Are HUD Stats Basically Useless?

Aleksei

Aleksei

Legend
Silver Level
Joined
Feb 20, 2013
Total posts
1,527
Chips
0
Frankly I think the standard base stats of VPIP/PFR/3bet are kind of borderline-useless. They just give you this super-stereotyped general image of how someone plays, and no real info that you can actually use to spot or exploit specific weaknesses.

Like, say someone has a VPIP/PFR of 38/15. What does that even mean? Does that mean they open-raise tight but then flat a ton? Do they limp a lot and raise the top of their range, but fold very often to someone else's open? And so on and so forth. It's just such a general stat that tracks so many completely different (and potentially unrelated) situations it doesn't tell you anything about any of them individually.

3bet preflop is much worse, because 3bets are an inherently polarizing situation so you want to have a good idea of exactly what someone is 3betting with in a spot -- which you will not get from a general 3bet stat because 3betting OOP is a completely different situation from 3betting IP, and then 3betting OOP vs early open is also completely different than restealing. so you generally wanna, you know, actually know how often Villain does each.

I don't even use the 3bet PF stat. I have a few 3bet vs position stats in my HUD and then use the popup for the rest. My HUD right now looks like this:

Notes/Player Name/hands/Live Amt BB
Limp/Call Any PF Raise/Open-Raise/Fold vs BTN Open/Fold vs CO Open
BB vs SB Open - Fold/Call/3Bet
3Bet vs BTN Open/3Bet vs CO Open/4Bet PF

Then I just use the popup for everything else.
 
benevg

benevg

Legend
Silver Level
Joined
Jul 22, 2007
Total posts
1,267
Chips
0
congratulations, you have now reached a point where understanding people's stats leads to understanding of their tendencies. this is really good and also potentially very profitable (i should know, it seems to be the only way i make any money ever).

you point about the stats not actually meaning much in a vacuum also has some merit. :stupido2:however, after a while you may find yourself buried in data which has not converged yet and is thus unreliable. so you would think you know something, and then it would turn out your knowledge is actually not quite true. it seems you are either prone to over-analysis, or still lacking the experience to synthesize how the different aspects of what you are looking at are actually the same. i am willing to bet that after a while you will grow back to liking the more general stats, even though the specific ones will still be more useful in any particular hand.

i could also make a different argument, but i think i'll leave that to the words of Andrew Seidman, aka BalugaWhale:
The world of online poker has become dominated by statistical analysis programs, hand history recorders and replayers, and HUDs (heads-up displays). For any given sample size of hands, you can find out everything from broad, easily used stats like preflop looseness, preflop raise percentage, and total aggression factor, to extremely specific statistics—fold to river check-raise percentage, etc. Worthless is a little bit too strong of a word, but in my opinion most statistics are extremely unimportant.
 
Deco

Deco

Legend
Silver Level
Joined
May 7, 2009
Total posts
2,544
Chips
0
My HUD:

VPIP/PFR/3B/Steal(CO)/Steal(btn)/Fold vs Steal(bb)
3b/F3b/F3b(co)/F3b(btn)
4b/F4b/cb/Fcb/Call open
SB Steal/BB Fold vs SB steal/Hands


My most used pop-ups are a PFR pop-up with open raise %s for every position and a 3bet% with positional 3bet%s and vs<position> 3bet%s. Both used in combination are great vs 3bet%s are likely the more useful of the two.


Frankly I think the standard base stats of VPIP/PFR/3bet are kind of borderline-useless.

I couldn't disagree with this more. The stereotype you speak of is far more useful than no information and for the most part if all we can use with recreational players as we rarely get the hands needed for more articulate stats.

Like, say someone has a VPIP/PFR of 38/15. What does that even mean? Does that mean they open-raise tight but then flat a ton? Do they limp a lot and raise the top of their range, but fold very often to someone else's open? And so on and so forth. It's just such a general stat that tracks so many completely different (and potentially unrelated) situations it doesn't tell you anything about any of them individually.

It means there a fish, this is the most useful information we can possibly hold. Whether they flat or open limp for all this we can find out using an open limp% or as I prefer the call open%. Although tbh that vpip/pfr gap is nearly always made up from flats. Fish who open limp at very large frequencies have a much larger VPIP or a much larger gap, 38/0 or 60/15.

3bet preflop is much worse, because 3bets are an inherently polarizing situation so you want to have a good idea of exactly what someone is 3betting with in a spot -- which you will not get from a general 3bet stat because 3betting OOP is a completely different situation from 3betting IP, and then 3betting OOP vs early open is also completely different than restealing. so you generally wanna, you know, actually know how often Villain does each.

I don't even use the 3bet PF stat. I have a few 3bet vs position stats in my HUD and then use the popup for the rest. My HUD right now looks like this:

Everything you say here is true. However positional 3bet stats take absolutly ages to converge. Until we've over 1000 hands the 3bet% is all we can use and is once again better than nothing.
 
vinylspiros

vinylspiros

PIRANHA-------->< (((º>
Silver Level
Joined
Oct 9, 2012
Total posts
4,393
Awards
1
Chips
1
Ill answer you with an example.

Lets say you are holding QQ. and you raise from under the gun to 3x.

would you feel more comfortable calling an all in from someone whos vpip /pfr is: (both of which u have 300 hand with)
10/8 or 65/ 49? you being able to answer this question is proof that HUD stats are not useless.

notice i mentioned general stats and not specific stats.
 
benevg

benevg

Legend
Silver Level
Joined
Jul 22, 2007
Total posts
1,267
Chips
0
Ill answer you with an example.

Lets say you are holding QQ. and you raise from under the gun to 3x.

would you feel more comfortable calling an all in from someone whos vpip /pfr is: (both of which u have 300 hand with)
10/8 or 65/ 49? you being able to answer this question is proof that HUD stats are not useless.

notice i mentioned general stats and not specific stats.
lol - for 300 hands, if i have not made a note and/or color-coded the 65/49 (assuming those stick around for that long), there must be something wrong with my method :p
 
vinylspiros

vinylspiros

PIRANHA-------->< (((º>
Silver Level
Joined
Oct 9, 2012
Total posts
4,393
Awards
1
Chips
1
lol - for 300 hands, if i have not made a note and/or color-coded the 65/49 (assuming those stick around for that long), there must be something wrong with my method :p
hehehhe, yea ok 300 hands is a small sample. point i was trying to make is that something is better than nothing.:rolleyes:
 
Deco

Deco

Legend
Silver Level
Joined
May 7, 2009
Total posts
2,544
Chips
0
lol - for 300 hands, if i have not made a note and/or color-coded the 65/49 (assuming those stick around for that long), there must be something wrong with my method :p

People multi-table were HUD stats become more necessary.
There are less extreme examples were even one tabling obeservation will fail. Otherwise I wouldn't bother creating a HUD for live games :p

The difference between a 70/20 and a 45/15 is vast yet cannot be kept track of well in our heads.
VPIP/PFR can be useful from as little as 4-9 hands. A 60/0 (5 hands) will more than likely turn out to be a fish.
Humans are increrdibly inaccurate with coming up with 3bet%s, people generally base how oiften someone 3bets based on how often someone has 3bet them recently.
 
Top