this is simple utility theory. assumes your risk valuation is not linear, and you are more affraid of a loss of $2 vs a gain of $2. so even if your expected gain is above $0, you are not willing to get into the game.

i thinks it's fine. (not sure if you should be playing 2nl then).

I'd be checking his "ELR" in PokerTracker first, but not if it's full moon obviously.

any results available yet? did he have JJ?

You might as well play roulette.

fixed your post.


Also, +1 to everyone letting Giblet make actual posts instead of dogging on him, although I just did too. I hope it's no pin in your cushion, Giblet.


As for an actual reply, Mr. Giblet, where you said you were due to get lucky from your string of back luck, I believe that view is called "Gambler's Fallacy" and it has to do with the past results not affecting the current/future results.

I could flip a coin 10 times in a row and have it land on heads. Although I know that over an infinite number of theoretical trials, it will all even out, the next flip still only has a 50% chance of landing on tails. Don't let the thought of "I'm due to get lucky" affect your judgment. Try to keep playing by the numbers and making good decisions and the luck will work itself out.

I see your point, but you won't get far in poker with that attitude. It's all about the +EV. If you don't have roll enough to flip for 2$ you should really rather do something else.

No. EV and variance are two statistical terms that don't correlate at all. You should only have to care about EV and not think about variance though, that's what bankroll management is for.

lol

You may be right, but it sounds wrong to me.

Suppose I'm a professional backer of horses. I have the skill to keep finding two kinds of horses: those on offer at 20-1 which should really be 16-1, and those on offer at 5-2 which should really be 2-1.

Every $ invested on the first group of horses returns $1.24, while every $ invested on the second group returns $1.17.

But I get fed up with the mind-bending losing sequences from the first group, and decide to stop backing those horses, and concentrate on the shorter priced group.

My winrate takes a hit, but my variance is greatly reduced, surely?

Think of it this way.

Excluding nosebleed players.

Pull out a graph of a 1ptbb/100 winner vs the 5ptbb/100 winner.

What you will find is that the closer the winrate is to 0 the greater the swings will be in general.

Now you may choose to look at the short-term variance and think ok by passing on these equity 'flips' I am reducing my variance....but what you are actually doing is making a likely +EV decision into a 0EV decision.
The more often you do this the lesser a proportion +EV decision will be amongst all your decisions.
e.g.
25% of your decisions are +EV
50% of your decisions are 0EV
25% of your decisions are -EV

Assume that all decision are of = value in units
And of your 25% of +EV decisions 20% are AK all in PF vs an Unknown.

The net effect is now your distribution of
decisions is

20% +EV
55% 0EV
25% -EV

So your +EV value decisions went from being 1 in every 4 being correct to now being 1-5 being correct.

Ok.

So your focus should be on reducing the number of mistakes you make first and because all decisions are not of = value - then to focus on biggest pots won or lost - if your goal is to reduce variance/increase winrate.

First, I don't play 2nl.

And, second, if someone is that risk averse, they shouldn't be playing poker in the first place, which is what I was getting at.

Actually, that's not even a scenario worth considering if you are in this for the long run.

How does this thread have 62 replies? This has been done 100's of times.

i must be playin at the wrong tables because i'd say its KK or AA 95% of the time............ YES........ AT 2NL (at least on FT)

sure some dummys will try it with mid pocket pairs, or even qq whatever... But I dont think its as often as everyone says.

I'd need him to "exhibit evidence of lunacy" before calling AK... Sometimes I wonder how long its been since people have played 2NL that give advice..

Since I saw this thread I have paid special attention to this. The last 5 times I have seen it and have gotten to see the hole cards, various Villains have had 87s, A6o, 94s, AQo and QQ. (Cake 10NL 6max)

You know, I've since thought about it, and I can't assume I see too many shoves like that while > I < have AK... Mainly because I have an A and a K - someone have KK is unlikely, and AK well... AA unlikely but still possible..

Anyway... I think im going to experiment with this and just see what happens maybe the next ten times i see myself in that situation.. Assuming I don't climb out of 2 NL again anyway.. Maybe Im just remembering the bad

I don't call it off, but I'll shove myself.

Does that make sense?

1ptbb/100 players aren't generally playing low-variance hold'em compared to 5ptbb/100 players. They're just playing worse hold'em.

DiamondDog's sports betting comparison is a good one. In poker, you're more likely to lose multiple times in a row when you're a 52% favourite than when you're a 60% favourite; even if by passing up on those 52% spots you're decreasing your total +EV, you're also decreasing the likelihood of losing ten in a row. You're not taking into account the fact that by playing a low-variance style, although less of your decisions are +EV and more of your decisions are 0EV, your +EV decisions are more +EV and your -EV decisions are less -EV (if you do it right). This reduces the likelihood and size of your swings, at the expense of some of your total winrate over a large sample.

Compare it to early-stage STT or MTT play. Your objective is to play tight and minimize your chance of busting out, and to accumulate chips when solid, majorly +EV opportunities present themselves. Often, decisions that are +cEV (i.e. +EV in a cash game) should be passed up on; calling a shove with AK in the fifth hand of a level 1 is a bad play here, even if you'll win 52% of the time. You try to keep pots small, play cautiously, avoid risky bluffs and only play solid hands, all with the objective of increasing consistency and survival at the expense of always obtaining maximum Sklansky bucks. Now, the reasons for folding are different ($EV versus minimizing swongs), but the principle is the same; reduce variance.

Another example would be if you folded every hand except AA, which you open shoved. Now, your winrate would be poor (well, negative), but your graph would be very consistent, a bit like a small-scale version of an MTT grinder's graph. It'd fall slowly, before either spiking upwards for 1.5BB most of the time, 1BI some of the time, 2-3BI every now and again and downwards for 1BI a small percentage of the time. Over a large sample, your variance will be practically zero because you're only taking spots where you're a massive +EV favourite and passing up spots where you're less +EV. That's obviously an extreme illustration, but you see what I'm getting at here.

This

Here is yet another counterexample to your assertion that reducing win rate necessarily increases variance.

My personal win rate in 9-handed NLHE online is 4.55 PTBB/100 with a standard deviation of 34.22 PTBB over 100 hands, which corresponds to a variance of 1171 PTBB^2/100. I'm not claiming that's a wonderful WR, but it very definitely is positive.

Suppose I were to change strategies to one of always folding. Yes, I would sit down and put real money on the table and play my blinds, but I would always fold. Let's assume the other players figure this out real quick and always steal my big blind.

My win rate using this strategy is going to be -0.75 PTBB/9 hands, or -8.33 PTBB/100. This is definitely less than my current win rate. My variance is going to be (-8.33)^2 - (((-.5)^2 + (-0.25)^2)*100/9) = 65.97 PTBB^2/100.

Not only would my win rate be lower than my current one using this strategy, but my variance is stunningly lower.

Do you know anything about mathematical logic, DiggertheDog? All it takes to falsify a flat assertion is to find a single counterexample, no matter how far-fetched or pathological. I have described a change of strategy that would, without a shadow of doubt, reduce both my NLHE win rate and my variance.

Therefore, your claim that reducing win rate necessarily increases variance is, without a shadow of doubt, absolutely, utterly and completely false.

Go away, and don't come back until you've learned some basic probability and statistics.

alan bostik -

theres two definitions of varience, mathamatical, and poker player

digger is talking about the poker player definition, which is less downswings, which is all you really care about

Jeez, wake up on the wrong side of bed today?

For me its all depending on the Villain and his history. If i have seen him play really bad i might call AI with AK, if not i will fold.

Againts a resonable range we are at best 50 % so why call when we at best win 50 % of the time, and in many cases we are a 1 to five dog ?

Also i think if people look at it like its only 2 $ is fundementally the wrong way to approch it. True its only 2 $ but the goal of the game is not to win $ here and now, but to have as high bb/100 as posibel. Wich a 50/50 situation does not help at all.

In cash games its much better to wait for a better spot, because it will always come if you are patient. You dont have that luxury in tournaments.

But ofcourse if we have seen villain play AI with crap before its and easy call.

wat

Variance is only a mathematical term, that poker players use. Read AlanBostick's post, sums up the discussion imo.

yea dont listen to this.

Call this everytime at this limit.

Why the hostility at the end?
Ok I concede the point.

But OP - do not worry about variance @ micro cash games - focus on correct decision making.
The more correct decisions you make the more likely you will have a larger win-rate and the effect of downswings in your poker life will most likely be ameliorated.

Ok I will go away now.

variance as a mathematical term isn't whats commonly used on 2p2

we use varience to mean downswings, its a slang term

Nearly all of my poker has been cash games - never played any kind of MTT, dabbled in some SNGs - but I'm pretty sure this is the complete opposite of how we should be thinking.

In cash games, assuming we're properly bankrolled and can simply buy another stack of chips if we lose our current one, there's no such thing as a 'better spot'. A spot is either +EV or it isn't. If it is, we should be taking advantage. Yes, the swongs are going to take some dealing with, but if it's +EV then walking away from the spot is costing us $.

In tournaments (as I understand things) we should be playing tight early on and loosen up later but we're still looking for +EV situations. The only difference is, cEV and $EV aren't always the same thing (the way they are in cash games) so stuff like ICM comes into play. Spots where we'd snap shove in cash can become folds, spots that are super easy folds in cash games become insta-shoves in SNGs etc.

(In cash games, Chip Reese, in SuperSystem (1 or 2? can't remember) talks about one special spot where it would have been the correct thing to do to turn down a +EV spot in a cash game but that was when he was starting out, couldn't reload if he went broke, and found himself playing against some drunk lagtard. Reese got it in as a slight fav, realising later that if he'd been more patient it was almost guaranteed that he could get it in later as a huge fav, given how the guy was playing. But that's a special case. With a bigger bankroll getting it in as any type of fav is the thing to do in cash games.)

Apologies to MrGiblet for somewhat derailing his thread.

i'd snapcall and hopefuly the fulltilt TIMER clicker shuffler thing wil flop me a fullhouse Aces fulla KINGS