Poker Efficiency Rating (very long, nerdy)

[piki]

190k sample, 3/6 to 10/20 6m with taggish stats 23/17/2.5 and 1.2bb/100

15.9% Win% General Info (see Totals section under Known Starting Hands)
45.5% W$WSF General Info
52.1% W$SD General Info
5.27 Average # Players Summary

26679 Seen Flop w/ 2 Players General Info - Filtered (Money In + 2 Players Seen Flop)
16595 Seen Flop w/ 3 Players General Info - Filtered (Money In + 3 Players Seen Flop)
5368 Seen Flop w/ 4 Players General Info - Filtered (Money In + 4 Players Seen Flop)
1158 Seen Flop w/ 5 Players General Info - Filtered (Money In + 5 Players Seen Flop)
97 Seen Flop w/ 6 Players General Info - Filtered (Money In + 6 Players Seen Flop)
0 Seen Flop w/ 7 Players General Info - Filtered (Money In + 7 Players Seen Flop)
0 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)

2.63 ASF Average Players Seeing Flop
0.835 EF1 Win Efficiency
1.194 EF2 Flop Efficiency
1.042 EF3 Showdown Efficiency
103.9 PER Poker Efficiency Rating

120.1 PER 2 ^.5, ^1.25, ^1.25

-----

17k hands at 20/40+, 28/21/2.0 and running at hot 4.1bb/100

18.6% Win% General Info (see Totals section under Known Starting Hands)
49.2% W$WSF General Info
55.3% W$SD General Info
5.05 Average # Players Summary

3012 Seen Flop w/ 2 Players General Info - Filtered (Money In + 2 Players Seen Flop)
1182 Seen Flop w/ 3 Players General Info - Filtered (Money In + 3 Players Seen Flop)
232 Seen Flop w/ 4 Players General Info - Filtered (Money In + 4 Players Seen Flop)
36 Seen Flop w/ 5 Players General Info - Filtered (Money In + 5 Players Seen Flop)
2 Seen Flop w/ 6 Players General Info - Filtered (Money In + 6 Players Seen Flop)
0 Seen Flop w/ 7 Players General Info - Filtered (Money In + 7 Players Seen Flop)
0 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)

2.39 ASF Average Players Seeing Flop
0.938 EF1 Win Efficiency
1.178 EF2 Flop Efficiency
1.106 EF3 Showdown Efficiency
122.2 PER Poker Efficiency Rating

134.8 PER 2 ^.5, ^1.25, ^1.25

-pix

[StellarWind]

Interesting post. I apologize for not having time to digest the entire thread but I wanted to toss a couple things out for consideration.

1. EF3 favors players who miss value bets on the river. Everytime you bet the best hand and Villain folds your PER drops. EF3 might be better if it counted every hand that sees the river and not just showdowns. You paid a lot of money to build this hand and the $64 question is whether or not it's any good. The actual river play doesn't matter very much unless someone folds the best hand. I don't know how easy this is to do in PT but there are some stats on the details page that may help.

2. I'm concerned that EF1 is a proxy for looseness rather than correct play. Two winning players who are very similar except that one plays a large number of zero EV hands (preflop and flop) will have very different EF1s with no real difference in results. Under this metric a tighter player is never going to look good compared to a 30% VP$IP player, regardless of who gets the money.

3. EF2 will be strongly affected by blind play because it does not factor in cost to see the flop. There is a tremendous difference between putting in 2 SB from LP (steal or isolation) compared to receiving odds to complete the SB or call one in the BB. Those who play a lot of blind hands shouldn't expect a high EF2 but they don't need one either--cheap flops and early departures. So if your style is slanted toward raising a zillion hands because you have the button you can expect a high EF2 even though you might be losing your shirt. Conversely you may be coining money by playing every cheap blind hand that comes along, but your EF2 will suffer terribly.

Perhaps the overall message is that PER--both this particular formula and possible improvements--will be most meaningful in comparing basically good players with a similar style. As soon as you compare two players with significantly different styles it will quickly become a question of whose style best fits the formula being used.

[piki]

Here is another one, 50/100 and 100/200, 60k hands at 2.5bb/100, 34/25/1.6

20.9% Win% General Info (see Totals section under Known Starting Hands)
45.7% W$WSF General Info
51.6% W$SD General Info
5.39 Average # Players Summary

12749 Seen Flop w/ 2 Players General Info - Filtered (Money In + 2 Players Seen Flop)
4737 Seen Flop w/ 3 Players General Info - Filtered (Money In + 3 Players Seen Flop)
932 Seen Flop w/ 4 Players General Info - Filtered (Money In + 4 Players Seen Flop)
139 Seen Flop w/ 5 Players General Info - Filtered (Money In + 5 Players Seen Flop)
7 Seen Flop w/ 6 Players General Info - Filtered (Money In + 6 Players Seen Flop)
0 Seen Flop w/ 7 Players General Info - Filtered (Money In + 7 Players Seen Flop)
0 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)

2.38 ASF Average Players Seeing Flop
1.128 EF1 Win Efficiency
1.087 EF2 Flop Efficiency
1.033 EF3 Showdown Efficiency
126.6 PER Poker Efficiency Rating

122.8 PER 2 ^.5, ^1.25, ^1.25

-pix

[TeeJayOrTJ]

147k sample - 2.26bb/100 this is 10/20

19.5% Win% General Info (see Totals section under Known Starting Hands)
44.2% W$WSF General Info
49.5% W$SD General Info
7.69 Average # Players Summary

26381 Seen Flop w/ 2 Players General Info - Filtered (Money In + 2 Players Seen Flop)
16881 Seen Flop w/ 3 Players General Info - Filtered (Money In + 3 Players Seen Flop)
5051 Seen Flop w/ 4 Players General Info - Filtered (Money + 4 Players Seen Flop)
793 Seen Flop w/ 5 Players General Info - Filtered (Money In + 5 Players Seen Flop)
63 Seen Flop w/ 6 Players General Info - Filtered (Money In + 6 Players Seen Flop)


2.60 ASF Average Players Seeing Flop
1.008 EF1 Win Efficiency
1.149 EF2 Flop Efficiency
0.991 EF3 Showdown Efficiency
123.0 PER Poker Efficiency Rating (with the new calculations)

And because I was bored I looked at a hot streak

63k sample - 3.70bb/100 this is 10/20

19.2% Win% General Info (see Totals section under Known Starting Hands)
44.4% W$WSF General Info
50.5% W$SD General Info
5.21 Average # Players Summary

10892 Seen Flop w/ 2 Players General Info - Filtered (Money In + 2 Players Seen Flop)
7193 Seen Flop w/ 3 Players General Info - Filtered (Money In + 3 Players Seen Flop)
2150 Seen Flop w/ 4 Players General Info - Filtered (Money + 4 Players Seen Flop)
339 Seen Flop w/ 5 Players General Info - Filtered (Money In + 5 Players Seen Flop)
28 Seen Flop w/ 6 Players General Info - Filtered (Money In + 6 Players Seen Flop)


2.61 ASF Average Players Seeing Flop
1.001 EF1 Win Efficiency
1.160 EF2 Flop Efficiency
1.010 EF3 Showdown Efficiency
127.1 PER Poker Efficiency Rating (with the new calculations)

[Alobar]

73K hands at 1.95BB/100

Win%: 15.13
W$WSF: 43.41
W$SD: 53.72
5.35 players
2 players seeing flop: 9650
3 players seeing flop: 6509
4 players seeing flop: 2261
5 players seeing flop: 449
6 players seeing flop: 43
total flops seen: 18,912
ASF: 2.65

EF1: .8095
EF2: 1.15
EF3: 1.077

PER (original formula): 100.2
PER (revised formula): 117.5

[gaming_mouse]

[ QUOTE ]
Interesting post. I apologize for not having time to digest the entire thread but I wanted to toss a couple things out for consideration.

1. EF3 favors players who miss value bets on the river. Everytime you bet the best hand and Villain folds your PER drops. EF3 might be better if it counted every hand that sees the river and not just showdowns. You paid a lot of money to build this hand and the $64 question is whether or not it's any good. The actual river play doesn't matter very much unless someone folds the best hand. I don't know how easy this is to do in PT but there are some stats on the details page that may help.

2. I'm concerned that EF1 is a proxy for looseness rather than correct play. Two winning players who are very similar except that one plays a large number of zero EV hands (preflop and flop) will have very different EF1s with no real difference in results. Under this metric a tighter player is never going to look good compared to a 30% VP$IP player, regardless of who gets the money.

3. EF2 will be strongly affected by blind play because it does not factor in cost to see the flop. There is a tremendous difference between putting in 2 SB from LP (steal or isolation) compared to receiving odds to complete the SB or call one in the BB. Those who play a lot of blind hands shouldn't expect a high EF2 but they don't need one either--cheap flops and early departures. So if your style is slanted toward raising a zillion hands because you have the button you can expect a high EF2 even though you might be losing your shirt. Conversely you may be coining money by playing every cheap blind hand that comes along, but your EF2 will suffer terribly.

Perhaps the overall message is that PER--both this particular formula and possible improvements--will be most meaningful in comparing basically good players with a similar style. As soon as you compare two players with significantly different styles it will quickly become a question of whose style best fits the formula being used.

[/ QUOTE ]

As usual, great analysis.

[disjunction]

[ QUOTE ]


1. EF3 favors players who miss value bets on the river. Everytime you bet the best hand and Villain folds your PER drops.


[/ QUOTE ]

Actually you can take this argument to a logical extreme. The correct strategy to increase your efficiency rating is, from the flop on, to keep your opponents in the hand if you think you'll win. Avoid betting in spots where you're ahead, this will probably even increase your EF when your opponent is drawing to something that will beat you. By checking through the turn with top pair you're taking a small risk to your EF1 and EF2 for a high-probability gain of some EF3.

But if you don't think you'll win, bluff as much as possible because it can only help your EF.

[disjunction]

[ QUOTE ]



3. You see a flop but fold before showdown
4. You see a flop and win without a showdown
5. You go to showdown and win



[/ QUOTE ]

Using these seem especially intriguing. Of course, I'm not actually interested in seeing you do this. [img]/images/graemlins/grin.gif[/img]

[luckyharr]

Nate: More data if you need it. The two columns at the bottom are the two different methods you've proposed.

67,293 10/20, 2.11 BB/100. VPIP 35.55, PFR 24.04 if that helps.

20.4% Win% General Info (see Totals section under Known Starting Hands)
42.6% W$WSF General Info
50.5% W$SD General Info
5.36 Average # Players Summary

12991 Seen Flop w/ 2 Players General Info - Filtered (Money In + 2 Players Seen Flop) 2
8285 Seen Flop w/ 3 Players General Info - Filtered (Money In + 3 Players Seen Flop) 3
2653 Seen Flop w/ 4 Players General Info - Filtered (Money In + 4 Players Seen Flop) 4
530 Seen Flop w/ 5 Players General Info - Filtered (Money In + 5 Players Seen Flop) 5
47 Seen Flop w/ 6 Players General Info - Filtered (Money In + 6 Players Seen Flop) 6
3 Seen Flop w/ 7 Players General Info - Filtered (Money In + 7 Players Seen Flop) 7
0 Seen Flop w/ 8 Players General Info - Filtered (Money In + 8 Players Seen Flop) 8
0 Seen Flop w/ 9 Players General Info - Filtered (Money In + 9 Players Seen Flop) 9
0 Seen Flop w/ 10 Players General Info - Filtered (Money In + 10 Players Seen Flop) 10

2.63 2.63 ASF
1.091 1.045 EF1
1.120 1.152 EF2
1.009 1.011 EF3

123.3 121.7 PER

[BabyJesus]

5/10 6 max
145,248 1.34 BB/100

18.6% Win%
41.3% W$WSF
48.8% W$SD
5.28 Average # Players

23009 Seen Flop w/ 2 Players
18621 Seen Flop w/ 3 Players
7849 Seen Flop w/ 4 Players
1752 Seen Flop w/ 5 Players
158 Seen Flop w/ 6 Players
0 Seen Flop w/ 7 Players
0 Seen Flop w/ 8 Players
0 Seen Flop w/ 9 Players
0 Seen Flop w/ 10 Players

2.78 ASF
0.982 EF1
1.149 EF2
0.976 EF3
110.0 PER
114.2 NEW PER