I just dont understand this fascination with SD. I just do not give a sht. Maybe cuz its im old school and did not have these apps etc when I started. I care about playing my best period.

Yesterday I played and got lost in a hand. This does not happen often. And it really pissed me off. I thought about it for a good slug of the night - and even woke up a few times thinking about it and trying to find out what happened/where I went wrong (it was only a 220$ pot). But it was enough to keep me up thinking/evaluating/retooling thoughts v villain types/and strats v those types.


Bickering about who is a stud and who is a moran is y this thread got nuked the first time. AS someone who had been around the block I can say with all honesty I really used to enjoy going into the casino etc. I no longer do - but I am too stoopit to do other stuff so I play poker.

All you young studs - give it 10k hours and get back to me. That is only 5 years of full time. Playing poker full time is a drag, and doing it for 5 years seems like an eternity. And if you do not have balance in your life you will not last

I personally don't care what my StnDev is either. I play the best I know how and whatever happens happens. Its not like Im ever in the middle of hand and Im about to bet the turn and say to myself "Oh crap, if I bet here my StnDev is going to go up"

However, when we are discussing poker in detail like we do here, sometimes there are ways to dig deep inside the numbers in ways that could improve our games. When I compare my StnDev to other good players, most of their StnDevs are much higher than mine. Maybe they are making certain more risky and aggro plays that I could use to improve my win rate. Maybe they are making too many of them and that's why their StnDev is so high but their win rate isnt any higher than mine. If I can figure out which of these more risky plays adds enough EV without adding too much risk, I can improve my game.

does anyone have a link/guide for correctly calculating win rate std dev in excel? my app is definitely not doing it correctly

It was used anecdotally intentionally. A better analogy would be if he knew xx non-important stat that is an effect (and not a cause, this is important) of his game.

Regardless, the point remains, the best live player I have ever seen did not know what standard deviation meant. He cleared $200K per year. This is not an exaggeration or used lightly. I'm not sure I've ever played with anyone that I believed could clear $150K.

There is a good story on these forums about when Phil Ivey first heard the the concept of EV and his reaction.

I think it's as simple as very large numbers of people run well below what they think is their expectation, and a few of them are correct.

I dont know what half the **** people here talk about means. Its because I learned from trial and error not from reading books and websites. I still win as much or more than most people. But that doesnt mean I couldnt get better if I understood and was able to apply some of the stuff people talk about here.

So again, I dont care what my StnDev is and the only thing I know about it is the higher it is the more crazy your swings are. But someone who is an expert on stats AND is as good as I believe I am at reading people, situations, betting patterns ect would be a better player (including your friend).

Still not getting it. Stndev is an effect, not a cause.

And there is nothing this kid could do to be better at poker, besides quitting when tired/tilted, which is on the scale of things you should be focusing on, is 20-30 knotches above stdv

I get it completely. The point that started this whole discussion is this:

JB has been mired in variance. He's been talking about it for months. It drove him back to a regular job. I dont know if he has posted his StnDev but I bet if he does you will see its super high.

His or anyone's StnDev being super high will give you a clue that he is playing a super high variance style. He may well be running well below average, but hes also bringing a lot of the variance pain on himself. Its self inflicted and its not increasing his win rate based on his own self reported results. That stat tells you that hes playing more big pots than he needs to in an optimum world.

High StnDev is normally associated with a high level of aggression which also leads to massive variance. Anyone with a high StnDev could easily lower their variance and most likely lessen their downswings by playing a less aggro game and if they can do that without hurting their win rate, why wouldnt they?

There are plenty of times where a shove or a fold is probably pretty close to the same EV. I might fold. JB might shove. If he constantly shoves in those spots he could easily lose a bunch of huge pots in a row and get crushed with nothing added to his win rate long term but it is soul crushing short term.

Early in 2017 I changed my game from already-nitty to extremely nitty (for example, I no longer feel playing 66-/Axs/KTs+/QTs+/JTs is profitable in EP and sometimes MP, so I open fold them; doubt you're going to find anyone more nittier than that on this forum). I also started BIing for just 66bbs instead of my normal 100bbs, again a step in playing an easy peasy shorterstacked and (hopefully) lower variance form of poker (although I'm not exactly sure if that is low variance style, tbh). It look like it started paying immediate dividends, as my break-even year to that point then took a 8.5 bb/hr upswing over the next 258 hours. And then I went on a cute little 3-7 session run over just 76 hours that tied my worst ever downswing of 955 bbs (of which I'm still digging out of).

I don't know anything about SD, and I'm only assuming mine is lower than most here given my nittier low variance style (I think?), but bad things can still easily happen. Course, I'm also guessing that my 955 bbs worst-ever-downswing is probably on the fairly small side for most players.

But I also realize a lot of this has to simply do with your skill advantage over the field you play in. The larger your advantage is (i.e. the better you are as an overall player and the more your opponents suck), the more you can probably do whatever you want and never really do that badly. But the more that gap in skill advantage decreases (i.e. the more average a player you are overall and the less your opponents flat out suck), the more you can probably fairly easily get in some meh downswings.

And of course, all of this over just a lol single lifetime sample size (which may be meaningless).

ETA: And I also don't want to just hilite the negative forms of variance. Remember, this is coming from a me, a guy who once went on a cute little 32-3 session streak, and at one time had a 23 month winning streak (think about the ridiculousness of that for a second given how few hours a once-a-week player gets in a month, noting that it would have been a 27 month streak had a flopped fullhouse held up). Pretty silly things can happen in this game.

GcluelessdownswingnoobG

Higher aggression (the word should probably be "better" aggression) reduces variance.

Mpethy (v/good poster/mod who did database analysis) posted on this a great deal in this thread and in chat.

Good aggression reduces variance. Higher aggression for the sake of being aggressive does not.

That's a strawman.

The fact remains. Two players, one tag and one lag, with the same win rates. The lag by definition has a higher VPIP/PFR/3b%. He has higher aggression.

And, perhaps surprisingly, lower variance.

Mike - it didn’t drive me back to a “regular” job. I planned on playing full-time for 1-2 years until I found the right opportunity and that’s exactly how long it lasted - 18 months.

As I’ve said numerous times, I never intended on making it a “career.” I am still playing as my primary source of income and it will likely be my primary source for another 12-18 months until I am more established in my new career.

Would it have been exponentially better if my stint as a full-time pro continued on the 45’ trajectory it was on? Absolutely. But I did what I had to do to preserve my bankroll, minimize the downside (while capping my upside) and continue to grind and earn. I’ve been able to supplement poker with some trading income in the market, and while it wasn’t a sexy 18 months it worked for me and with my bankroll intact and money socked away I’ll be ready to ride the next wave up if and when it ever decides to return.

Since I’ve started posting here there are very very few people, maybe no more than a handful, that have been a constant presence on the forum week in and week out consistently logging volume like I have. There is something to be said for the consistency required to keep playing week in and week out through good times and bad. Go back in 1-year increments and see the amount of people that seemingly disappear, likely due to bad streaks, BRM and not managing the variance.

Yah it didn’t go to plan but I’m still pretty damn proud of what I’ve been able to accomplish and the mental fortitude I’ve developed along the way. There’s no doubt in my mind I will be back to where I want to be eventually and I’ll be a stronger player for the experiences I’ve developed along the way.

I'm not convinced of this. It's possible your whole year might simply revolve around how you ran in big pots a lot more than you think.

For example, I had a horrendous year this year, only shipping about 17 BIs (my second worst rec year ever, slightly less than 3 bb/hr, assuming a BI is 100bb although technically I'm now BIing for less). In the recent thread regarding coping with bad beats I gave an example of how I was recently 1 outed in a set over set case; we "only" got in a little over a BI all-in on the flop, but with the dead money the swing overall for me was about ~2.5 BI, which is fairly significant given what I shipped overall this year. It doesn't take all that much "running bad" in your dozen big pots for the year. I started off the year with a thread regarding how I questioned whether playing with a maniac was worth it for a rec player given how the losses in a few hands could have a huge detrimental affect on overall winrate; if I recall, I believe I ran something like -3BI below expectation in just a few hands, or something like that, which again is a whopping percentage of what I ended up shipping for the year (and even what I would hope to ship for the year). I didn't do any all-in EV a la M / Mike Starr / etc., but I'm fairly convinced I ran pretty horrendous overall in the big pots I was in this year, which can have a pretty crushing affect on winrate over a lol rec yearly hours sample size.

But, admittedly, the more of a crusher you are with regards to the field then likely the less you'll feel this affect.

I think it was SABR who popped into the chat thread earlier this year and said his whole year was salvaged by a *single hand*. Course I'm guessing SABR plays in much bigger stacked games and can get himself into situations where there are stupid monstrous bbs on the line (which is unlikely in my much smaller stacked game).

Gmayyourunwellinbigpots,becauseitislikelytheonlyth ingthatmatters,imoG

Please don't take anything Ive said as an insult or any disrespect whatsoever. I have much respect for your journey.

Over the long term Yes. But the LAGs long term could easily be much longer than the TAGs long term making it seem like statements like "you cant play enough hours in a lifetime to get your true win rate" to seem to be true.

Assume a LAG plays for stacks 5 times per session
The TAG plays for stacks on avg once per session

They both get it in as a 70/30 favorite. If it takes 1000 all ins before they both reach the expected EV then obviously the LAG will get their 5 times earlier. But if they both run equally bad for 10 sessions, the LAG is going could experience a downswing 5 times deeper (even though theoretically he should come out of it sooner).

Also, the LAG is going to be getting it in with less of an edge than the TAG. The TAG may be getting it in with an avg of 70% equity and the LAG may only avg 50% equity or less since he is pot committing himself with bigger bets and has to call more all ins when behind which can easily cause even deeper downswings.

So over the long term, the LAGs variance is probably lower because hes playing more pots but when he runs bad he will get crushed. At least thats the way my non stat expert mind interprets it.

<Mod edit: see below for corrected math>

Note- I didn't attempt to randomize the 10% component, just assumed that Player 3's game "gets good" exactly 10% of the time.

I can't edit my last comment regarding running-good-in-big-hands from above, but I also don't want to just leave it hiliting the negative part of things. Last year when I shipped ~6.6 bbs/hr (at the time my second worst yearly result but also I thought perhaps a lot more in line with what was now possible in my changing poker environment) I mentioned how ~80% of my profits for the year came in just ~10% of my sessions (or something like that). I can't recall those sessions, but was a large part of those sessions simply running good (or not running bad) in large pots? Honestly wouldn't surprise me at all if that was the case.

GcluelessrunninggoodinbigpotsnoobG

Thomas Bayes would tell us that a player who won $75,000 over his last 4000 hours and thinks he's Player 3 above, is very likely incorrect and actually has a significantly lower expectation.

But the player in question might be right.

Sure, but if Bayes can *see* the players in the table splashing around like drunken donkeys, he might have a different opinion.

While it's hard to get statistics on a game, we can (sometimes) get an extremely fast feel for the kind of action we'll have at a table when we sit down.

Assuming this is true...what point does it prove?

MODS - Can somebody delete my poast (#20529), I fukked up the math.

CORRECTED THE MATH (HOPEFULLY)

Consider 3 sample players, all 3 are playing 2-5, with a long term WR = 8BB or $40 per hour.

Player 1 always sits in a game where his WR is $40/hr and his StnDev is 55BB per hour.

Player 2 always sits in a game where his WR is $40/hr and his StnDev is 75BB per hour.

Player 3 - 90% of the time he sits in a game where his WR is $30/hr and his Stn Dev is 75BB per hour, and 10% of the time he sits in an awesome high variance game where his WR is $130/hr and his Stn Dev is 150BB per hour.

So, all 3 players have an expectation of $160,000 after 4000 hours. What are their run-bad exposures over a 4000 hour sample?

PLAYER 1:
1 in 10 will win less than $137,500
1 in 100 will win less than $119,500
1 in 1000 will win less than $106,000

PLAYER 2:
1 in 10 will win less than $129,500
1 in 100 will win less than $105,000
1 in 1000 will win less than $87,000

PLAYER 3:
1 in 10 will win less than $125,500
1 in 100 will win less than $97,000
1 in 1000 will win less than $77,000

(I welcome any checking of the math, of course)

The poast where I screwed up showed that running well in big pots is super important, while the corrected version shows that the effect of big pots is a bit significant, but probably overestimated by most in the long term.

Not proving anything. I've seen claims about how important running good in key situations is important and how low variance play is so much greater than high variance play. I prefer to analyze claims like those by trying to put numbers to them. Thought I'd share what I found.

All in all, I mostly agree with squid and others who say ruminations about StnDev aren't very productive. But for everyone who thinks they are running bad, or trying to make life decisions regarding how poker fits into their income plans, I think trying to understand the variance component can be worthwhile. Also, knowing the numbers better helps me in limiting entitlement tilt or whatever else it's called these days.