Multi-tabling and variance

[DavidC]

What's the SD on a coinflip? (I guess 0.5)

How about on 100 coinflips (variance away from the mean of 50h/50t)?

I'm kinda curious about this, since it's a standard distribution, but it IS random.

--Dave.

Or for this, would you have to take like 1k samples of 100 flips to measure the sd/100 flips?

[DavidC]

[ QUOTE ]
It wouldn't theoretically reduce vairance - big downswings within a certain amount of hands are just as likely - the only thing that would be reduced is the length of time you were in the downswing - ie: if you are going to experience a downswing over 1000 hands, those 1000 hands will take less total of your time to get through with several tables -

the flip side of that is that the downswings can seem more sudden and therefore interperted as more severe if you are multitabling - multitabling you could go thru that entire 1000 hand downswing all in one afternoon -

[/ QUOTE ]

An afternoon, try a sitting! (2.5 hours 8-tabling)

[DavidC]

[ QUOTE ]
[ QUOTE ]
I think it is but I wanted to check with more experienced players.

[/ QUOTE ]

Don't multitable until you become yourself more experienced.
You will never improve your game by MTTing

[/ QUOTE ]

Agreed. [img]/images/graemlins/frown.gif[/img]

[Dave G.]

[ QUOTE ]
Variance per hour increases when multi-tabling, but variance per hand does not.

[/ QUOTE ]

Yes, I believe I covered this when I stated that your variance on any one table does not change.

When people speak of variance like this, they mean bankroll variance. It is very clear to see that your bankroll is going to vary in direct proportion to how many tables you are playing.

[xenthebrain]

Just wanna throw it in.
If you multitable (as winning player) and you play the same at each table like before you don't have a higher risk of ruin with your bankroll i.e. you don't need a larger one.

[turaho]

Maybe I misunderstand the concept of variance, but I thought it is how we describe fluctuations in "luck" over a particular sample size.

Let's say for the sake of argument that a player who plays 100% prefect poker all the time can expect to make 3BB/hr at a micro table. If that player can continue to play perfect poker at 4 tables simultaneously, their expected win rate should not change.

But since there's an element of luck in poker, you are not guaranteed to win 3BB/hr every hour even if you play perfect poker. Draws won't come in, opponents suck out, and so on. Actually some hours you'll win 1BB, some you'll win 15BB, and some you'll lose 20BB. But in the grand scheme of things, we can average that out to determine our win rate.

To use a metaphor, if I flip a coin, I expect heads to come up 50% of the time. If heads only comes up 40% of the time in 100 throws, that's variance in action. For the short-term session of 100 coinflips you're coming in below the expected win rate, but over the long term the ratio of heads-to-tails results will be very close to 1:1.

If that's so, then how would playing multiple tables affect those random occurences? That's like saying if I flip 4 coins at the same time, I expect that ratio to become worse. If anything, multitabling should reduce the appearance of variance because it generates a larger sample size, and the larger the sample size, the better the chance that results will approach the mean.

Of course, variance also means you can run cold for 10k hands, so when you're losing a lot it seems really big because it happens so fast. But I say that as long as playing multiple tables has no effect on your poker playing skills, then it should have no effect on your variance.

Perhaps the question should be, "Does multitabling affect your poker playing ability?" The answer being the old standby, "It depends." For some players, following the action at mutliple tables reduces their ability to get reads on their opponents or prevents them from making sound judgements based on calculating pot odds. For others it helps because it keeps them from entering pots that they shouldn't be because they're bored and crave action.