Infinite River Aggression

I've been following the Absolute Rigged thread lately, and while I'm staying out of the fray for the time being, I thought I should clarify what I think is a misunderstanding of one of the Poker Tracker stats which, in its face, seems to be the most damning evidence against Absolute and the alleged culprits, namely the infamous infinite river aggression. I've heard both Chris (Dorkus Malorkus) and F Paulson, two well respected and usually well informed members make the same erroneous statement about what this stat means.

" The infinite aggression factor means that they never call on the river, only bet/raise or fold."

This is incorrect and as it seems to be the linchpin in the argument against Absolute it's important to get it right. In Poker Tracker, the AF (Agression Factor) is calculated as Raise % + Bet % / Call. An infinite number is achieved when one attempts to divide any number by zero. In this case, the zero is the number of times the player in question called a river bet. So where's the problem? DM and FP have left out one more option. CHECKING.

OK, so how does that change the situation as it exists? I mean this guy never once in 190 hands at $200/$400 called a river bet. Isn't that suspicious enough in itself? If all he did was Bet Raise or Fold then yes, but unavailable in any of the documentation I've read is the number of times this person GRAYCAT checked the river. I've done a little investigating and I've come to the conclusion that infinite river aggression factors for players with even several hundred hands are neither unheard of nor rare. In fact I was surprised to find that our very own Tenbob had an infinite aggression factor with over 400 hands in a tournament I railed at FTP.

Well, what about the super high win rate at showdowns? Again, very good, but not unheard of. In 207 hands of NL at Full Tilt, John Juanda never once called a river bet and had a showdown winrate of 63%. This compares nicely with GRAYCATs 67%. Or how about Chris Furguson? 325 hands at $5/$10, Inf. AG Winrate 61%. And why is this surprising when every good poker book I've ever read has suggested that one should play the river in just this manner?

I have no desire to reopen the RIGGED debate but I do have more to add to this one part of it. Unfortunately this is all I have time for tonight. I do look forward to reading some of your comments.

-FD

Best post on this topic so far Four Dogs. This is something i havnt really been following with a massive interest, just somewhat of a slight interest. Your correct totally in saying that the sample size is simply too small. Having someone like youself that data mines quality players like Chris Furg and myself (hic) gives your data a little more credibility. However, first off tournament play is a totally different beast than the rings, and you can be sure that over 400 hands on a ring table i wont have an inf river agg. In tournaments, its my style of play, i rarely call a river bet, im either check/fold, check/raise, or bet/fold, stack sizes usually dictate it. In a ring game calling down light is a much better proposition for lots of reasons too complex to go into now.

The one thing i wont do, is knock some of the wisdom of the high stakes players on 2+2, some of these guys are simply awesome, and if they say something is true, it usually is. Nice to see a different point of view on the whole thing, if anything its got me thinking about how im playing the river.

Got me thinking too. I've always been a firm believer in checking the river and then calling with any potential winner. Now that I take a closer look at some of these super high river AF's accompanied by high winrates I'm starting to wonder if there might be some lost value there. F Paulson posted a thread a while back expressing the same concern. I think I may have advocated a more passive approach to river betting; check/call as opposed to betting out. A wonder if I haven't been brainwashed more than a little by Sklansky and Harrington. Time for a little out of the box thinking.

Nice post.

Without actually seeing the related DBs I can't comment conclusively, but this most likely applies.

If the Ferguson/Juanda stats are for full ring and they're playing a relatively standard TAG game, it's entirely possible that in 200 hands they're seeing maybe a dozen rivers. Compare this with Greycat et al, who was primarily playing HU or shorthanded, and playing at 90% VPIP. In his 400 hands, he's probably seeing 200 rivers or so (this is just a guess, but you get my point).

Now sure, someone could have an infinite river AF over a dozen hands and be well up - that can be chalked down to short term variance and a small sample size. When we start talking in terms of a couple of hundred rivers seen, though, we start delving into the realms of the results being too distant from expectancy even accounting for the large standard deviation playing an incredibly loose style brings.

Just like someone having a W$SD% of 60% over say 1000 hands is not in itself unusual, but someone playing with a VPIP of 90% and a PFR of 70% having that same W$SD% over the same sample size would be, for example.

I'm sure there's more I could say on this, but I'll dig it up again when I'm more awake tomorrow.

Edit: FWIW you're right about the checking, and I was aware of it, but just saw (and indeed still see) it as unimportant.

i dont want to turn this awesome thread into a poker stratety thread, but why wouldnt you check call the river? Let's say you have a mediocre hand and you do not want to bet it because you will only get called by a better hand, or get raised and not be able to showdown a possible winner....learned this in one of Chris Ferguson's articles
VVVVVV
http://www.chrisferguson.com/article-2

its a pretty good article, enjoy

again sorry to make the post seem more strategical and not as relevant to the OP