I will also put here what I wrote in that post about how I understand these stats so far. Please tell me if you think the following represents an accurate descripion of these stats, as when I wrote this in the other post I got no response.

AggressionFactor indicates how a player puts money into the pot. If he wants to be the driver (do the betting), then AggrFact is high, but if he is willing to let his opponent be the driver (he is content just calling), then AggrFact is low.

Next, AggressionFrequency indicates how often the player likes to START "driving the betting". If he starts to drive the betting often, then AggrFreq is high, but if he rarely wakes up to drive the betting, then AggrFreq is low.

Finally, VPIP / PFR determine how "fake" the player's aggression is (i.e., how full of bluffs his aggression will be). When VPIP and PFR are high, then his postflop aggression will be "fake", or without a real hand; and when VPIP and PFR are low, then his postflop aggression will be "real", or with a real hand.


An easy way to summarize all of the above is:
- aggression-factor is the "money-in-aggressively-factor"
- aggression-frequency is the "will-be-aggressive-frequency"
- VPIP / PFR is the "preflop-determined aggression-will-be-fake factor"

Please comment on what I have written here.

Is AggressionFact the correct scale for indicating a calling station, and not AggressionFreq?

Nice thread topic curious to find how to analyze bluffing frequency with these stats.

Think u will find its very dependant on vpip and im not just talking lagfish 60vpip Vs 10vpip EG 13vpip vs 17vpip changes it alot. (i think)

Example i was playing 15/10/3.5/36% (approx iv lost all hands due to harddrive crash) at 100nl 200nl obv alot of c-bets bluffs etc

Iv cashed all out and want to rebuild from micros and play live for a change.

Anyway at 1c/2c only 800hands im playing 22/10/2.7/31% i can find like 5 spots iv bluffed!

For teaching leakbuster has AF at 2.4-3.4 optimal Afreq at 28-38 optimal not sure if they adjust by stakes?

For working out bluffers/stations.. WTSD or maybe saw river when saw flop stats are also needed for stations. But for bluffers VPIP and AF and AF% are important.

Also fairly sure i read once an AF of like 1 is really aggressive for a 70vpip player so 0.5 doesnt actually make them passive whereas a 14vpip with 0.5 is very passive.

GW, I think your assumptions in your previous post are pretty good. AFact is largely about your attitude to play when you elect to play or continue with a hand. AFreq, on the other hand, is your attitude to overall play of all hands dealt to you. So, to answer the question above...it's both.

A player that is 15% and 0.8 is likely to be a typical calling station. We know this because he only takes an aggressive action around 15% of the time meaning he must be calling or folding a lot. We also know from his AFact that when he does elect to play a hand (i.e. not fold) he still selects passive options.

However, to truely understand our 15/0.8 we probably need to look at some additional stats. For example, what do we make of him if he is:

a) VPIP = 34%, PFR = 4%, WTSD = 30%
b) VPIP = 10%, PFR = 3%, WTSD = 17%

If our 15/0.8 is Player A then he is clearly a calling station. He likes to limp, likes to see flops and likes to go to a showdown. His 15/0.8 stats reflect his willingness to limp, call raises, call bets and generally not release hands or draws. He's not a great candidate for bluffing.

Player B on the other hand is some sort of super pansy nit/rock. He's learned the 'tight' side of TAG play but has over-applied it and failed to learn to be aggressive. He doesn't get involved until he's really happy about his hand and circumstances and, when he does, he allows others to control the pot for him. He limps, checks and calls, will give up if the board rolls out scary, etc. This guy can be bluffed liberally but never forget he is generally playing a tight range of solid hands and he's highly likley to slowplay.

You're unlikely to come across too many 10/3/0.8/15% in your time, but Player A is far more common. The purpose of this post is to try and make that point that sometimes you need to give a little context to the stats to paint a better overal picture of your opponent.

For example, what are our thoughts on player C: 10/3/4.0/45%.

This is basically the same rock/nit that we identified in Player B but he's more suitably aggressive. Player C realises that by playing largely premium hands he can play then aggressively (usually for value) post-flop. When Player B enters a pot you know he has a hand but you also know he can probably be bullied off anything but the near nuts. Player C on the other hand also has a hand but he's far more aware of it's true value and is less likely to release it.

Acevader,

Could you release version 2 of the table that you showed at the beginning of the post? If you still do not quite like the values that you have in the table, then just call it version 1.1 and not 2.0!

Thanks

Ok, here is my second version of the table. I revised it after running a report in HEM. FWIW, I used to buy hand histories so I have epic samples on many regs - granted most of the hands are a year old now. Based on the results it seems that any AFq below 20% is extremly passive with anything below 25% being passive. Neutral seems to be approximately 25-29% with 29% being the exact average of all players. FWIW, I resticted sample size to 1000 hands or more and 5-10 players to try reduce the impact of any short-handed play. 30-36% seemed about right for the 'aggressive' cateogry and, as you'll see, a good number of winning regs fall into this range. 37%+ is generally very aggressive but not necessarily a bad thing. I suppose I could technically add a "too aggressive" cateogry, which would probably be 50%+ but I don't believe it's required for what I'm trying to achieve.

As for AF, that needed some more serious tweaking. I now regard anything below 2.0 as passive and neutral is around 2.1-2.5. Aggressive is 2.6-3.6 and very aggressive 3.7+. The majority of winning regs are 2.6+ and, I suppose, 7.5+ would be "too aggressive".

The purpose of this chart, as I said before, is to give total newbies something to shoot for and categories to avoid. Having completed this exercise I believe that you should aim to play approximately 30-36% for AFq and 2.6-5.0 for AF. Furthermore, you'd generally benefit more from falling into the "very aggressive" as opposed to "neutral" cateogry. In terms of optimal win-rates "aggressive" is best, followed by "very aggressive", "neutral" and then down. In addition, there is more scope to fall well into the AF "Very Aggressive" cateogry without ill effect than there is to be in the "Very Aggressive" AFq category. A player who is at 42% is defo starting to push his luck and is possibly playing at too many pots. On the other hand a player with an AF of 5.0 is well into the AF "Very Aggressive" scale but not neccessarily at risk of making crucial errors.

The following player names should be familiar and in all cases the sample size is 25,000 hands or more. The majority of these samples are, in fairness, a year old. If your name features and you want it removed them please contact a mod (I'll only be able to edit for 20 mins or so) and have them change your name to "Joe Regular".

McAllin: 27.2%/3.20
BareAce11: 29.7%/2.50
Kush789: 36.9%/4.42
Myulchee80: 29.5%/1.80
Bigkawik: 30.6%/2.89
Cwarno1: 28.7%/3.20
Fittsmurf: 37.8%/5.53
MClark79: 31%/2.84
LostOn4thSt: 34%/3.15
MoreMrNiceGuy: 32.6%/3.27
Al_lupner: 29%/2.76
BerlinsBest: 32.2%/3.11
LRidge: 32.5%/2.48
Mawrich: 32.1%/3.10
Jolohu: 34%/3.97
Elaplayhard: 35.2%/4.55
Torr99: 30.0%/2.66
Turtlelicious: 34.4%/5.84

I'm sure you'll agree there is a good collection of solid winning regulars there the majority of whom play a predominantly 'aggressive' style of poker. As you can see most of them fall into the "Aggressive" categories for both statistics and the purpose of the table is to try and get our newbies to shoot for those cateogories and avoid, like the plauge, the "passive" or "very passive" categories.

Thoughts please

GW have you seen this yet? bttt

This must be why I play 12/10 and regs call down my huge bets with 88 on a JA4 board all the way to the river.

Ace, thought you might want to see these reports from players in my DB with >100 hands:

30.3% aggression / 2.56 factor 200K hands from 10nl to 200nl

If using PT3 35% is passive.

That's interesting/worrying? Why is that? Are the stats calculated differently in HEM and PT3? Does PT3 include 'checking'?

http://forumserver.twoplustwo.com/45...ffrent-614799/

The avarage fulltilt 10nl full ring player has a PT3 AFq 46.31% acording to my database.

Could you tell me the name of the report you ran, and the name of the columns that you used for both AggrFreq% and AggrFact? Was it the "Overall" report? And did you select the two additional stats of "Postflop Aggression Pct" and "Postflop Aggression"?

Click on "Players" tab in HEM, then set #Players to 6-10 to remove short-handed play and set date ranges. Set the minimum number of hands - I recommend 1000 as that will allow for some decent samples. Use 'Default Stats' and then run the report. You'll see there is a column for both Postflop Agg and Postflop Agg%. Click on either of them and you'll get a breakdown in "results analysis".

Play around with the reports, stats, etc

are these numbers legit for pt4 afq? it seems most people run above 40-50% afq , are they all very aggressive? can anyone make a similar column for pt4 afq?

[x] very passive

i tot everything under 3 was passive

I guess im alright with my 2.8 afterall

Less than 40% is passive and anything above 55% is aggro.