Wednesday, January 14, 2009

More winners and fewer losers in PLO?

I've always been interested in the maths of poker. Short term variance can be very misleading when we rely on experiential learning. Plays that come off can seem right and those that don't we can shy away from. Maths can help us gain a more profitable understanding of the game quicker.

One stat that I read early and have since observed to be very true is that only 5% of poker players are long term winners, the other 95% feeding the economy. In my early Poker Office days I would scroll down the list of observed players overall results and eventually, sure enough, about 5% would play big volume as winners and there'd be a long tail of small users.

I also know that in NLH we can be presented with opportunities that represent large edges pre and post flop, maybe 80/20. In Omaha this is more rare, the norm being closer to 60/40 as an edge that you want to push. So what effect does this have on the distribution of winners and losers?

After a very small sample size, just over 5000 hands observed, the winner/loser split in my PokerTracker Omaha is exactly 40.06% and 59.94%. If anyone has a larger sample size, perhaps you can share with me if this phenomenon holds true. If so, inferior players can be fooled more by short term variance into thinking their strategy is good. Just look at my recent premature entries.

In terms of my recent PLO results, they have slowed down to 9BB/PTBB. Last nights session was very LAGGY and the variance massive. I dropped 2 BIs early and recovered by the end of the session.

4 comments:

James P McAteer said...

As I write this its still last night in the UK which feels very weird. I wonder how big an edge a mathematical perspective gives you? Apparently Mozart was fantastic with numbers yet was obviously very creative.

Not sure what that last comment has to do with anything...

TiocfaidhArLa said...

IMHO, maths plays much more of a role online as betting patterns are your biggest guide. Understanding the player tendencies is easy with tracking software and calculating outs and pot odds is primary school stuff.

More nuanced maths is for the professionals where the gap between playing standards gets much less as a percentage. Even then, rules of thumb arise when the hard work has been done, eg bankroll mgmt etc.

parttimebonuschaser said...

In your original post my interpretation is that you're saying that since the edge is smaller in omaha (60/40 versus 80/20 preflop) there are potentially more winners in the game.

Over the short term this may be the case, but as more of the players get better and they keep pushing these small edges harder, your percentage of winners is going to drop, as the sharks will be feeding on the fish.

I guess for example in fixed limit the edges of any random cards versus No limit are smaller because of the fixed nature of betting allowing more draws to be correctly drawn to. However, there are still probably only 5% of players that are winning players all up in fixed limit.

Another factor you could consider might be that if players can't have much of an edge over other players, then rake will make proportionally more players losers as they are only marginal winners.

Food for thought.

TiocfaidhArLa said...

You're spot on.

I found the 60/40 split more coincidental and as you said probably symptomatic of the small sample size where variance fools us all.

Few people play chess for money because the best player will almost always win. Backgammon, the dice gives amateurs a fighting chance but not over many games.

Hold'em is perfect because it gives the perception of being able to beat Ivey if you get lucky in a single short session.

Omaha I am hoping will be an extreme version of this where it looks like we are coin flipping all the time but the better players will be on the right side of the crooked coin.

I'll be interested to observe this divergence and the playing styles that ultimately win.