The wikipedia has a great write up on texas Hold'em statistics. These tables expand on that theory and are very useful for "heads up" play.

The following table shows the winning percentage of a hand vs.a random hand. In my opinion, this table is mainly useful in near heads up situations as people do not in generally play random hands, the play select or premium hands. There are 47,008 possible distinct head-to-head matchups. Jasbo has calculated the exact W/L/T values for all these. The exact ranking of all 169 hands is given below. This table matches up closely with Justin Cases' table in Percentage Hold'Em, which used simulation to estimate the hand values. However, the difference is that Jasbo counted ties as worth 1/2, while Justin's table counted ties as 1.

Note: These numbers do not give indicate how hands match up against each other. To calculate two hands agains each other use the poker calculator here. Rather this table shows how each hand will do independently against a random, unknown hand. For example, K9 Suited (.600) ranks above A5 Off-Suit(.577) in the table: meaning that if you are all-in preflop against a random, unknown hand, you would pick K9s instead of A5 Off-Suit. However, if you have a proposition bet where you can take either Ac5d or Kh9h against each other, then you should take Ac5d, since it is favored against Kh9h. I have taken the time (too much time) to hand color code this chart based on a Green-Yellow-Red gradient. Think of it like a traffic llight. The suited combinations are on top while the unsuited are on bottom. So for example, the odds of winning with K4 suited against a random hand is 54.9%.

Eastbay at sitngo-analyzer.com noted that because people don't normally call big raises with random hands, but with select hands, a new chart and set of numbers was needed. He ran the numbers for the odds of a hand vs a random hand from the set of Any pair or any ace. Again I took way too much time to hand color code the results in the chart below. Suited cards are on top and unsuited are on bottom.

After realizing the overwhelming weakness of A2o-A6o and that many people call with suited Broadway cards, the set was revised to Any Pair, Any Ace Except A6o-A2o, KQs, KJs, KTs or QJs. Eastbay at sitngo-analyzer.com ran the numbers again for every hand vs. a random hand in this set and I took hours to color code the gradient on the same Green-Yellow-Red scale. The results are extremely valuable in determining which hands to push all in with late in SnG. For Example, assuming most people will only call from this set, you are better off pushing with 10-2 suited than A2 offsuite. A very surprising result indeed. Check out the chart: