Hey all,

I'm working on my own HUD at the moment which I'm hoping to share with the world some day. I'm not aiming to do anything revolutionary, but want to make some small enhancements like highlighting stats in a way that makes the most relevant stats for a given street the most discernable. For example, highlighting Fold to C-Bet frequencies of our villains when it's our turn on the flop and we're the preflop raiser, or highlighting the C-Bet Frequency on the flop of the original raiser when we're not the PF raiser.

One small innovation I want to try to make is the concept of a 'normalized' or 'weighted' VP$IP to indicate the overall tightness or looseness of a player. It might be overly ambitious: VP$IP is one of those essential HUD stats, but the thing I never liked about it is that we have to expect it to vary based on the size of the table.

Naturally we'd expect a much higher VP$IP in HU than we do 6-ring where we still expect a higher VP$IP than a full ring game. And it's all too common at the micro stakes I play for a 6-max table to temporarily become extremely short-handed at high frequencies after a few players bust out or sit out, leaving the table 2 or 3-handed and likewise throwing off VP$IP measurements.

I want a measurement that stays constant regardless of how short-handed a table becomes by making it somehow proportional to the size of the table. I need some help with this from very experienced poker players who have a lot of experience with HUDs.

A simple way to do this is to multiply VP$IP by the number of players at the table to come up with a 'Weighted' VP$IP. For example:

A player who plays 19% of hands at full-ring would have a WVP$IP of 171 (19 * 9 = 171).
A player who plays 28.5% of hands at 6-max would also have a WVP$IP of 171 (28.5 * 6 = 171)
4-handed, a player would have to play 42.75% of hands to get the same WVP$IP.

The problem is that this doesn't sound exactly right to me. I'd expect a 19% reg playing full ring to probably play around 25% 6-handed, not 28.5%, and maybe around 33-35% 4-handed, not 42.75%. Likewise to get the same WVP$IP of 171 HU, he'd have to play a whopping 85.5% of hands which I think is way too much for someone who plays 19% 9-handed.

So it's not exactly linearly proportional: a player's natural VP$IP and the number of players at the table, and what I want to try to come up with is a reasonably solid formula that can capture a player's overall tightness/looseness in a way that doesn't fluctuate considerably with the number of players at the table.

I imagine the reason it's not strictly linear is due to positional constraints. Regardless of whether we're dealing with full ring, 6-max, or 4-handed games, we're still dealing with positional constants like the button, SB, and BB. A fancy way to try to calculate this would be to weight it based on position. This way, the number of players is irrelevant, only the position of the person putting money into the pot. Yet I think this gets overly complicated to try to come up with a single number from all of this if it's weighted by position rather than table size.

I was wondering what you all think about this idea, and also I'd like help from experienced people. For example, given a player who plays around 19% VP$IP full ring, what sort of average VP$IP would you expect him to play at 6-max? How about when two players leave and the table becomes 4-handed? How about 3-handed? If I get enough samples and thoughts, I might be able to come up with some formula that accurately captures a player's overall looseness/tightness independent of the number of players at the table.

It's ok, or you can just use steal vpip which isn't that vulnerable to number of players (only when it's HU and 3 pl maybe).

If you want to follow your idea you must find how that function looks. If you choose position weight, keep in mind that vpip from BUT 3 handed could be different from BUT vpip 9 handed, cause 9 handed there are less unopened hands for BUT. So if you go that road use some other stats, like unopened VPIP.