I mainly donk around between $2/$4 and $10/$20 online. I've started looking up the stats on PTR about the players at my tables to see their win rates. Upon looking at the "report card" function, which shows their "stats" in grade format, as compared to the top 10 winning players at those same stakes, I noticed that many players have stats that appear impossible to attain, for humans. What I mean is, it seems really unlikely that aside from the random outlier, it would be very tough for any human player to play a style that would line up so close to optimal on all betting rounds of a holdem hand. For example, here is the report card of one particular player:

PTR Poker Report CardPrepared for: XXXXXXXX

This report card is based on 108,053 hands tracked at $5/$10 FL Hold'em (6 max) by PokerTableRatings.com

Subject Grade

Pre-Flop Tightness
You are 0% looser than optimal. Good Work! A+
Pre-Flop Aggression
You are 0% more aggressive than optimal. Good Work! A+
Flop Aggression
You are 0% more aggressive than optimal. Good Work! A+
Turn Aggression
You are 3% more aggressive than optimal. Good Work! A+
River Aggression
You are 1% more aggressive than optimal. Good Work! A+
Showdown Frequency
You showdown 3% more often than optimal. Good Work! A+ Rakeback earnings:
You would have earned $26,164.72 if you had
PokerTableRatings.com Rakeback!
$26,164.72
Overall Grade A+

This kind of report card is not that uncommon. From all the players I've looked up, it seems that more than half have stats that are at least in single digit differentials to "optimal" play. While this is not necessarily an accusation of bot infiltration in the limit games, it does seem unusual. Are the humans all really playing such a similar style? If so, these games simply cannot be beaten anymore by anyone playing at less than an expert level.

I play a pretty high level game myself. I am a winning player (although barely) at these levels ($2/$4 through $10/$20). I also consistently beat live games up through $100/$200. However, some of my stats on PTR can be as much as 20% off the "optimal" figures for certain streets, although my sample size may not be significant enough to draw too many conclusions.

The reason I'm posting this is just to get thoughts on how to interpret what those PTR stats mean, and if any conclusions can be drawn from them.

Your huge fail is seeing the word "optimal" in neon lights and not being able to logically figure out how they are using it. They are using the top 10 players as optimal. This is just your grade of your stats compared to theirs.

That doesnt mean that the top ten players play optimal, they have just won the most money and you are being compared to them. Optimal play might not ever show up if the top 10 winningest players at that stake put in 10x the volume as the guy playing your definition of optimal.

To conclude, you are making this way more than it should be. Again, just because someone uses the word, "optimal" doesnt mean that is the case.

Does PTR define "winning" in terms of total dollars won or in BB's/100?

Still, I don't think my fail is as big as you think. To the unlying issue, don't you find it unusal at all that even if PTR defines "optimal" as to how the top ten winners are playing, then how do you explain that so many other players are playing a style so incredibly similiar to those ten players? Is the whole world not only watching the same training videos, but able to execute the strategy with such consistent balance? Seems odd to me.

It's total dollars won, not BB/100.

What do you mean so many others? It shouldn't be unusual that the top ten winners play similarly. In fact, the next ten on the list are probably only ranked 10-20 based on variance or volume (since it's limit, I'd go much further down the list). A good heater brings them in the top ten, or simply running 0.5BB over EV while a 1-10 player runs 0.5BB under EV.

Optimal for some players wouldnt be optimal for others. Many different styles can produce various winning strat's

yeah this, and it depends heavily on the other players at the table



for example this reg (on left of image) who i play with on a daily basis has better stats than me on that PTR report card but there's no way he's playing more appropriately than me given all the fish in our games

he's also definitely not a bot but is pretty close to their optimal stats. i don't think it would be very difficult to have good stats on this PTR page but still be a so-called 'badtag', having a good understanding of basic strategy and having good stats but still making consistent human errors and playing far from GTO.

it's an interesting question, you might want to post in the FL forums

edit: this is not on FTP, AP or UB

why are we even talking about this, PTR doesn't even track FTP anymore, and misses a ton of hands for AP/UB. IDK about stars since I don't play there, but the missing hands combined with other things would def. skew these stats.

You can pretty much throw out any conclusions you are making based on 30K hand samples.

ldo. really more of a 100k sample over all stakes, but mostly i was talking about the way i see him play hands day in day out

Meh, I was talking more about yours.

clearly aren't playing in the same games as me then troll :P. i couldn't achieve that on stars or FTP, but it's standard where i am

I'm not trolling. Do you think your EV is 5BB/100? If so, how big is your sample to substantiate that claim? It's a legit question. I'm not being antagonistic - I don't play micro/small games. What is your standard deviation?

I'm usually within 10 % points in the various games I play and usually between a B+ & A+ in overall grade and I've never been coached, etc in my life.

On partypoker, if the .25/.50 games stay as they are, I think my winrate is somewhere between 3.5 to 5.5BB/100. Std dev BB is ~ 16.5. My sample size is only 40k, and I'm aware of confidence intervals and variance, but there are just very few regs, for example not to brag but on the top 50 winners list on PTR (pretty accurate for party), if you filter by descending winrate, mine is the highest of those. The highest all time hands played count is only 175k, with the average of those 50 players being 67k (on stars it's 475k). I beat .10/.20 on FTP for 2.5BB/100 over 80k hands fwiw, and found the games there much harder. the highest .25/.50 FL all time $ won for stars, FTP and party respectively are $10k, $2.7k and $1.6k respectively. Each site has roughly equivalent rakeback/VIP reward systems for high volume microstakes players, it must just come down to the fact that US players are restricted that party is so much softer. If my winrate drops after I have a >200k hand sample I'll happily eat my words.

I was thinking the same thing when I saw your post. Could be cause I'm on a 130k hand break even stretch.......

as you're prob aware, the smaller the edge, the longer the possible breakeven stretch. check out donjuan's awesome thread where he beats the stars midstakes games for a small winrate 24 tabling over millions of hands but has 200k BE stretches if you haven't seen it. but when you can sit at 1/2 with 3 players with vpip>40 at party it's hard to lose. anyway /derail