Prospective stat for Poker Ace: PFR magnitude

[wslee00]

I think this stat would be very useful, since I see the occasional player always raising either 2x or 3x bb. Having this stat would tell me whether his 20% pfr is actually him being hyper aggressive or someone that just likes to sweeten the pot

[ QUOTE ]
A PFR magnitude does nothing for me peronally and I really don't understand the point. Do you really think that someone lets say raises 50 times and that they will always only mini-raise and then suddenly raise a large amount and you shouldn't ALREADY be aware of this if you've looked at the game at all. Not to mention this thing called the law of averages and the fact that it's unlikely that something will deviate enough to allow you to make a solid conclusion.

Sorry I just don't get it. I mean I understand what you're saying, I just don't see a way to make it applicable enough to apply a read. Understanding raises is more about seeing what people raised, and in what position, and matching that with some sort of pattern based on the hands they show down.

[/ QUOTE ]

Your points would apply to most all HUD stats. Do you recommend not using pfr or vpip?

[Bukem_]

Its really wouldnt be that helpful.

The problem is some people minraise certain hands, and make pot sized raises with others. If you are just looking at the average, you don't realize this.

Paying attention to what people are raising with is much better.

[ QUOTE ]
Its really wouldnt be that helpful.

The problem is some people minraise certain hands, and make pot sized raises with others. If you are just looking at the average, you don't realize this.

Paying attention to what people are raising with is much better.

[/ QUOTE ]

Yes, paying attention is always better, all else being the same. Averages don't address variances or complex patterns, of course. But that is true for most of the stats we use. e.g., vpip and pfr don't address position. I haven't heard anyone dismiss these stats as unvaluable.

[Mercman572]

[ QUOTE ]
[ QUOTE ]
Its really wouldnt be that helpful.

The problem is some people minraise certain hands, and make pot sized raises with others. If you are just looking at the average, you don't realize this.

Paying attention to what people are raising with is much better.

[/ QUOTE ]

Yes, paying attention is always better, all else being the same. Averages don't address variances or complex patterns, of course. But that is true for most of the stats we use. e.g., vpip and pfr don't address position. I haven't heard anyone dismiss these stats as unvaluable.

[/ QUOTE ]

I'm with you Tera. I notice some people minraise sometimes, but I don't notice what percentage of times. 40% wouldn't say much, but 90 would be nice.

I think this would be a valuable stat. I know plenty of players who never minraise, and plenty who always minraise. If 2 is a minraise, any stat between 2.0-2.8ish or above 3.8ish would be valuable to see if the villain does anything out of the ordinary.

[grouchie]

i thnk something like this would be kind of valuable but like someone else said, you are usually going to notice that type of thing and have a not that he almost always mini raises.

[Cosimo]

[ QUOTE ]
But that is true for most of the stats we use. e.g., vpip and pfr don't address position. I haven't heard anyone dismiss these stats as unvaluable.

[/ QUOTE ]

I think PFR > raise amount. PFR is a very course measure. As we add more and more stats, they obviously get more specific -- the sort of stats useful in fewer circumstances. I think PFR is a great stat because of its courseness. Stats that are general in nature will converge faster and be applicable to a wider range of situations, and hence preferable. Although an accurate position-relative PFR (etc) would be keen, I would need a ton of hands before it started to converge and the main thing it would tell me is if villain is position-aware, which only loosely helps me play against him.

Currently, I use AF, flop aggression %, and flop fold % together to assess bluffiness/passivity/what to do when he bets or raises the flop. I wish there was an easier way to do this, but I donk think stats alone would help. I bring this up because I think a "% minraise" would be great, and it could be combined with a "average more-than-minraise amount" stat. But the latter is too specific. You'd need 100+ raises to get something good out of that stat. Even more handy would be something like "% chance 3bets with bluff" or "likely hands when bets more than 4xBB" or "uses 4BB+1/limper" type stuff, which is of course WAY too grossly specific, difficult to engineer, or would require huge hand samples before it was useful.

So, I'd say, "% minraise" would be the most useful stat along the lines being discussed here. This could probably be broken down into preflop, flop, turn, river, and postflop average stats -- and I'd likely only use the preflop value and postflop average.