyper aggressive tournament crusher is a combination of exploitive play and survivorship bias.
So much for the old saying; "Play for the win!"
Amazing content. Thank you!
Simply amazing! U guys are awesome!!
How did you switch between formulas in GTO software with a press of a key?
You can "Star" solutions in the solution selector, filter by starred solutions, then press up and down to switch between them.
For this video however, I simply opened many GTOW tabs and used the ctrl+tab hotkey in chrome to cycle through the solutions.
@@GTOWizard Amazing - thanks!
So if the buyin is 1000 bucks, you are making 30 bucks as the ICM and -130 as the cev strategy. The first one will play what - 10 hours on average and the second one 2 hours? If you value your hour at let's say 30 bucks / hour, the first one is losing 270 bucks on average. The ICM guy is losing 190 bucks on average.
Yeah, I think it's pretty clear which strategy is better...
I agree icm starts to matter when 40% of the field is left more or less. But dont you think it's much more significant when you have more starting stacks? It makes sense folding more if you have around 4 or 5 starting stacks, because that stack may be enough to cash the tournament. However if you only have 1 or 2 starting stacks you shouldnt be tighter because of icm since that stack isnt worth preserving? Isnt there a relation between chips and icm?
If you watched the whole video, you would notice that this question has been answered. Short answer: yes it matters. For long answer: watch the whole video
I have watched the whole video. Stackdepth isnt the same as number of starting stacks. The avarage blinds of the tournament doesnt really matter.
@@mariodiaz3976 The lower your stack depth is the more ICM matters to you. Every chip is extremely precious when you have few of them and you really need to pick your spots. You can afford to wait much longer than people generally think, do not go shoving / calling with 15 bbs just because you are short. The bigger your stack is, the more freedom you have, even according to ICM. If you have big stack in final table, ICM even expands your range of play since others should be more careful!
Great content !
Imagine the new poker tech we will have in 10 years... I don't know if poker is screwed or getting better
It's screwed online, but live is alive and well. And honestly, most players dont have the time or brainpower to absorb an entire ChipEV strategy, let alone adding a whole new constantly shifting strategy like ICM. Like, the ranges change based on how many are in the field and also based on both absolute stack sizes and how big your stack is relative to the rest of the table. How many ranges can you memorize? And that's just preflop. GTOWizard doesn't even have a single postflop ICM spot yet, though i'm pretty sure HRC can calculate those spots.
@@lemmyorleans that's what I figured
There will be a shift away from NLH, which has already been deeply analyzed & studied, to mixed games, which can be constantly evolved. Poker fundamentals and adaptability will outperform chart memorization.
@@lvzee 5 Card hi/lo pineapple is the future
I'm an on/off mtt player for nearly 15 years and this is how I always found success. It goes hand in hand with the old timey mtt pros preached about increasing aggression when approaching bubble to win more tournaments at the cost of variance.
Do you calculatre the field size at the beginning of your tournament or at the first hand after no re-entries are allowed? I assume after re-entry.
These were simulated without reentry
Does latereg have ANYTHING to do with this? I seem to unconciously switch to "icm" when latereg ends. Keeping a healthy stack and taking less marginal sports because i feel like Im working towards the bubble since that point.
Late reg has the effect of artificially increasing field size since there are more entries. The impact of ICM on these spots depends at what point latereg ends.
It does make a huge differnt it is push fold u can have a get chip stargety without shoving right before the money with this sim all the icm aren’t going to go all in at all near the bubble on the sim so of course the chip stargety is going bust before the money
if you wanna be rich, you gotta be a nit
so the meta is to play tighter ?
So it's always better to max late reg in mtt's?
And stall every single decision.
what kind of risk premium is assumed for 100% ICM early game in the sims you ran?
It depends on the stack and field size.
We've recently added ICM sims for earlier MTT stages, as well as a features that displays the bubble factors. blog.gtowizard.com/pko-solutions-and-new-dashboard/
Mr wizard how mutch does ur coaching cost monthly?
All premium members have access to our weekly coaching! You can also view previous coaching seminars on our website via the coaching page. As of today, we have 85 hours of coaching content!
I could never figure out why chip ev was good in tournament i could only see a longterm strategy being good at cash games but im a noob i donno nothin
Can you add in ICM to start switching to cEV at various breakover moments?
I’m interest in ICM until the bubble and getting more risky once bubble bursts.
ChipEv is good for sattelites where only 1 player get paid
Tight is right..
sadly all these examples loses vs rake :D
luckily we dont play vs these sims.
good vid
Thanks for turning poker into bunch of nerds using calculators instead of their brain :D
TLDW: Tight is right after a while.
Can anyone use the simple words to explain how to play ICM in a real tournament? Thanks!
Play tight, play aggressively, and remember that survival is key
Skip to 20 mins everything before is a waste of time
ggwp
how can I take anything this says seriously when you say heigTH?
What part of the video are you referring to?
not using gtowizard mid mtt?
You never mention the percentage of the field that gets paid @gtowizard
15% of the field is paid. Details of the payout structure are shown at 2:55
@@GTOWizard got it thanks . Should be made clear in the application too on the menu