Why You Shouldnt Play HUSNG with 20 Buyins

[cwar]

[ QUOTE ]
you should not be recommending that those people play with 20 buyins.

[/ QUOTE ]
Did you read the title of the article? [img]/images/graemlins/grin.gif[/img]

[TNixon]

er, yeah. duh. lol

[BarryLyndon]

[ QUOTE ]
only 20 buy-ins? hell i should be playing the 10Ks then

(thinly veiled brag post obv)

Indy

[/ QUOTE ]

Really not that thin [img]/images/graemlins/wink.gif[/img]

I play HUSNGs with much more than 20 buyins because I play other games; also, HUSNGs, of all the poker games, have the greatest potential to get under my skin. When you run bad, you are going to have to deal with being DOWN 2:1 a lot of the time. It's totally unpleasant, but if you stick with it, you can learn a lot about how to play the "short stack." Still unpleasant though. Climbing from 1K chips to 3K chips at 20/40 is a long road.

You will have to make a lot of tough decisions as to whether to push/raise/fold at awkward stacks - say you're stack is 740 and the blinds or 40/80. Totally different from cash whereas you can simply reload. Unpleasant decisions lie ahead.

HUSNGs are intensely psychological. Try losing three or four of them in a row and then going up against ONE person where you have to make a decision on every hand to win or lose a blind. And then you are down 2:1 again. Or up to 2:1 but then lose all-in preflop AJ v. A9. And now you are down 2:1.

I'm a winner in HUSNGs, but I've stopped playing for a while just because I think I lost like 6 out of my last 10 or something and was just tired of them. Not on a tilt or anything, just needed to move on to another game. Too many awkward stack situations and not enough players to protect me from being impatient.

So, if you are going to focus on HUSNGs, proceed with caution. Of all the forms of poker I've played, it puts you the closest to both your opponent and your own psychological underpinnings.

Barry

[LordMushroom2]

Someone asked if your style of play (LAG vs. TAG) affects your variance in HUSnGs. And the answer is that it doesn´t. Your winrate is the only thing that matters on variance in HUSnGs. One style may yield a higher winrate than the other, though.

I have looked a little into this Kelly criterion and it is interesting, but I have some questions about it.

1) Winrates are not constant in HUSnGs. They increase if your bankroll decreases (you are forced to play lower and easier levels). And they decrease if your bankroll increases (you move up to higher and harder levels, until you reach your most profitable level and then you won´t move further up no matter how big your bankroll becomes).

It seems to me the Kelly criterion is not designed to take that into account, and that it is thus recommending a too aggressive bankroll-strategy.

What do you think?

2) If the Kelly criterion recommends a too aggressive bankroll-strategy, can anyone make a modified formula to fix that?

3) If the Kelly criterion doesn´t recommend a too aggressive bankroll-strategy, can anyone make a formula where the uncertainty of the winrate is included?

(I mean include the standard deviation and the number of games played at that particular level so you can safely use a full criterion which takes uncertainty of winrate into consideration instead of resorting to the half-assed half criterion)

[daveT]

TNixon:

I have also been using the 10 bi rule over the past 300+ games. Pretty much every time, I move up, I get smacked. Take last week, I was down to $10, ran up to ~250, then ate a 16 bi downer, so back to playing the fives again.

In a previous post about br management (one of like 50, right?), you stated that you think it is abnormal for a player to experience multiple 7+ bi downswings. Is there anyway that you can show in your simulation how many downswings a person should experience in x many games. Is it normal to run

wwwwwwwwwwllllllllllllllllllwwwwwwwwlllllllllwlwlw lwwwwlwww

while having a 58% wr? I know the above looks silly, but that is honestly how I run. If it's psychology based, I must be the worlds best poker player, and I know that's not the truth.

[vodoo]

thats quite a good post, nh cwar.

Im playing with 100 buy-ins, but im really a br nit.

[soop]

You got me thinking with that post Dave. I bet that's the norm and the swings in HU poker (due to psychology/tilt) are bigger than they would be if you were just flipping coins.

Let's do a little experiment. Find your win rate in games where you lost the previous match and compare it to your winrate when you won the previous match. (use every match - no selection bias). If you're a robot, they should be the same over a long sample. But of course, my bet is that most people have better win rates after wins and are worse after losses.

If we all do this, we can can sort of quantify tilt, and how much you are likely to experience (even if you don't think you ever tilt).

I'm gonna look this up and post it when I get home.

[TNixon]

Well, unfortunately, wwwwwwlllll is just as likely wlwlwlwlwlw.

It sucks that the losses are coming at just the wrong times, but if you're analyzing your play, and taking into account the fact that the players are going to be at least slightly better at the higher level (although to be honest, there's probably not a huge difference between your average $5 player and your average $10 player), then it's just a matter of time. 58% is more than enough that you'll get through the next level at some point.

One thing I do know from personal experience is that it's *very* difficult to still be playing good poker after a string of 5 or 6 losses. Even if you don't feel like you're playing poorly, you might be adding a little extra oomph to the downswings.

I was having a similar experience moving from $30s to $50s trying to play 10 buyins. I'd get up to the fifties, lose a couple in a row, and have to move back down. Last night I finally broke solidly through. After winning 7 $30s in a row, I had 14 or 15 buyins for $50s, and lost 2 of my first 3 $50s, but had enough to keep playing, and ended up at 60% in the $50s for the night, over 15 or 16 games or something.

Whatever bankroll rule you're using, there is a very real statistical advantage to leaving yourself enough room to play a few games at the higher level before moving back down. So if you're moving up at 10, move back down at 7. Or if you move up at 15, don't move back down until you hit 10.

[cwar]

[ QUOTE ]
you stated that you think it is abnormal for a player to experience multiple 7+ bi downswings

[/ QUOTE ]
This happens to me multiple times in a week, I wouldnt call it rare at all.

[daveT]

[ QUOTE ]

Let's do a little experiment. Find your win rate in games where you lost the previous match and compare it to your winrate when you won the previous match. (use every match - no selection bias). If you're a robot, they should be the same over a long sample. But of course, my bet is that most people have better win rates after wins and are worse after losses.



[/ QUOTE ]

This is the first 200 or so games for me, this is when I first started playing HUSNGs, so perhaps these numbers don't reflect what I should be doing now.

lwlwlwwllwwwwwwwwlwwwlwlwwwwllwwwwlwwlwwlllwwlll
wwwlwllwwwwlllwwlwwwllwlwwllwlwlwwlwwwwllwwwwlwl
wlllwwwlwlllwllwllwwlllllwwlwwwllwlwllllwwwwwwwwl
wlwwllwlllwllwlwwlwlwlwlwlllwwwwllwlllwlwlwwwwlwl
lllwwlwllllllwwlwllllwwlllll

Here are the last 100 games or so.

wlwllwwwlwwwwwwlwwwllwwwwlllwwlllllwlwwl
wlllllwllllllwwlwlwllwwwllwwwwwwwllwwlll
llwwwlwwlwwllllwwlwllllwwlwl

the first sample:

w:115
l:108
Sequences:
ww: 58
ll: 48
lw: 55
wl: 56

Sample 2:
w:54
l: 54
Sequence:
ww: 30
ll: 30
wl: 24
lw: 22

Incredibly, my career has been bookended by two break-even stretches. Go figure. The bright side is that it makes the math easier.