Theory: Variance vs. Style
Posted by Phil Galfond
Posted by
Phil Galfond
posted in
High Stakes
Theory: Variance vs. Style
It seems to me, the general consensus is that players who 3bet and 4bet very frequently, who barrel off more often, and who make more call-downs have higher variance than those who are super TAG. Obviously, results seem to show that LAGs do very well in tourneys, which I suppose people could explain in one of a few ways:
1) There are more LAGs nowadays
2) People who succeed as LAGs are better players, because it takes more skill to be a LAG
3) LAGs do have much more variance, but they have so much more EV that it's worth the trade-off
I'd like to enter another explanation and see what you guys think of it (or perhaps people already think this… I'm not in touch with the MTT strategic community):
Good LAGs experience significantly LESS variance than the standard/strong TAGs.
Strong TAGs get most of their edge by playing better hands than their opponents. However, these edges come into play in large pots. After all, if you have to wait around to catch a hand (folding your antes and blinds), you need to win some medium-large pots when you actually do have a hand (yes I realize TAGs can/do bluff too… I'm simplifying). Given that TAGs fold most hands, then play medium-large pots with a big edge, their results are very dependent on winning these big pots. So, even if I'm a TAG and I get KK in v ATs, or top set in vs. a flush draw, I'm out of the tourney 1/3 of the time in just that one pot. If I run KK into AA, or an overpair into bottom set, I'm out. This part is true of LAGs as well, but….
Yes, LAGs do play large pots, but most importantly, they play a ton of small-medium pots (with an edge). When a LAG makes a +EV blind steal, +EV 3bet, 4bet, or flop c/r, he's playing a small pot with an edge. Yes, he finds himself in more large pots, but the fact that he has so many opportunities throughout a tournament to realize small edges makes him not nearly as dependent as his TAG counterpart on winning the big flips, 60/40s, or cooler situations. By entering many small pots with an edge, he's essentially decreasing his average bet size and making more bets (bets meaning, like, +EV gambling bets, not actual flop/turn bets).
Anyways, I think you get the point. Do people agree/disagree? Is there any way to model this to provide evidence to support or refute it?
Loading 52 Comments...
I think the real key here is understanding when it is and isn't ok to be a LAG. Here's an example that I've been discussing with a friend.
28 players left ITM of a massive 10k live event. We're chip leader with 120bbs and play is 6 handed today. Another player at our table has 115bbs, and the other 4 guys have between 20-60bbs a piece. The remaining field is very soft and keeping our variance down should be a top priority. I'm under the impression that it is ok to be as LAG as we'd like to be against everyone at the table except for the other big stack. I think theoretically vs the guy with 115bbs we should not have an all in preflop range because if we both made enough raises to get 115bbs in, one of us has to have aces and the other guy has to be making a huge mistake. I know that sounds weird and remember I'm trying to talk in theory. I'm sure in practice plenty of guys will get in 99+ AQs+ against you and you can exploit them by having an all in preflop range of KK+ or whatever, but I have a hard time finding a way for it to be ok to get all in preflop in this scenario without AA. Our stacks are worth too much and taking a chance of being crippled feels very wrong. I think there should be a lot of call pots with 1-3 streets of betting between the two big stacks, and very little raising. Now once the stacks get further from each other, say we grow to 140bb and the other guy drops to 80bb, sure, go after him now. Ill end my ramble here. Curious what you guys think.
Ben Tollerene
It's tricky, because in the situation you described, it's a constant game of chicken. Both you and the 115bb stack would HATE to get a lot of money in against each other, which means that your actual value ranges (to stack off) should be exactly AA, as you said. Given that, the first one to shove wins the pot a substantial amount of the time, especially postflop where AA isn't guaranteed 80% eq anymore.
I think it probably would be beneficial for a big time MTT reg to get the reputation for spite calling/jamming vs. other regs in these spot (and to continue doing it anyways). Seems like the benefit of always winning the game of chicken would outweigh the times you run into AA.
As to your post, yeah we definitely want to reduce variance and keep the post small. In practice, especially if you have postflop skills (like you do), and against players who will incorrectly put too much in preflop (hurting you both), you'd probably be better off with your strategy, rarely raising and playing a 3 street game.
of course, this is ICM 101. if either of those players GII with 99 through eh QQ even, AQ suited, ak off, that is a real big potential mistake. now, of course there is poker to play, and if your opponent is getting it in light and you have a strong read you adjust accordingly.
In the example of Benjamine I have 0 problems, if there is some agressive gameflow, to get it in with AK vs the other bigstack in a cashgame. But it's a massive leak doing that in tournaments. Even vs the moorman's and NSBs in the world, there is a better way to play your AK then gii preflop.
There are people who see a +EV spot to three bet 100% of hands at that moment, there are other regs who will wait for a playable hand before three bet bluffing, even if you think you can exploit the opens of this villain by three betting. I am definitely in the latter camp, but I wonder if I am wrong. In cash games you can afford to wait for a good bluffing hand if you notice an exploitative tendency. In tournaments you might only get one chance to exploit someone, before one of you busts, your stack size changes, your table breaks, table conditions chance etc. My problem with that strategy is you need to be sure you are right in seeing an exploitative strategy, if you are wrong and he folds to 3bets 45% of the time, you may end up playing a 3bet pot with 82o.
I think LAGs would experience more bankroll variance because of their slightly reduced ITMs, but not by a significant amount.. However, you are talking about stack size variance within a given tournament: that they can build stacks to protect them from being all in for their tournament life. I am not sure if that's true, I have seen LAGs building stacks and TAGs bleeding off chips and vice-versa but have no idea the frequencies I see them. I am not sure that LAGs do better in tourneys, there is a huge survivorship bias in tournament poker and there is so much variance anyways that it's tough to draw meaningful conclusions based on the types of players who seem to do well.
"There are people who see a +EV spot to three bet 100% of hands at that moment, there are other regs who will wait for a playable hand before three bet bluffing" i like this point you make and for the most part im also more in the latter group here. I view it as kind of insurance, to have not too wide of a range when you can only speculate to a certain degree given circumstances that this would make a good 3bet spot. so that for the cases you are wrong, you on average still not make the mistake 100% of the time, plus when you have a bunch of blockers in your 3bet range you increase the likelihood of being right, hence also reduce variance. So i like using this at least as one of the indicators for differentiating between lags and tags.
Im actually also not sure if nowadays the LAG style in mtts is still the best, there are a bunch of big winners even in HSMTTs with high roi out there who play actually pretty tight. At the same time though, looking at players like pessagno or thebattler (unfortunately not playing anymore on stars, loved his style), they play super lag even to the degree that some regs are wondering how the plays they make can ever be profitable, then you look them up and expect a high variance graph and the opposite is the case.
Overall i think neither style is necessarily superior to the other, one just has to make the right adjustments with whatever style one chooses. At least i dont personally see any style being particularly more profitable nowadays, while few years ago would have def said LAG is best.
ps: oops, guess i came late to the discussion, thought that was a new thread
Also from a time efficiency side, LAGs tend to either bust quickly or run up a big stack which is much more time efficient than playing a whole tournament very tight and perhaps squeezing into the money. It is why when I play tournaments I go full out rambo style and just try to play a lot of pots in what I think are mildly +ev situations/steals etc where I just want to run up or bust. I think in the short term playing a high variance LAG style is much more +EV and like you say lower variance than a TAG style. You may cash more frequently and be losing less money in the short term, but as long as long term profitability and variance goes playing LAG is much better imo. Also when you play a lot of hands and play very LAG it gets into a you vs table dynamics since people are gonna be playing so many pots vs you, and this can lead to players playing somewhat straight forward to me in my experience [at the lower stakes, mind] since they think we're a maniac and just wait for a hand, and when they get a hand it is super obvious and they aren't winning nearly enough off of us
87 <3's, aren't i popular
A lot of LAGgy players credit their playing style for their huge swings but it's actually because their edge isn't very large, often due to making huge ICM suicides and +cEV but -$EV plays (punts!) that they simply chalk up to being their style or "playing for the win". Both LAGs and TAGs variance is affected by their winrate but typically the best LAGs will have smaller swings simply because they're better poker players and their ROI is higher.
The problem with the TAG style is that traditionally (circa 2004) they used to get all-in often with massive edges against super weak players, often times with their opponents drawing dead. But over time as the weaker players improved, these TAG-based exploitative situations become far less appealing as edges were smaller, risks increased, and more early tourney lives lost. In walks the LAG. Taking a lot of minor +EV spots with maximum aggression didn't make a whole lot of sense until the alternative (playing TAG) became less attractive. I think the game's evolved to the point where (as Sam said) it's criminal not to take any substantially +EV play provided it surpasses the threshold of profitability with tourney life factored in.
"I've actually been thinking about this kind of a lot for a guy who doesn't play hardly any MTTs. I think a big part of it is that winning chips without showdown is much lower variance than winning chips with showdown. I would say there's no variance in winning pots without showdown, but you still have to take on risk and put money in the pot to win the NSD pots." - Ben
I think this quote really nails it on the head with respect to variance. I think it would be very interesting to take the top 10 online MTTers (over a substantial sample) and compare their red-lines. Intuitively, I think the game has evolved to the point where the best players are simply the ones that are closest to maximally exploiting NSD situations while being sharpest in all the forced SD situations. The nature of MTTs in general offer a lot of largely exploitative based +EV opportunities with individual hands in a vacuum in contrast to online cash games where ranges tend to be more evenly distributed and its more a game of optimizing range vs. range exploitation. This would seem to imply that the *truly* best MTTers are likely doing very well in NSD relative to other MTT regs.
Should have prefaced this by saying this is all very speculative based spewage given that I've played all of 0 online MTTs in the past ~2 years.
As a winning small-mid MTTer on the softer US sites who considers himself to be somewhere between a standard TAG and very good TAG, I think this post totally nails it.
"it's criminal not to take any substantially +EV play provided it surpasses the threshold of profitability with tourney life factored in" -THIS
I think as I alluded to earlier what separates LAGs from TAGs is a willingness to choose to exploit people's tendencies regardless of their hands. There are times when people go overboard, doing stuff like opening 84o UTG or, but there are also times they find a profitable three bet with a hand that a TAGs would snap fold 100% of the time.
One thing I notice good lags do frequently, is 3bet very wide ranges as bluffs, when they are only commiting say 10-20% of their stack, including basically pure bluff hands, i.e. 84o k5o etc. The reason for this, as far as I can see is with antes, and players opening small, you end up getting a very good price on your bluff, particularly vs certain players who will have not only a high fold to 3bet, but a high fold to flop cbet on basically all textures, because they are not doing enough rebluffing, and they don't protect their range by flatting 3bets strong.
Once antes kick in everyone's opening range from every position should be substantially wider. Most of the time antes change the dead money each hand from 1.5bbs, to nearly 2.5bbs which obviously requires a large strategy adjustment. I think a lot of what the laggier regs do is exploit these spots vs weaker players who are playing too tight, however trying this at a tough table, your likely to be, 3bet, cold 4bet or 4bet by the opener often enough that looser plays cease to be profitable.
I think some people take this to too much of an extreme and try to bully people in spots where it is completely inapplicable (e.g. some reg trying to battle another reg too hard in the early ante stages of the 109c isn't going to work out well for them) but when you look at players destroying people on big bubbles or in big ICM situations where they have the big stack it's because everyone else is caged and too scared to play back.
There is a general consensus that turbos, especially the high BI ones are of more variance than reg speeds.
You will find though that if you take all the small edges possible vs fish and most Reg Speed regs and add them up, you will find that in the end, Turbos are possible to have smaller variance than Reg speeds.
Playing a high variance style means in terms of Game Theory taking a big risk which pays off a big reward.
Problem is that with the RNG&the huge fields involved you can end up being massively +cEV but still coming up short.
That means that playing LAG when you are CL or just Table CL is profitable in a lot of ways because all TAGs do effectively waiting around for an 8% hand and going with it.
Again as Jono said above, its all about punting stacks when its simply not worth it ICM-wise, because yeah he does play 25/22/9 over 2k hands but still he can show up with AA and you will bust 80% of the time.
Last but not least, I totally agree /w Scott on the Top 3 fact. If there wasn't for min-cashing we all wouldn't have a br to play with.
Cheers.
My personal point of view is that LAG nor TAG is the best way to plays MTTs. The most profitable style is obviously a mix of both. Someone said it here before as well, play loosers vs the weak players and tighten vs the regulars who mostly won't assume you do this and have wider ranges themselves and put you on a wider range then you actually have.
Some players who are big winners and get lots of respect from the MTT community are just horrible at adjusting vs there table and villains or are just to arrogant and think they are way better then they are.
One example is the LAG player being in the HJ and behind him are two good lags and a (imo) top MTT reg. The LAG player opens with 20BBs with 89o.
On an avg table, this will work a lot since you dont get flatted a ton and you feel like you can read ranges from people well enough to know when to fire or when to shut it down.
But with that line up behind you, you have to give credit to people, especially when your own image is pretty screwed up and people know your game and know how to play back vs you.
Most LAGs don't adjust well and to often put in the last bet where it's obvious that the player who they play against is 100% knowing what they are going to do and anticipated/widens his value range to much for it to be profitable to play back at.
To finish back on topic; Most of it has been said here about LAG vs TAG and variance. Optimal strategy is a good mix of both. This should create the highest $EV and lowest variance.
I feel like in most games $109 and under, you pretty much have to be just LAG enough to beat up the fish at the right times and not have regs tag you as 'uber nit who always has it', while also being TAG enough that you're not getting stuck in a bunch of spots where you 3bet a reg with whom you have history, he 4bets you, and all of a sudden 5bet jamming 40bb with K7s becomes completely standard for you every time a reg opens in middle position. There are just so many edges to be had outside of these spots that you're throwing away a ton of equity when they do occur, because you're not only throwing away your current stack X% of the time, but you're also sacrificing the lower-variance opportunities to exercise your edge which are going to come along later in the tournament.
I think there's some confusion creeping into the discussion because people aren't properly distinguishing between two types of variance, first standard deviation which is measured in bb/100 and second qualitative variance which is how big swings you take (either up, down, or break even).
Take two hypothetical players, Tagbot and Lagbot. Tagbot plays like a TAG, and Lagbot plays exactly like Tagbot in all cases except for those vaunted 'low variance non showdown spots' we've been discussing. To take an example of a spot, antes are on (there are 2.5bb of dead money in there), Lagbot is on the button and the two blinds fold to steal % sums to 60% for a stealsize of 2.5bb- so a steal with any two cards wins at the very minimum .4(-$2.5bbb)+(.6)(+2,5bb)>0. Since by assumption Tagbot and Lagbot play identical in all situations apart from this steal, stddev(tagbot)=(stddev)(lagbot) before considering the range of hands lagbot plays in excess of tagbot in this spot. We'll call this posterior standard deviation number 1 for simplicity's sake. What's tagbot's standard deviation in this spot for the set of +ev hands he does not play but lagbot does? Well, it's 0, because he always folds the subset of hands we're considering. So, tagbot's standard deviation is still 1. Lagbot however plays some more hands in this spot. The standard deviation for raising here will be in excess of 0, so for this case, the standard deviation of lagbot is necessarily higher than the standard deviation of tagbot, StdDev(lagbot)>1.
Ok, Sauce, how exactly does this support your thesis ? Well, I'm hoping this example brings a little more precision into this discussion. What it proves is that if hands were considered in a vacuum and the taggier players plays are a subset of the laggier guy's, then the standard deviations are simply additive, therefore the laggier guy always has the higher standard deviation. Even if this is true though, that doesn't mean the laggier guy will have "more variance" in the qualitative sense which refers to his losing/breakeven stretches. Afterall, those non showdown low variance spots are often very numerous, and can really stack onto the laggier guy's ROI. They also might result in his opponent's playing much worse vs him (or much better! why do we always have to play worse vs Lags?). The thing is, at this point it makes more sense to throw off the whole TAG/LAG distinction because what this example shows is just that the Lagbot is a much better player and his ROI is way higher. In poker, winrates usually have a much larger range than ROIs- the range of ROIs for decent pros in MTTs is probably 25-125%, and the range of standard deviations for MTT players is more like 45bb/100 to perhaps 125 bb/100. Which is to say that there's more variance in winrates than in standard deviations for the population. 'Qualitative Variance'= is some affect of ROI and standard deviation. The guy who has the least of it is probably someone with a high ROI and reasonable taggy/lagginess (and hence a reasonable standard deviation). But there is absolutely no logical reason why we should think LAG play has lower variance. In fact it's just the opposite, in the presence of equal ROIs the LAG guy almost always has higher variance. It's only when we lump higher ROI into LAG play that we start to think the Laggy guy has lower variance. My guess is that people lump these things together because if two players A and B play 10% and 15% of hands respectively and each player by hypothesis only plays spots which are +EV, then B has a higher ROI (assuming the magnitudes of the +EV of the overlapping 10% are fairly similar which is true in poker, in fact we expect B to win more from the overlapping 10%). However, this sort of thinking shouldn't mislead us into thinking the Lagginess is the cause of the reduced variance in the LAG population, it's just the opposite.
Glad this discussion has been brought up - I've been an MTT pro for a while and felt that the best way to describe the most profitable way to play MTTs is the chip utilization theory. This isn't a groundbreaking concept but never hear it mentioned and its clearly very important. The top MTTers are very LAG naturally and will adapt from there based on table dynamics,stacks,etc. These top players have very large non showdown winnings which is the core of winning MTTs - knowing how to play 20 bb stacks is important too but in my opinion not the key...the chip utilization theory basically states that these players when they get a big stack will create such massive +ev scenarios vs everyone else that getting that stack in a neutral or even -ev manner might still be very profitable. The importance of the big stack is based on effective stacks,average stack sizes, and the skill of the remaining field - say we have 100 people left in a sunday major - we have 30 bbs and we're in a hand with a villan that covers us - given how soft the field will be on average and how many +ev spots we can create with 60 bb, we could take a -ev spot now to create huge +ev spots in the future. Some of the biggest (online) MTT winners aren't the best hand readers but you'll see them take crazy bluff lines that may cost them money at the time but offer massive future value if it goes through. I'd imagine if you brokedown the stats of big MTT winners their NSDW(non showdown winnings) would be through the roof when over say 40bb....thats the gist of my thoughts - would be curious to hear what you guys think.
Great thread. To take Ben's post above and consider this from a different perspective - that an individual MTT stack is a bankroll.
The good LAG, who has a greater ROI because he is playing additional small pots well, will encounter greater variance but will not suffer as great a chance of busting because of the enhanced ROI (Kelly criterion) providing his ROI to variance ratio is above a certain (magical) number.
Consequently the LAG will have a lower risk of ruin (but greater variance) in the tournament, without withering. Add up the result of all the MTTs and the variance in the LAG's results should actually be less. That makes a lot of sense to me. Does it make sense to anyone else?
I believe that a reason why LAGs seem to do better in tourneys is that because they have a style with higher chip-variance... they are more likely to win them. A LAG player is more likely to accumulate chips, but also more likely to bust before a TAG player. I believe the result is that a good TAG player is more likely to cash, but less likely to win - and since this is a "seems" thing - you are basing it off of basically who "seems" to final table and win. This doesn't necessarily mean that a LAG players' ROI is higher - although it could. That is a different question. That's something Scott Siever jumped in to try and address - in the negative, in his response - namely that you don't have to win/top-3 in order to have a good tournament ROI ... that plenty of it comes from cashing. However, because a LAG style means that you are slightly more likely to win a tournament, according to my logic (and something I believe) - it SEEMS like they do better in tournaments.
I personally believe that the actual EV of the player, both from plays - but also from how badly the player gets others to play against him - are what dominate the actual ROI of a player... and that LAG players are actually more likely to win tournaments.
It actually seems kind of simple, and is something I've assumed for a while ... but I could just be totally off here. If you have a zero-sum / can't add on situation, which is exactly what a tournament is... isn't it true that a more aggressive strategy that goes boom or bust (but with the same EV as a more conservative strategy) will be more likely to win the tournament (but also more likely to bust earlier)? It's the basis of what I'm claiming here - and I believe it... although I haven't proven it mathematically - and definitely would be interested to hear why I'm right or wrong.
On the crappy post-Black Friday American sites I think TAG may actually be better than LAG because due to those MTT structures the avg. stack size is much more shallow than PStars and FTP and so there is less margin for error when playing a LAG style.
Also, lower buy-in MTT's are probably more conducive to playing TAG since a big part of LAG play involves bluffing/semi-bluffing which in turn relies on your opponents either being extremely timid, or thinking players who won't station you in spots where it appears you can have little to no bluffs in your range.
Anyways ...
I think that's just a small piece to this discussion but felt it worth mentioning because I think a big mistake less experienced players make is applying high level concepts to low level games and that usually has a very unhappy ending.
Without being able to quantify it, intuition says playing a LAG style should lead to winning larger pots with premiums since your range is much wider than a TAG player and the TAG player may be forced to run into spots where they cooler their opponent or hold vs 3 bet shoves stacks at their tables.
Tom, regarding post BF American sites I'd have to disagree with you about TAG being a better strategy. Yes stacks are more shallow and there is less margin for error, but there is much more benefit to busting early or winning. The payouts are very top heavy and the fields are so small that a final table doesn't yield you much of a reward, you need to top 3. Now that they have implemented ridiculous late reg and reentries it caters even more to a LAG because you can beat on people enough to force them to make a lot of mistakes and take the small edges when they present themselves knowing you can just get back in.
I also feel the skill gap is pretty huge on these sites right now. You have a lot of people that either gave up playing for an extended period of time or just have went about it casually since BF. Then there is the people that have never played the major sites with everyone else in the world. There are a ton of bad regs and fish but very few really good players. The majority of the field just doesn't adjust to LAGS well enough for it not to be the best style. As long as the long is adjusting to the small # of people who do have a clue they run over the majority. Look at the biggest winners on these sites since BF, most are LAGS. I think the gap is only widening the further away from BF we get.
on WPN the BIG 10 tournies have a favorable structure for TAG (5k and 15 min levels) and they seem to be more consistent with deeper finishes
Agreed. Some of my friends who are quite good at MTT's are routinely getting to the last two tables or better in the 50k.
This entire thread is great, thread porn if you will. Agree with everything being said and it's refreshing to hear everyone agreeing with this point of view. If everyone starts at the same stack and (over time) gets the same hands, then a LAG will clearly fluctuate way higher and lower causing more bustouts but more wins. Obviously there are road bumps along the way (money bubble, ft bubble, icm, etc.) causing changes in optimal play style but if the LAG is putting him/herself in more situations to make chips then like you said they will be able to withstand some bad luck later on whereas the TAG usually finds himself out of the tourney.
I will say that nowadays a lot of good players aren't TAG or LAG but whatever they need to be for the given situation, like Sam stated.
All that is said in this thread is pretty interesting.
As for me, I feel that since games are getting more laggy, with more people willing to steal / resteal, play small ball, etc, a good strategy on aggro tables with competent players can be to cultivate a very nitty image (not taking obvious steal and resteal spots) mixed with a very agressive postflop game and / or a very tricky and inducing postflop game.
Anyway, we all know that poker is a game of adaptation. If everybody is playing a bit spewy preflop, then patience increases in value, and traps / slowplays as well.
As for the variance question, having a very conservative preflop image can allow for huge fold equity postflop against decent players. I believe than taking less spots but having a higher chance of success in our plays decreases our variance => we should be more likely to accumulate chips without showdown.
I colour code the player type you describe. The problem with playing that style is that because it is incongruous it is quite easy to identify, even in a small sample, and then adjust to imo.
One of the most interesting threads ive ever read and with all the poker IQ in this thread Im not going to even attempt to add anything. I thought Ben Sulsky's example was super interesting and proves alot and it was refreshing to hear scott siever state the importance cashing.
So if accumulating chips to place top 3, and cashing are both huge factors that impact ROI% would it be reasonable to assume that its optimal to play more TAG and take less marginal spots before the money bubble and then switch to a more LAG once your are ITM in general? Even though its not that simple because of stack size, tables game flow and other factors.
My idea is couldn't it be that the style to play a tournament should depend on how big the field is in a given tournament. IMO you should play more TAGish in small field tournaments where you have a bigger chance to get in the money (cause lets say you just need cooller or few to go your way, to win one flip or two and so on so then you have huge chances to make it) and just getting in the money will affect your ROI much better compared to later payjumps (I mean late payouts for ex. Final Table payouts) which will not affect your ROI as much cause these money isn't so big (compared to buy in) and hasn't near close enough impact as late stage payouts of huge fields MTT's (where it's hundred times x your buy in) where in those your mincashing means sh*t to you ROI and you should not play for mincashing at all and ONLY focus your play to get to the Final Table for exz.
So my point is does these assumptions have any corealations when choosing playing styles (TAG or LAG) and if so, might it be that style should switch (what I think at the moment) to more and more TAG'y (unless you can exploit someone pretty hard) as huge fields tournament gets closer and closer to it's end.
For practical exz. Scott Seiver plays more small fields, bigger buy ins, tinny field highroller NL and other games MTT's so his mincashing has bigger impact on his ROI so his style should be based on TAG'y side cause he needs to reduce variance at the early stages (as we asumed that LAG's might bust out earlier more often vs TAG's) and should be more concentrated on geting in the money at first while if we take some other guys who plays mainly sunday huge fields majors they need to play more LAG'y since the beginning of the tournament and they should concentrate ONLY on making it to the Final Table (taking all even barely +EV risks along theyr way) and not focusing on mincashing at all also not waiting for "huge" hands, which would be pretty hard to get cause they'r way is much much longer than Steves and you need to be at the top of your variance if you are realying your win on big cards, coollers, etc. to your side.
And looking from one more perspective, lets take a critical exz.: you are deep in huge field tourney and lets say you are there 10 times in such a spot so (as I told exz. critical) you CAN or SHOULD 9 times out of these 10 oportunities take a risky spot to get your ships in making even barely +EV play and lets say you HAVE to (or you are) sacrifice(ing) (or just LETS you sacrifice) these spots that even if you get to the Final Table 1 time out of 10 where the biggests payouts that one time it will still be for you (and for your ROI) so +++EV that it will ransom your risky plays given how many times you will multiplicate your buyin. While on another hand a guy like Scott Seiver cant afford such an oportunity to bust out 9 times out of 10 (and he doesn't need to) cause that 1 time when he gets to the top his bustouts won't ransom his huge risks.
Hope I made it clear. What you think guys? :-)
P. S. No. 1. It might be just an imagination that LAG's are doing better cause theyr wins are more often made in huge field tournaments where prizes are very big itself so they are much more "noticeable" when somebody wins them. While TAGs are winning smaller prizes itself, but does it more constantly than LAGs do.
P. S. No. 2. Can anybody advice/talk/discuss how to play top pairs in MTT's? Question can appear silly or some of you might laugh, but I'm asking seriuosly :-) Cause if you think deeply enough it might be the hardest range to play with (ofc we need another thread on this one).
Paging Nick Rampone...
gotta be the most interesting post ever.. especially given the number of clever cookies who have posted here.. top man sauce
"We all can agree that in an MTT in which we have a large skill edge, we'd like to reduce variance as much as possible (without sacrificing too much EV)." -Phil Galfond
I've been pondering on this thought for the past week or so. Why is it that we want to reduce variance as much as possible? Why don't we want to push every single small edge? I cant find a good answer to this question. Someone please chime in.
Its not like if we bust the tourney we get shot or something, then i could see why reducing variance would be beneficial. Also if you had a last longer bet with someone then i can see why reducing variance is beneficial. But when we are trying to build a chip stack why is reducing variance beneficial?
Great thread. I agree that a lot of what Phil is saying is the LAG's advantage in the OP will be quantified via a higher ROI, not via lower variance. Sauce's post on TAG/LAG should clear up the standard deviation angle I think. A TAG with 40% ROI will have lower variance than a LAG with 40% ROI, but the TAG may still have greater variance (in terms of downswings and overall bankroll fluctuations) than a LAG with 70% ROI.
@daniel...busting the tourney is sort of like getting shot isn't it? We lose all of our equity and the buy ins mount up working against our cashes
What a great thread. I know it's old but I read every word. maybe commenting on it will help other new members find it. Any chance we could this topic rehashed? That would be great!
Super interesting thread, read about 75% of it before realizing it was 6 yrs old!
While some of the concepts discussed are a bit dated at this point, I think there is still some super useful takeaways.
The thing I found surprising reading through this is all of the discussion centering around LAG vs TAG. I kind of feel like in 2020 that is a false dichotomy and that the best players could be described as "TAG" in some stages of the tournament and shift their strategy to what could be considered "LAG" in other stages.
I have always felt min cashing is pretty important in tournaments, more so than most of my peers. It seems like over the years the trend has been that people are starting to value the mincash a bit more. I think in general the right approach to MTTs is to have a more "TAG" approach all the way up to the bubble, then shift to what could be called "LAG" from the bubble all the way to final few tables or final table, and from there on you are basically at the mercy of ICM and your stack size dictates a lot.
Once you achieve the mincash, a tournament (pretty much any tournament) is just going to be extremely volatile all the way up to FT or Final 2-4 tables depending on structure. In this phase volatility needs to be embraced and you kind of have to just go for it. As soon as the bubble bursts, strategies should and do shift dramatically.
I like a lot of the pros and cons for playing LAG, but I think the main reason I have tried to change from being fairly TAG about 2 years ago to now is down to how poorly average regs adapt to a LAGy style. vidmate save insta
Be the first to add a comment