Poker Efficiency Rating (very long, nerdy)

[PartyGirlUK]

131.8 (last 20.5k hads)

[kurosh]

Some random database I have with hands from 2/4 to 300/600. 85k hands, 1.22BB/100.

24.6% Win% General Info (see Totals section under Known Starting Hands)
45.2% W$WSF General Info
49.5% W$SD General Info
4.75 Average # Players Summary

23186 Seen Flop w/ 2 Players General Info - Filtered (Money In + 2 Players Seen Flop)
6572 Seen Flop w/ 3 Players General Info - Filtered (Money In + 3 Players Seen Flop)
1544 Seen Flop w/ 4 Players General Info - Filtered (Money In + 4 Players Seen Flop)
388 Seen Flop w/ 5 Players General Info - Filtered (Money In + 5 Players Seen Flop)
50 Seen Flop w/ 6 Players General Info - Filtered (Money In + 6 Players Seen Flop)
4 Seen Flop w/ 7 Players General Info - Filtered (Money In + 7 Players Seen Flop)
2 Seen Flop w/ 8 Players General Info - Filtered (Money In + 8 Players Seen Flop)
0 Seen Flop w/ 9 Players General Info - Filtered (Money In + 9 Players Seen Flop)
0 Seen Flop w/ 10 Players General Info - Filtered (Money In + 10 Players Seen Flop)

2.35 ASF Average Players Seeing Flop
1.170 EF1 Win Efficiency
1.062 EF2 Flop Efficiency
0.989 EF3 Showdown Efficiency
122.9 PER Poker Efficiency Rating

[gaming_mouse]

[ QUOTE ]
[ QUOTE ]
Nate,

Good post. I've been wondering about stats like this ever since I saw the "Luck Factor" a couple years ago in the original version of the Poki software. I think that was a more complex formula but I have no idea how "accurate" it was.

Anyway, two things.....

My PER for my last 24K hands of 10 data is 96.4. I'm running at -.24 and don't know how much is due to bad play vs running bad. I did think this was interesing....

EF1: .8
EF2: 1.1
EF3: 1.1

How should I interpret this?

Thanks,
gm

[/ QUOTE ]

gm,

Well, I think it's too early to say how to interpret specific numbers (except that >=100 is probably desirable), or even whether this metric is useful at all.

However, one thing that I thought was interesting was that I calculated my PER for my last 4000 hands, over which I've run (played?) fairly poorly and lost -1.34 BB/100. My PER for these hands is 107.2, which is lower than my usual number, but still comfortably above 100.

[/ QUOTE ]

That's interesing... suggests that the stat is doing it's job. Can you explain a little bit more about why this stat should account for things like being on the bad side of AA/KK or getting 2/3/4-outed on the river fairly often?

Thanks
gm

[PartyGirlUK]

25.7% Win% General Info
47.0% W$WSF General Info
53.2% W$SD General Info
4.23 Average # Players Summary

5840 Seen Flop w/ 2 Players General Info - Filtered
1822 Seen Flop w/ 3 Players General Info - Filtered
350 Seen Flop w/ 4 Players General Info - Filtered
350 Seen Flop w/ 5 Players General Info - Filtered
2 Seen Flop w/ 6 Players General Info - Filtered
0 Seen Flop w/ 7 Players General Info - Filtered
0 Seen Flop w/ 8 Players General Info - Filtered
0 Seen Flop w/ 9 Players General Info - Filtered
0 Seen Flop w/ 10 Players General Info - Filtered

2.43 ASF Average Players Seeing Flop
1.086 EF1 Win Efficiency
1.141 EF2 Flop Efficiency
1.063 EF3 Showdown Efficiency
131.8 PER Poker Efficiency Rating

[Nate tha\\\' Great]

[ QUOTE ]
I used my 10/20 hands:

BB/100 1.9

ASF: 2.7
EF1: .786
EF2: 1.113
EF3: 1.125
PER: 98.5

[/ QUOTE ]

Okay, this worries me a bit because you've been a winning player and yet your PER comes in at lower than 100. (I assume that these numbers represent a reasonable sample size).

I suspect that the beta version PER might be biased toward a loose-aggressive style of play; it might give too much credit for winning pots and not deduct enough for spewing chips when you don't have the best of it.

Something like...

PER = EF1 * EF2 * (EF3^1.5) * 100

...would place more emphasis on winning showdowns and might show a better correlation with winrate. For example, your PER under this version of the formula would be 104.4.

[disjunction]

I'm flooded right now so I can't think about this until tonight, but my first inclination is to change your parentheses and see what I get. For example, the overall Efficiency Rating is something like

(Win percent)*(Win % when saw flop) * (Win % at showdown) * (Avg Players) * (Average # seeing flop) * 2.

So wow, winning a pot that goes to showdown sure does help your rating, it gets multiplied 3 times. Or conversely, losing a pot that goes to showdown sure does hurt? I'm not commenting one way or the other yet, like I say, I'm writing this in hurry. Just observing.

[gaming_mouse]

[ QUOTE ]

PER = EF1 * EF2 * (EF3^1.5) * 100

[/ QUOTE ]

That puts me at 109.

Nate, what about a stat that explicitly takes into account the kinds of bad beats we were talking about (2-4 outers, AA vs KK, etc)? This would have to be a custom query (and the PT data structure, which sucks, might not even have enough info), and it would be alot harder to develop, but it might be really useful for gauging bad runs....

[Nate tha\\\' Great]

Kurosh,

Your EF1 is really high... what's your VPIP?

[mantasm]

[ QUOTE ]
I suspect that the beta version PER might be biased toward a loose-aggressive style of play;

[/ QUOTE ]

I think that's probably true, I play a pretty traditional TAG preflop style (26-27/18-19), and postflop I go to showdown less than many people (34-35 wtsd, 55-56 w$sd)

[Nate tha\\\' Great]

[ QUOTE ]
Nate,

Very interesting. All these metrics are focused on frequency of winning pots. Have you given any thought to metrics which reflect skill at extracting value?

[/ QUOTE ]

I suspect that's very difficult to do, though maybe you could look at something like average # of BB won when winning a showdown (for example my rating in this department is 4.60).