Friday, August 8, 2008

The Origo-based Win-rate Fallacy

There's a silent misconception that I see implied in poker posts here and there around the intertubes. It's very subtle, but I definitely think it's there. The reasoning goes something like this:

"I'm a 4 ptbb/100 winning player while playing 4 tables. If I try to play 8 tables instead, my win-rate will probably go down a little, but assuming it stays more than half what it is now, I'm still making more money."

Do you see the danger in that paragraph? Like I said, it's subtle. Arguably semantic. But I think it's more than just semantics.

The potential fallacy lies in the word "half." Many people argue over their win-rates as if the lower boundary of their win-rate is zero, implying that if they play half as well, they win half as much. When I put it like this, it's fairly obvious that this isn't the case; if I play half as well as I do now (make twice as many mistakes or make twice as costly ones) my win-rate isn't going to be half of what it is now. It's very likely to be negative.

So while adding another table may indeed be profitable, there's certainly nothing guaranteeing this just because you're winning at the number of tables you're playing now. I make this post to serve as a reminder to all (three) of my readers that our edge in these games is much, much smaller than we think. If we compromise that edge in one way or another, we may no longer be winning players at all.

It doesn't take much for a winning 4-tabling player to become a losing 8-tabler ("tabler?" is that a word? really? Ok.) All it may take is an extra distraction here and there, because our edge isn't that big. That's not to say that we shouldn't play more tables, it's just that we must not dismiss the effect of those extra dollars that we miss here and there. They may constitute our entire win-rate, and more.


BelgoSuisse said...

I'm pretty sure you have a lot more than three readers...

A very good point you make, as always. Although with myself as a guinea pig, I also notice the opposite: my win rate sometimes goes down if I play too few tables, mostly because I get bored by the lack of action and end up playing too loose. In the end, I think i'm playing my best poker when I play between 300 and 400 hands/hour, which is between 4 and 6 tables depending on FR or 6max.

Fredrik Paulsson: said...

I think this might hold true for a lot of players, actually. People get bored or start to overthink things or overestimate their own greatness and think they can make great bluffs and huge laydowns. Very often, just playing ABC poker would have been better.

So yes, there's definitely the case that playing more tables can increase not only your overall profitability but also your win-rate. :)

Anonymous said...

Well, count me as reader number 2.

And like Belgo, it depends on the situation for me. I need to multitable when I'm working at my short stack game because I have to play so tight and boredom can cause me to play too many hands. But if I multitable too much with PLO I'll get myself in trouble because I'm still learning the game.

And then if it's a CC game or tourney I won't multitable just because there's so much chatting to pay attention to :)