Yes,because you are pushing the edge from your full range in several spots-and that means sometimes stacking off with the worst of it.

"If you basically always stacks off with the best equity,you are playing too tight and leaving money on the table." -Eric Seidel.

Sent fra min SM-G991B via Tapatalk

Variance is like shorting a stock in the sense that the downside is theoretically limitless (though in the case of variance, the upside is also limitless), there is no upper boundary to how bad variance can actually get.

With that said, we have access to a variance calculator that can simulate how bad it can get over massive time periods, and luckily a few super geniuses in the thread gave everyone a quick refresher in middle school math (that funny little number to the top right of your term is called an "exponent", and believe it or not, those funny little numbers don't always have to equal 1. When they don't equal 1 they do some pretty wild things!) so everyone here should be able to simulate what their variance looks like. The great thing about it is all this **** about playing bad when losing or not playing your a-game are all going to be accounted for in your winrate and stdev numbers, so the precision of the variance calculator is likely quite a bit higher than you might think.

you're right, I'm wrong. You made me look up the definition and rethink it.

Certainly, can be. My understanding is that a 'cooler' is a standard spot where theoretically correct play inevitably leads to a large loss. Not practically correct play - which would result in folding KK preflop for 100bb when your opponent has AA. Obviously if we could see the cards we would open fold, but we can't, so we don't. Instead, we play correctly and 3-betting or 4-betting... finding all the money going in the middle with a significant equity disadvantage.

Or am I misunderstanding what a cooler is?

This is exactly what I was thinking, that once we have a large enough sample (pretty impossible to do, but gets better with more data) our times of 'bad play' will be accounted for within the simulation because our win rate and std dev already reflect those times. If we had samples of only our 'A-game' for thousands of hours, I would expect my win rate to have been quite a bit higher.

The entire premise of your argument seems to completely ignore my previous statements.

The fact that you push every edge while most grinders do not is exactly what I was talking about in terms of stylistic differences that could result in you having a long breakeven stretch and them being incapable of having a long breakeven stretch. This does not mean they are better or that they even have a higher hourly despite not having as long of losing streaks. I thought I made that clear before:

You're talking about losing flips and all-in equity when there are literally players that never put their stacks in the middle until they have the effective nuts.

I still have a hard time believing that someone of your skill level could play 100bb 1/2 full time for a year and have a losing year due to variance. That seems absolutely insane. If I lose a couple sessions in a row, I adjust my game. Is this optimal? Absolutely not. When I make adjustments, am I playing every spot perfectly? Of course not. I don't like having a losing week much less a losing month much less a losing year. I make adjustments to ensure that sort of downswing doesn't happen. A year long downswing would break most poker pros.

So what you are saying DC is basically that you are choosing to leaving money on the table,or accept a lower winrate longterm in order to protect yourself against the worst negative variance.

Fair enough choice i guess.

Sent fra min SM-G991B via Tapatalk

I'm curious as to how variance would cause the nit in the theoretical example below to go on a long breakeven stretch, or even a short one for that matter. Neither player is good at poker.

Theoretically a nit is playing a heads up match against the worst player to ever play poker.

This is the worst player's strategy:

1) Play every hand
2) Never raise, never lead, always check behind.
3) Always fold postflop when faced with a 1bb bet
4) Always go all in on the river when faced with a bet larger than 1bb.
5) Tell your opponent your strategy.

Nit's strategy to exploit the worst player:
1) Play every hand.
2) Never raise preflop
3) Bet 1bb on the flop/turn/river when he doesn't have a chance at the absolute nuts.
4) Bet 2bb on the river whenever he has the nuts.

This is the choice that nearly every single live grinder makes. Very few take every edge possible. Most live low stakes pros are close to broke. They can't risk going months without an income. I wouldn't say it's a "fair enough choice." I'd say it's the absolutely logical choice for most.

BTW, that's leaving theoretical money on the table, not actual money. Big difference. In reality, the player that is trying to take every edge possible could be the one making less money.

Sure.

Sent fra min SM-G991B via Tapatalk

Yep exactly. Ultimately the imprecision of the calculator is going to come mostly from the imprecision of the numbers themselves because, as you pointed out, it's practically impossible to have a sample size large enough to accurately reflect your true winrate and stdev.

I'm obviously an idiot, but I would guess that it might be actually be quite difficult (both theoretically with a lot of careful consideration as well as practically in the heat of the moment) to know with 100% confidence that every "pushing our edges" move is actually indeed correct.

Gnothatin',justsayin'G

This dump of a casino Encore Boston charges 10$ per hand rake and Im pretty sure they take it at $50, in 2/5 PLO do you think you could fold every hand blind UTG and it wouldn't effect your winrate and maybe improve by leaking EV in unprofitable hands?

When I played it the players were bad, but not like the worst Ive ever seen. Also just say you can make 40/50$ an hour in the game, how much would a 10$ rake effect your winrate opposed to say a 5/6$ rake. I'd say we used to get around 15 hands per hour ??? I was gonna boycott this dump but have nothing else to do really and its close by.

I'm really not at all convinced sacrificing winrate for lower variance is actually an effective means of lowering potential downswings or break-even stretches from a mathematical perspective. Goes back to the fact that standard deviation scales with the square root of time while winnings scale linearly with time. The longer you play, the more your winnings will be affected by your winrate rather than your stdev in relation to each other. Like it'll probably get rid of some of those shitty -400bb weekends or week long break-even stretches that are more flash in the pan than anything else but anything longer than that I feel like the hit to the winrate ends up mattering far more.

Is lol $5 maximum rake still a thing in some places in 2022?

Ghasn'tseena$5maximumrakesinceearly2016G

Still is in Vegas. Southpoint is $4+2.

You are 100% correct.

I typed up a bunch of responses along these lines - got too wordy and never posted.

Simply put - the best prevention of long breakeven and deep downswings is a good winrate. A good winrate comes from pushing edges, not dodging variance.

My (admittedly rather uneducated) guess is Vegas is one of the last remaining areas to have lol $5 max rakes?

Bottom line: Rake is always going up and if your LLSNL game ain't raking off ~$10 now then it soon will be.

GcluelesscomplainingaboutrisingpricesnoobG

Totally agree. Win rate is everything in the long run. There may be some consideration given theoretically to flipping 10,000bb at 50.5%. Made up the numbers, but I think you get the point. If the edge is small enough and the flip amount is big enough there is no doubt some place where DC's thought process may become relevant. Personally, I don't encounter these situations in actual play because - 1) Imperfect information prevents me from knowing my edge so exactly beforehand, 2) Bankroll management prevents me from flipping for obscene amounts.

As an aspiring player, my main concern is how to understand the variance in the game to use better bankroll management to prepare myself for the downside. Thankfully, I have a job I can fall back on if needed, but want to avoid that if possible and make more money playing poker. And I don't see poker working out in the long run as a viable career path without getting to the top of the field, which by definition is going to mean 'pushing edges'. I just need to have the bankroll to withstand the variance from that.

So do we still think that the standard "1 year of life expenses + 50 BI's (5,000bb's)" advice is sufficient for a professional player?? I'm leaning toward a larger bankroll than that if I want to insulate myself. For transparency sake, I'm right around the 1 year of living expense and a 50K roll on top of that, while playing 2-5, and 5-T as my main games.

At some level, edge passing is just a Kelly problem- there is always some (small) amount of edge x that we should pass on given bankroll y. At the end of the day you are trying to maximize ev over a long term horizon, not on a hand/session by session basis, and reducing risk of ruin can improve long term ev (downswing stops you from playing or forces you to move down), depending on how much wr you are giving up vs variance reduction. The actual cost of the downswing varies a lot person to person though (including if you are likely to reduce hours while on a downswing) so there isn’t really a single right approach to the wr/var trade off imo.

With that said, I do think it’s suboptimal to make these assessments on a hand by hand basis (ie this spot is close so I’ll fold to reduce variance) bc our estimates of edge are so imprecise and (imo) it’s very difficult to stop this from biasing your thinking at some level (also practically speaking I think the players that want to do this are less likely to be able to accurately assess that the spot is 0ev instead of ++ev, so it ends up being very net negative). Instead I think it’s better to make predetermined strategy adjustments and then try to play optimally within that, ie when shot taking deciding ahead of time on tighter than normal preflop ranges, but not considering variance in hand (particularly post flop). This isn’t really groundbreaking stuff, but I would definitely disagree with a blanket statement that passing on edges is always suboptimal, and imo it should even be common for (good) players who are looking to aggressively move up stakes.

I would guess a mathematician could create a proof that you are mathematically correct. But people don't play mathematical, computer programed, consistently correct poker. It is a game of emotions, tells, reads, luck, and some whacky stuff that occasionally makes thinking players shake their heads and wonder what the heck is going on.

My point is for me to play my best, I have to be playing with confidence and focus. I don't always have that for whatever reason. Sometimes I have to walk away for a walk around the parking lot to clear my head, sometimes I have to go into lock down mode, sometimes I'm very active and dominating. Sometimes making plays that reduce my variance is necessary for me to play well. I'm not and never will be 100% in control of my emotions. I wish I could say otherwise.

I have to completely disagree. SABR mentioned losing lots of flips, losing KK to AK, running KK into AA etc as examples of runbad. I can imagine if you are a TAG playing super deep that can be an absolute killer. Your entire month could come down to just a few of these spots. In the games I play in one can buy in for 1600bbs+ at times and achieve a higher winrate, but I'm not pushing those edges. I choose to buy in for 100bbs. Losing in these all in spots when I have big holdings is going to have little effect on whether I have a winning month/year given my playing style and relative short stack. Over a 1 month period I once lost with AA AIPF 7 out of 7 times and I still had a good winning month.

It *can* lower downswings. I think the majority of people who think they are benefitting themselves by folding in +EV spots are doing the opposite though. People underestimate the value and overestimate the risk associated with "thin" edges. Most edges that poeple think are "thin" aren't really. For example, I've seen people in this forum stating that it's reasonable to skip on 55% equity 100BB all in "coin flip" spots, which is kinda nuts if you recognize how little long term risk there actually is if you ONLY played these. It's the same risk as someone winning 10BB/h with a 100BB/h standard deviation, stats many people would envy, which we've seen is pretty low. It's a really good spot.

There are methods to incorporate risk into decision making which are more methodical and mathematically justified than "whelp, I crush these games so hard I don't have to flip with the fish here." I don't think it's that useful to learn because just taking every spot isn't going to be that much worse than an optimal risk averse strategy for someone who is appropriately rolled for the stakes they play. Just take every single spot. It's better for your image when you give action anyway.

To be clear, I'm not saying that. In my last session, I 4bet to $300+ at 2/5 with 77 to isolate a shortie with a weaker range. My main point is simply that not everyone plays the same so not everyone has the same chance at an extended downswing. A super-deep stack value-oriented TAG is not the same as a short stack bluffy LAG. Also, player pools can be different as well (not just in skill, but also stylistically).

Are they waiting to have the nuts on the river before they pile money in?

This is especially true in PLO, but you can "wait for the nuts" and still be stuck huge at all-in EV. There is no way around it.

I'm not strictly talking about $1/2 or 100 bb poker. It doesn't matter what the stakes are. You can run bad for much, much longer than you think it's possible or reasonable to.

Sent from my SM-G991U using Tapatalk