The following was sent to me by a friend who called it the "Hutchinson Omaha Point System", a system for evaluating starting hands in Omaha. Someone on 2+2 asked me for it so I thought it would be worth posting.


Hutchinson Omaha Point System


The purpose of this system is to provide a simple means of evaluating starting hands in Omaha poker. It was developed in several steps:

First, Mike Caro's Poker Probe software was used to determine the win percentage for various four card combinations when played against nine opponents. This was accomplished via a Monte-Carlo type simulation with a
minimum of 50,000 hands being dealt for each starting hand. The assumption made in this type of simulation is that each hand is played to the finish.
This is, of course, an unreasonable expectation, but , in the absence of detailed knowledge of each player's starting requirements, method of play, etc., it is the best means of approximating a hand's strength and earning potential.

Secondly, a number of components were examined in an effort to determine their relative contribution to the value of each starting hand. Eventually, it was decided that the primary determinants of good Omaha starting hands related to the rank of the cards and whether or not they were paired, suited, or connected.

Finally, a type of regression analysis was conducted to try and determine the relative weighting of each of these factors. The system that follows is the result of quantifying the contribution made by each of these various
components.

Once the calculations are made, the resultant point total, WHEN DIVIDED BY TWO, is an approximation of the actual win percentage for a particular hand--when played to the finish against nine opponents. The correlation between point totals and win percentages, while not representing a one-to-one correspondence is, nevertheless, quite high. In fact, in about 70% of the cases the actual win percentage will be within just one point of the total points awarded by this system. This means that if the system indicates that a given hand earns, say, 40 points, you can be quite confident that the actual win percentage for this hand is between 19 and 21 points. It is very likely to win more often than a hand with 38 points and almost certain to outperform a hand with 36 points.


STEPS IN CALCULATING POINT TOTALS

FIRST, to evaluate the contribution made by suited cards, look to see if your hand contains two or more cards of the same suit. If it does, award points based upon the rank of the highest card. Repeat the procedure if your hand is double suited.

If the highest card is an ACE award 8 points
If the highest card is a KING award 6 points
If the highest card is a QUEEN award 5 points
If the highest card is a JACK award 4 points
If the highest card is a TEN or a NINE award 3 points
If the highest card is an EIGHT award 2 points
If the highest card is SEVEN or below award 1 point.
If your hand contains more than two cards of the same suit, deduct 2 points.

SECOND, to factor in the advantage of having pairs,

If you have a pair of ACES award 18 points
If you have a pair of KINGS award 16 points
If you have a pair of QUEENS award 14 points
If you have a pair of JACKS award 13 points
If you have a pair of TENS award 12 points
If you have a pair of NINES award 10 points
If you have a pair of EIGHTS award 8 points
If you have a pair of SEVENS or below award 7 points

Award no points to any hand that contains three of the same rank.

THIRD, when your hand contains cards capable of completing a straight it becomes more valuable. Therefore, If your cards contain no more than a three card gap, add the following points:

For FOUR cards, add 25 points
For THREE cards, add 18 points
For TWO cards, add 8 points

From these totals, subtract two points for each gap, up to a maximum of six points.

To account for the special case represented by ACES, deduct four points from the above totals when an Ace is used. This is necessary because an Ace can make fewer straights. However, when your hand contains small cards that can be used with an Ace to make a straight, the hand's value increases. Therefore, when your hand contains an Ace and another wheel card, add 6 points. Add 12 points for an Ace and two wheel cards.

FINALLY, a determination must be made as to which hands qualify as playable. This becomes a function of how many points one decides are necessary before entering a hand. My suggestion would be to only play hands that earn 28
points or more. It can be argued that, ignoring the rake, any hand with more than a 10 percent win rate (i.e., those with 20 points or more) is potentially profitable in the long run. Still, I have the prejudice that most players, and especially those who are relatively inexperienced, would
be better advised to forsake marginal hands and to focus on those that earn 28 points or more. Recalling that a random hand will win about 10% of the time in a ten-handed game, it can be seen that playing only premium
combinations of 28 points or more insures that you will always have a hand that is 40% better than a random hand. The total required to raise or to call someone's raise must also be determined subjectively. I feel that 32 points is the appropriate level, so, in summary,

YOU SHOULD CALL WITH 28 POINTS OR MORE AND CONSIDER RAISING WITH 32
POINTS OR MORE

SOME EXAMPLES FOR CLARIFICATION

The hand that has the highest win percentage in Omaha contains two ACES and two KINGS and is double suited. A hand containing the AS, KS, AH, and KH would earn 54 points under this system--calculated as follows: under step
one above, the two double suits headed by the two aces earn 8 points each for a total of 16 points; step two awards 18 points for the pair of aces and 16 points for the pair of kings, or a total of 34 more points; under step three, the ace-king combination earns 4 points for its straight potential. (NOTE: The two consecutive cards earn 8 points but a deduction of 4 points is made because one of the cards is an Ace.) The resultant total of 54 points, when divided by two, closely parallels the actual win percentage for the hand which is about 26.65.

Assume you have the 9S, 8S, 9D, and 8D. Step one awards a total of 6 points for the two double suits headed by nines. Under step two, the pair of nines earns 10 points and the pair of eights earns 8 points. The last step awards 8 points for the 9-8 combination. The total of 32 points, when divided by two, is the same as this hand's actual win rate of 16 per cent.

With the QS, QD,9H, and 9C, no points are earned under step one as there are no suited cards. Step two gives 14 points for the pair of queens and 10 points for the pair of eights. Step three awards 8 points for the Q-9 combination but then calls for a deduction of 4 points because of the two card gap that exists between the two cards. The final total is 28 points and, when divided by two, it again closely reflects the actual win percentage for this hand which is 14.5%.

An example of a hand that tends to be somewhat over-rated by novice players is AS, KD, QH, and TS. Under step one the hand receives 8 points for the suited ace and ten. Step two is disregarded as the hand does not contain any pairs. Step three awards 23 points for the straight potential of the four connected cards. The final total is only 31 points, making this a marginally playable hand. It actually wins about 16.2%.


Finally, consider AS, 3S, KD, 4D. Step one awards 14 points, step two awards none, and step three grants 12 points for the A-3-4 combination and 4 points for the A-K combination. This total of 30 points corresponds with
the actual win rate of 15%.

Three things. 1) Thank you for posting it so people can have access. 2) This system will likely be unusable do to the speed of online play and the failure to account for position. 3) As you note, the system is meant for 9 handed to the river play. The online game is now 6-max.

8877 ds is a raising hand, but this system asks you to fold it as it only has 27 points.

This is a ridiculous and useless thing. For full ring just play hands that make nuts, no?

I love description of AsKdQhTs as "marginally playable"

Me no likes mwette. Too complicated and rigid.

Actually, I'm kinda curious about a "reverse hutchinson system"--Instead of coming up with points based on cards, I want to infer hand ranges based on poker tracker stats.

For example, suppose I come across an opponent that I have in my database, but haven't really observed first-hand. He seems to be a winning 20/6 (FR) player.
What types of inferences can we make about what hands he's raising with?
Suppose I run across a good 40/10 player. What additional (4% more) hand types can we guess this person is raising with?


Are there stats published anywhere that has info on the probabilities of being dealt certain Omaha hands types?

I'ts not my stuff. I just "quoted" from another source.

There have been recent posts about the probabilities, one obvious one would be the Omaha page on Wikipedia.

As for their raising ranges, if you know some SQL and have a large enough PT database history with the player that you feel it is accurate enough to your liking, you can at least get the hand ranges where he went to a showdown.
That will give you a subset of the hands he is raising with (given that more of those will be seeing showdowns for the most part) and from there you can probably expand it yourself a bit since you can see the hands they are considering.

You will want to take position into consideration as well, which shouldn't be a problem as that is tracked in the db.

This system is very interesting, but I think that you should rather think how often you hit the flop instead of thinking how often you win after river in all-in alike situation. Most of the moves and decisions comes up on the flop, so...

And finally IMHO pre-flop raising isn't about getting 16% winrate at all, because you don't make most of the money pre-flop and you will be called too often. It depends much more about stack sizes and such.

What you guys think?

once you play long enough you'll know what's good and what's not and the decisions will come instinctively
If you're using mental energy to "tally" your hand every hand you'll slow the game down and every1 will quit, sessions won't be fun and you'll probably burn out with these decisions rather than actual poker decisions.
Also you'll probably make a bad fold a very good % of the time (especially 6 max)

This System is an excuse not to learn the basics of hand evaluation.

Ya, I'm raising and reraising like anywhere ever.

bump this and screw you guys...this is usable if you are intelligent.

To clarify, I did not and would not read the analysis

People are always looking for a system they can follow to be winners and that's why they'll always be losers.

i mean this system is meant for FR so I would assume most hands w/ an ace in them just aren't worth re-raising pre. I'd assume you can just lower the point totals you need for 6max so that AKQTss is in your re-raising range. the system seems like it could be pretty useful to compare hands against each other more than to actually tell you which specific hands to play.

Short and to the point!

Did you even read the post in its entirety.

This is an elementary preflop starting hands guide and has nothing to do with
hand evaluation.

Its easy to reject math, because its hard to do. But its math you do in your head when you evaluate hands. The difference is that some are good at it and some are not.

I see this as an interesting line of thought. raw maybe, but a reasonable starting point for some off the table study.

Well the system is a basic hand evaluation system actually, and probably a reasonable one for a newbie since its objectively grounded and prevents them from making mistakes by constraining their choices, which is important for an beginner.

Obviously a more experienced player should deviate a bit from this - it ignores stack sizes/position/postflop playability on hand selection, which is why 8899 is a fold under this system but actually playable, and you can raise much more junk from LP.

I think for a newbie, the system is fine, its like training wheels to keep them from falling over until they learn what they do on their own through experience, but is of limited use for most people on this board.

I think the system is also decent as its a reasonable way to think about hands - different things (rundowns, nutflush, pairs) are like little income streams and if those streams add up enough to be profitable then you play the hand. I certainly go through a somewhat similar (but quicker) process.

[/] system sucks

This.

But on a more realistic note: before I found Jeff Hwang's book (all pause for the choir of angels singing at mention of his name....) This guide did help me out a lot in the conversion from NLHE to PLO. A year later, I'll be the first to admit it does have major flaws but is a horribly valuable guide for any new scrublet that has basic questions.

PS: i'm currently on the upside of a major downswing, but am running 28.4/5 and am winning consistantly enough to mention it (even now with the 12 buy downswing over and accounted for, I'm at about 19BB/100, was at 35bb/100 at .02/.05 tables). However, the 5 is deceptive because at the scrub stakes I'm at, there is almost always someone who has raised it already PF, so depending on my starting hand, I'm likely to just call most of the time.

lol 12 buyins is not a downswing...esp not a major one, 12 buyins is a farily bad session. How fin good do ppl run?

lol... Fair enough, but when you're in a bet with someone to turn 20$ into 300$ and 12 buyins was a massive portion of the bankroll... it suddenly does turn into a downswing