actually there was a decent amount of playing at the beginning of that garf, it was just in such a narrow range of +/-



that spike was basically cashing in on sitting through the above, and then back to BE mode

I finally got around to trying the oldest trick in the book: reboot. That seemed to clear up the problem.

GwhoopG

I'm a rec player but I try not to be a huge fish. I read books, listen to podcasts, watch some videos but I have a full time job and can't exactly dedicate my life to getting better at poker. I play what is probably a too nitty, low variance style and I am sure that I don't get all the value I should be getting out of my good hands. I think I should probably be losing more sessions, but winning more overall. Interested to hear any thoughts on that.

Started tracking in 6/2015 so this is about two years. I was between jobs for a few months at the end of last year / beginning of this year and got to play more and had slightly better results in that stretch. Graphs filtered to NLHE (and it's pretty much all 1/2), because I play a 2/4-3/6 FL game with a bunch of old guys a couple times a month and trying not to muck up the stats.

Congrats dude.

I've had a higher bb/100 winrate in 2/5 than 1/2 in basically any year. To go to a casino, see no 2/5 or PLO and then play a 1/2 game where you buy in for $400 and the rest of the table has $600 combined usually means I go home. Even the worst $2/5 games have enough money on the table to make money just winning blinds.

I cross-posted this from another thread to prevent further derail.

Common sense and math. There is an absolute limit to how much variance a game can have, and this correlates positively with the stack sizes. Smaller and smaller stacks definitely don't lead to higher and higher variance. For a silly example, a .2BB stack game obviously has lower variance than a 200BB stack game. How can a shorter game be lower variance when there is literally a limit to the amount of "variance" you can experience in an individual hand? In a 200BB game compared to a 100BB game, you are still winning/losing 100BB hands about as often as you are winning/losing them in a 100BB game, you are just not all-in and have the potential to win/lose even more, which leads to higher variance. There are a lot of times where the stack size limits your win or loss in a game, which directly limits your variance.

There are times where someone might just take a 50BB loss in a 200BB max game where they might commit a stack instead in a 100BB max game, but I don't think this is enough to counteract other effects.

I'd love to look at such a database, but I don't know where they exist. My own online database isn't very big and has hardly any short-stacked hands.

Also, we're talking about short vs. deep games, not big vs. small games. I definitely agree that variance is higher in smaller games in terms of BBs.

Mathematically and logically it makes no sense to me that shorter games would have higher variance.

No, i believe its correct. Im pretty sure its fairly well documented that good shortstackers had much higher variance than good deepstack lags. Mostly bc of your third paragraph, which im too lazy to try to quote now that ive already started typing quick reply.

The term variance is always a bit confusing to me as it relates to poker. But i believe this is true bc shorter games rely on showdown winnings more and deeper games rely on non showdown winnings more and non showdown smooths out your wr more.

Could be off on that last bit.

In shortstack games, a lotta time you're doing << 60/40 preflop flips with some dead money, so pretty high variance.

GIthink?G

This discussion made me curious so I decided to go and check my online MTT database(I never play shortstacked in online cash so don't have a sample) to find out what the relation between stack size and variance looks like. This is what it looks like in graph form:



Total sample is about 175k hands, only the 250bb and up stack sizes don't have a significant sample size. Now obviously online MTTs and live cash are quite different but I am very doubtful that shortstacking increases variance because the fact that you can win/lose a lot more bbs when your stack size is larger matters a lot. Y axis is std.dev per 100 hands. X-axis is stack size from high to low

I think the idea that shortstacking increases variance comes from winrates being higher when deep, and higher winrates result in a lower possibility of losing. The variance probably also feels higher because you are all in more often when short.

Thanks for sharing! I agree with the last paragraph.

Here is the image which didn't show up properly.



It's still really hard to see the axes labels, though. It's here http://imgur.com/2yZYddF

The middle part is correct obviously. If I buy in at a deep home game that has no max, I'm maybe all in once or twice all night, if that.
Now, if we're playing a shorter game, our profits are much more reliant on showdowns, and specifically all in showdowns.

My hours are mostly PLO and the difference in variance isn't just the coming together of hand equities, it's that most hands reach showdown. It's a game reliant on making the best hand. When you're shorter stacked in holdem, you're reliant on the same thing as well.

Online MTTs are completely night and day compared to cash wrt variance dependent on current stack size. The difference is bigger than between PLO and NLHE.

The swinginess of a game is going to be contingent on how the players play the game. In online poker players are generally quite willing to get it all in preflop with short stacks. More showdowns = more swings. It sounds like the low stakes in Cali where everyone is short may be like that too. However, my experience playing short in normal live low stakes games is that you rarely get in flip situations pre.

Variance is defined as the expected value of the squared deviation from the mean.

For example, whenever you fold your variance for that decision is 0, as your EV at that decision point is 0 so there is no deviation from the mean (although your variance for the hand from the start is not 0). If you make a 0 EV 200BB river bluff shove into a 100BB pot, the variance is (200-0)^2*1/3+(100-0)^2*2/3 = 20000.

In deeper stacked games greater deviations from the mean occur which are not even possible in shorter games, which is why it does not make sense that shorter games would have higher variance.

Standard deviation is the square root of variance. We typically prefer to use SD for any of the applications related to poker.

I did just look at my own online database, and it is too small to draw any reasonable conclusions, unfortunately. There doesn't seem to be any correlation at all between stack depth and variance in my tiny 5000 hand sample.

At least it's evidence relevant to the discussion. It would be cool if you could provide some.

Right.

The problem is that when many players use the term "variance" they're talking about the magnitude of their swingyness, not the deviation from the mean. We have this discussion periodically ITT and I always feel like I learn a little more about the math when we do.

So in that context a short stack game has smaller swings because of the smaller pots.

All that runs in my room is 1/2, so heres my mid year results so far for 2017.

Hours: 345
Profit: $8710
Profit/hr: $25.25

The last 140 hrs have been break even but I might be running good overall. Too lazy to post a graph.

Overall sample is 630 hrs @ $18.5/hr, but been feeling good about some progress lately, so I'm hoping my 2017 results are indicative of that.

Looking to start making the drive to the casino often and actually start playing live. Ibe been around poker a while but feel like beating 1/2 at a casino on a Friday is a lot different to poker online which is where I've mostly played here and there over the years. Would any regs give advice on maybe some books aimed at this specifically, or maybe a video series or youtuber or podcast or whatever for just general strategy aimed at low stakes lI've NL? I think the last time I read a poker book it was by Phil Gordon if that gives you some idea.

Not sure why you posted that in the Winrate thread, but check out the stickies ITF. There are several good threads on online-to-live transfer.

With all due respect I am familiar with the stickied threads, and many are 5+ years old, some posted before I joined in 2012. Not saying they don't translate but I'm just looking for material I haven't already read. I looked at some other threads and they appeared pretty dead and thought this was a good place as it is actually frequented by regulars that play live and manage bankrolls.

Would love your input on other resources or publications I should look into Garick that have helped your live game.

Please try the Chat thread and/or the Books and Publications forum. This thread is for Bankroll Winrate and Finance discussion only.

I will say, though, that low-stakes live really hasn't changed much since 2012. I wouldn't worry much about resources being "outdated" at 2/5NL or lower unless they are pre-poker boom.

Well, I'd say you're better off throwing out a lot of the advice in LLSNL that something is standard/good because that's just how it is in low stakes, without any evidence/math or even an argument why.

It's changed in the sense that blindly following a list of idioms was enough to be a winner in 2012 but isn't in 2017 and worse, will ensure you are totally unable to learn and improve.

So... if you're playing 30 hands an hour...

($25.25/30)*100=$84 per 100 hands

Meaning at 1/2 you are earning 42bb/100hands.

That's a ridiculous winrate. Well done.

It's only ridiculous because you are comparing it to games that aren't comparable. Also, it's a minuscule sample size.

I would have to consult Excel but I think it was roughly 500 hours @ 1/2, 100 hours @ 1/3, 350 hours @ 2/5 and 150 hours @ 5/10.

I had a horrible first month as a pro (March '16) ending up -12k after the doom switch was activated and going on a -15k swing. I've had one other losing month (November '16) while going through another nasty downswing in October/November.

My mentality has certainly done a complete 180' though from playing recreationally vs. professionally. It's hard/impossible to explain in my opinion - it's just one of those things one has to experience them self if they decides to go pro. There is a bit of a learning curve but I've settled nicely into the grinder mentality. No intention of doing this all that much longer but at least I know I can always do it if need be.

SOOOOO much this. Cant be explained, as much as i try.

Not sure if this is exactly the right spot to post this but I think it's right...

Background: I've been playing 1/2 casino poker for about 7 years-- never full time and probably averaging one 5 hour session a week, with some months long breaks and some periods where I've been able to play 3 nights a week. I'm a winning player at what I estimate is about $15/hr. Because I'm not playing for a living, when my bankroll has shot up above 20 buy-ins, I've used it for personal expenditures.

Based on what I see in others' experiences, I am becoming convinced that I am sacrificing +EV for lower variance. While I've won on average $15/hr, I would estimate that I win ~85% of my sessions, but have won > 2.5 buy-ins in a session just 2-3 times over those 7 years, which seems extremely low to me. I definitely acknolwedge that I have "accountant's syndrome" in that after I get up a certain amount, I tend to play tighter and am more apt to leave, but I also think my style of play may be lower variance at a cost of a lower win rate.

Question: I've identified that light 3! is currently completely absent from my game. Given my background and description, what other leaks might you expect a player like me has?

Also,with my bankroll solid, I want to start adding 3! light to my game and based on this forum it seems the only consensus recommendations are 3! light with A-rag suited from late position or the blinds against Vs who have big enough stacks and have a fold button (also preferably with dead money to gain in the pot). Any other situations/cards where folks are finding 3! profitable, even if higher variance?

don't do any of this at 1/2 imo