I have been looking at the hand database from PokerFTP to find the optimal VPIP level. I was expecting the best VPIP to be somewhere between 15 and 25. However, there is no significant difference between the average BB100 of players with VPIPs from 10 to 30.

On average, players lose a little, but that is to be expected due to the rake.

The only pattern I can see is that the variance increases as players get looser. That is probably because they see more hands and have more chances to win or lose lots.

I've put my results into the linked chart.

The x axis is the VPIP, the y axis is the BB100 for players with that VPIP.

The blue line is the mean, and the green and red lines are the mean with two standard deviations added/subtracted.

The dip for VPIP 4 is just due to a very small sample size The same can be said for the eratic results as the VPIP gets above 40.

Do other peoples' Poker Tracker results concur or disagree with my findings?

it's all dependent on table conditions. if it's not between 18-30 longterm, you're going to have trouble beating SSNL, though.

I think your sample size is a big problem here. 1000 hands, even when its the bottom of the range, is far too small a sample for players. I don't know how big your db is, but you need to expand that part of the graph upwards, to at least 50k.

There is almost no winning players playing a 30 vpip game in most SSNL games i've played, in the past 6 months i've only seen one player i would consider a "good" lag, and he's a former 3/6 player. Most are breakeven or losers according to PTR.

Also, the laggier players are experiencing more variance because their winrates are lower, not because they're playing more hands. This is a common misconception. VPIP/PFR are also not really that closely linked to how good someone is except at the extremes of each stat.

Whenever i play 21+VP i lose whenever i play 16 i print money. True story.

fyp

I've been actually wondering this for quite some time now, are some of the 16/13 etc SSNL regulars really winning any significant amounts? At least most of them are breakeven (at best) rakeback pros.

Hello,
Thanks to everyone who's replied to this thread. It's a bank holiday in the UK today, so I've taken some time to do a deeper analysis of VPIP, based on your comments.

Unfortunately my database is too small to contain many players with 50k hands. However, for the purposes of this analysis, 1000 hands is not too small.

If we are analysing a large group of players with 1k hands, then while their individual BB100 scores may be less reliable than if we had 50k of hands for each player, the average BB100 for the group will still be the same.

Also, if we only include playeres with 50k hands, then we add survivorship bias to the results, as we'd only be looking at serious players. Assuming that players average 60 hands an hour, and can play 3 hours night for 5 days a week, it would take them about 56 weeks to play 50k hands. That's about a year. Or if they two table, that's half a year.

I've looked at the distribution of players' BB100 statistics for different VPIPs. And unfortunately the results do not match your statement. Below are two charts that show the distribution of players by BB100. One is for players with VPIP from 20 to 25. The other is for players with VPIP from 40-45. Both charts have distributions that are close to the normal distribution
(http://www.statsdirect.com/help/dist...stribution.htm). Both have a similar mean, though it can be seen the looser chart contains a higher proportion of players at more extreme BB100 values. And it is this that causes the higher variance for looser players, not a lower winrate.




These stats that I've been looking at seem to show that trying to get your vpip into a certain range is not really going to help your play. A loose player is just as likely to win as a tighter player.

Next, I'm going to look at the player types (Poker Office and Poker Tracker have filters to automatically categorise players) to see if there are any interesting patterns there.

Have a good easter,

Jon.

lol this is so dumb, vpip is directly correlated to the games you are playing in. These stats you are 'analysing' are essentially meaningless since they all come from different populations...

if u play this sort of style though then u can play a ton of tables and have a much higher hourly.

yea fo sho, but its soooo exploitable and i expect with the games as they are, for any reg to be crushed playing MSNL playing this style.

true, but my post was in reference to the guys comment about ssnl, which is still easily beatable playing that nitty.

I wonder if it is a factor that a lot of players playing 22/18 are relative newcomers or the type of stubborn people who just use a handchart rather than considering each preflop situation on its own merit.

VPIP matters, but its not the only thing

/thread.

If you can dig up the old FTP regs threads when datamining was allowed somebody used to give monthly/weekly lists of the top winners along with their VPIP and other general stats over much more meaningful sample sizes. In general the top winners at SSNL were (are?) much tighter than the standard 22/18/7 tag clones. It's hard to say just using something like PTR now a days since they mix in HU with 6-max.

ike i like how you post the least info of anyone and then do a /thread
rather cute