Poker Hand Groupings

In their book titled "Hold'Em Poker for Advanced Players", Sklansky and Malmuth give a set of 8 groups of hands organized by their success in poker according to their experiences. A couple of sites on the internet offer a regrouping based on various criteria. Here, I summarize two of them:

Hand groupings based on pokerroom.com stats

pokerroom.com provides a list of the expected value for each starting hand here: Starting Hand Expected Value Statistics. This lets us group the starting hands according to their expected win percentages according to their website.


Unsuited
Suited
Groups
Pairs
Connector
1-Gap
2-Gap
3-Gap
>3-Gap
Connector
1-Gap
2-Gap
3-Gap
>3-Gap
1
9-A
A




K, A
A
A


2
7, 8
K
A
A


J, Q
Q, K
K
A
A9s
3
5, 6
Q
K
K
A
T
J
Q
K
A3-8s, K8s
4






9
T


K7s, A2s
5
4
J
Q


A9
7, 8
9
J
Q
K6s
6
3
T
J
Q

A8, K9
5, 6
8
T
J
K5s, K4s, K3s, K3s,
Q7s, Q6s

Interestingly, the only significant difference between these groupings and that of evgr.com is that the pairs are in higher groupings and a couple of the unsuited hands are in lower groupings. I think this is because when you hold a low pair, and hit it, it is very profitable and folding them is easy. The six groupings are fairly different than the other sets of groupings I have seen. The first three groupings are all profitable hands, partitioned into 3 groups with (relatively) significant differences betweeen the groups, but relative homogeneity within the groups. The only exception is the first group, which isn't really a group and can't be treated that way under any system. Group 4 is the set of hands with 0 expected profit. Groups 5 or 6 are negative expected winnings, but some of the hands in group 5 (and to a lesser extent group 6) might be worth playing if you have position, but only on an ideal flop.

Odds for groupings based on pokerroom.com stats


Group
Description
Hands in group
Hands at least as good
as hands in group
Hands in group in top
x% of hands (pre-flop)
Strategy according to Player's Guide
1
Top 11 Hands
64
64
4.83

2
Next 12 Hands
76
140
10.56
Starving
3
Last Profitable Hands
104
244
18.40

4
0 Expected Value
16
260
19.61
Tight Expert
5

66
326
24.58
Professional
6

110
436
32.88
Savvy Gambler






All

1326
1326
100.00
Idiot

Hand groupings of evgr.com

The best regrouping that I have found is on evgr.com. evgr.com is a poker website that has a fairly technical outlook on poker. They have a whole tech section. Rather than going by opinions by long time players, they tested the playability of the top 84 poker hands by doing a huge simulation in 1996. The outcome of that simulation was the Texas Hold 'Em Poker Player's Guide. In addition to the groupings, they also have recommendations about betting according to position and how many other players are in the pot, but this stuff gets very advanced quickly. I have converted their word document into html so that it is more widely viewable. Their guide gives the following table:


Unsuited
Suited
Groups
Pairs
Connector
1-Gap
2-Gap
3-Gap
>3-Gap
Connector
1-Gap
2-Gap
3-Gap
>3-Gap
1
J-A





A




2
10
A




K
A
A


3
9

A



Q
K
K
A

4
8
K

A
A

J
Q


A9s, A8s
5
7
J, Q
Q, K
K


T
J
Q
Q, K
Axs, K8s
6
6




A9, A8
9
T
J

K7s, K6s, K5s
7
5
T
J
Q
K
A7, A5
8
9
T
J
K4s, K3s, K2s,
Q7s, Q6s, Q5s
8
4

T
J
Q
A6, A4, A3, A2,
K8, K7, K6



T
Q4s, Q3s,
J6s, J5s

Every entry in the chart is according to the higher of the two cards. By n-Gap, we mean the number of "empty spaces" between the cards. That is, AQ is a 1 gap, T6 is a 3 gap. For all but the >3-Gap entries, this uniquely specifies the other card. In the >3-Gap columns, both cards are expressed. By x, we mean any card not expressed above it in the column (and still qualifying as having a >3-gap).

The hands which are within the same grouping all have approximately the same odds pre-flop, with obvious differences within a group such as the difference between AA and JJ.

Odds for evgr.com groupings

We get the following table of odds for the evgr.com groupings. Note that I use the definition of a hand as one of the possible 1326 hands whereas evgr.com uses definition of a hand as one of the possible 169 hands.

Group
Hands in group
Hands at least as good as hands in group
Hands in group in top x% of hands (pre-flop)
Strategy according to Player's Guide
1
28
28
2.11

2
30
58
4.37

3
34
92
6.94

4
58
150
11.31
Starving
5
114
264
19.91
Tight Expert
6
54
318
23.98
Professional
7
118
436
32.88
Savvy Gambler
8
146
582
43.89
Good Gambler





All
1326
1326
100.00
Idiot

learn-texas-holdem.com groupings

This chart expresses exactly the same thing as the "Modified Hand Groupings" at http://www.learn-texas-holdem.com/texas-holdem-hand-groupings.htm but in a more digestable version.


Unsuited
Suited
Groups
Pairs
Connector
1-Gap
2-Gap
3-Gap
>3-Gap
Connector
1-Gap
2-Gap
3-Gap
>3-Gap
1
J-A





A




2
10
A




K
A
A


3
9

A



Q
K
K
A

4
8
K

A
A

J
Q


A9s, A8s
5
7
J, Q
Q, K
K


T
J
Q
K
Axs
6
2-6
9, T
J



8-9
9, T
J
Q
Kxs
7

5-8
T
J, Q
K
A9
3-7
4-8
7-10
J


Odds for learn-texas-holdem.com groupings

So, every poker strategy has some set of hands it folds pre flop. The question is, how good is the best hand you automatically fold compared to other possible hands? When I first started playing I was playing all kinds of crap. According to this chart, I was playing hands well below the 50% mark. This website says that an ideal number of bets and calling of bets should be 17%. That means you should only be playing the top five groups here from a purely statistical standpoint.

Group
Hands in group
Hands at least as good as hands in group
Hands in group in top x% of hands (pre-flop)
1
28
28
2.11
2
30
58
4.37
3
34
92
6.94
4
58
150
11.31
5
106
256
19.30
6
122
378
28.51
7
168
546
41.18
8
780
1326
100.00




All
1326
1326
100.00

I am pretty sure that as the number players and the ratio of blinds to average pots vary, the hands that you want to be playing also changes. For example, in a game with $0 blinds, you should only ever play pairs of aces. Because there is no penalty in waiting for them. In a game where the small blind is 1/10 of the total chips at the table, you should be playing at least the top 7 groups, if not some of the 8th, because the penalty of waiting is enormous. I would assume that the stats from the above site refers to a typical game on the internet or in a casino with more than 6 players and blinds that are a small fraction of the average stack.

The Point System

All this can be difficult to remember when a poker player is beginning. So, I have developed a rudimentary point system for determining the group of a hand, similar to the point system in bridge. This system will only work for non-pair starting hands. For pairs, you will just have to memorize them. Begin with the following values based on the highest card:

High Card
Points
A
1
K
2
Q
3
J
4
T
5
9
6
8
7

Then, add the following number of points on the following conditions:

Condition
Point Addition
Unsuited
+1
1-Gap
+1
2-Gap
+2
3-Gap
+3
>3-Gap Kicker is 7-9
+4
>3-Gap Kicker is 2-6
+5

Which gives us the following table:


Unsuited
Suited
Groups
Pairs
Connector
1-Gap
2-Gap
3-Gap
>3-Gap
Connector
1-Gap
2-Gap
3-Gap
>3-Gap
1
J-A





A




2
10
A




K
A



3
9
K
A



Q
K
A


4
8
Q
K
A


J
Q
K
A

5
7
J
Q
K
A

T
J
Q
K
Ax
6
2-6
T
J
Q
K
Ax
9
T
J
Q
Kx
7

9
T
J
Q
Kx
8
9
T
J
Qx
8

8
9
T
J
Qx

8
9
T
Jx

As you can see, the system is not perfect, but it makes for some simple rules. There are no hands that are mis-categorized by more than 1 group (assuming the addition of a 9th group for all other hands). The one place where this method doesn't do very well is distinguishing between group 8 and group 9. Here are some rules if you want to play group 8 hands, but not anything worse:

Here are my comments about distinguishing between group 8 and group 9: The only hand that is playable that doesn't have a shot at a straight or a flush is A9, cards below 10 are category 8 unless they are a suited connector. For the most part, if you don't have a face card or a suited connector, you've got group 8.