Got into a debate last night about ranking low hands. A player understands how we rank low hands. He just thinks the reasoning is flawed. Hang in there with me, and I will try to express the opinion his side of the debate.

This player maintains that low hands should be ranked from the lowest card up. He says that since we don't consider the lowest card when we rank high hands, we shouldn't consider the high card when we rank low hands. So in his mind, an A2347 should beat an A3456.

I can see his point, from that perspective. I'll grant him that low is not the mirror image of high. If poker had first been played as low, maybe his approach would be used. But it wasn't. It was added as an alternative.

I told him that another way of thinking about low is that the better of two low hands is the worse high hand, excluding the dual rank of ace as both highest and lowest hand. 7432A wins over a 6543A in high, so it loses in low.

He thinks this is inconsistent. I told him that if we applied his rules, then an A2348 would beat an A3457 for both high and low. How consistent is that?

Uh... no.

We're not playing who can collect the most low cards, we're playing who can make the lowest hand.

We most certainly do consider the low cards in high hands. AAKTJ beats AAK98.

This.

In low-hand games, hands are ranked exactly the same way as in high card games, with the one exception that aces are low instead of high. Then, based on standard rankings, the worst hand wins.

In a high-hand game, 7-high beats 6-high, regardless of the other cards. Therefore, in a low-hand game, 6-high always beats 7-high.

Think you meant AAKJT beats AAK98

I don't think you're out of the box yet. Why wouldn't AKQ98 lose to AQJT9 in a high game? AQJT9 has no card lower than 9. AKQ98 has an 8, so it is not as high. Why don't we start with the lowest in a high game if we start with the highest in a low game?

Again, just playing Devil's advocate.

Let me re-this this. This is indeed how we score low hands. We don't assign some numeric value to each card and then score the total hand, giving the win to the lowest score. If we did that, in high A3456 would lose to KQJT8.

I get what you're saying, and it's an interesting observation.

KQ432 is a higher hand than KJT98, so why isn't A2JQK a better hand than A3456....

e.g. AKxxx beats AQxxx, so why doesn't A2xxx beat any A3xxx

...interesting observation aisde, you're evaluating them backwards.

A wheel is 5432A, not A2345.

While I think you are following me, the above is a jump back into the box. I am not evaluating them this way. He is saying that is the way it should be evaluated.

He was really just arguing for argument's sake. I think he enjoys stirring the pot. And I enjoy rising to the challenge. But once you get his point, you have to come up with a good justification. It was interesting enough that I thought I'd share it.

Ask your friend why an A-high flush beats a 7-high flush. The 7-high is much more rare so it should win. Then watch his head explode.

I originally thought that avoiding situations where the same 5 cards could win high and low might be a justification, but that possibility exists under current rules because of the typical low-hand exceptions (A is low, straights and flushes don't count).

More likely, I think the best argument for doing it the way we do it is consistency. With the above exceptions, low-hands are evaluated just like high-hands, and the worst hand wins.

I don't have a better answer.

His argument (AKxxx>AQxxx, so A2xxx should be > A3xxx) makes sense in a bizarre way. I get his argument, but...

But if you were to list all possible 5-card hands evaluated as normal "non-low" poker hands were (a) straights don't count, (b) suits are irrelevant, and (c) aces were low, you'd have a bunch of hands listed that looked something like...

KKKKQ
KKKKJ
KKKKT
....
....
7432A
65432
6543A
6542A
6532A
6432A
5432A

The ones on the bottom are the lowest hands.

The reasoning is not supposed to be analogous to normal poker. The lo ranking is designed for playability of the game.

Consider, if lo-hands were ranked bottom up: Then if you have A2x in razz you're usually going to the river. There's little play beyond hand selection.

Whereas, under standard rules, even A23 in Razz is not a sure thing and you have to have some hand reading skills to play it well.


The same idea applies to Kansas City and Badugi rankings.

Exactly. Razz rules rule, here.

Did he accept your reasoning, or not?
Or, is he a fan of that "2nd best" game? :P

I guess he doesn't care that a hand and win both, or lose both.

Maybe I'll tell him we're going to play according to his rules, but for 2nd best. Then we're back to normal outcomes.

But as I stated above, that's not a valid argument. The same hand can already win both high and low due to the low-hand exceptions (A low, no straights or flushes). Take 5432A suited, for example; straight flush for the high and nut low.

I still think consistency has a little bit to do with it, but dismal's playability argument is the best answer I've seen so far.

I don't see how anyone can argue with:

"This is a list of all hands possible from high to low. The ones on the bottom are the best low hands."

Yes, I think low was seen as an alternative to the original high game. As such, the opposite hand would win low that would win high (giving the nod to Schmendo that some hands can win both). It was designed to have that outcome.

If poker were originally play as low, it's quite possible that A2378 would beat A3456. (Pali, in this alternate history you would likely put them in this order, to make scoring easier). Who knows? And maybe we would score the alternative high game to be the opposite of that.

Anyway, I told my friend to go change the rules of poker around the world, then come back and submit the change to our game. Then we'll definitely listen.

Maybe they should change the name of the games. Instead of Omaha High/Low, we'll call it Omaha Best/Worst. Then he won't complain.

You're right, after I posted I realized it wasn't the same.

The problem is that it's a false problem. It sounds good but it makes no sense. He's setting up a comparison on things that don't compare. It's like he's complaining you can't pinch hit while playing defense in football, and we're sitting here discussing the differences between offense and defense as a way to explain it. He's focusing on a detail to distract from the broken construct.

Lowball is LOWBALL. It's the lowest hand. And it's very very old, ancient here in California. Used to be the only game we could legally play.

It's not about kickers, it's about hand rankings. You can't compare it to five unpaired high cards because that's not the end of the ranking spectrum in that direction. A more logical highball comparison would be claiming AAK84 beats 77665 because Aces are a higher pair. Ask him to explain that, and he has the answer to his own question.

I'm all for thinking outside of boxes, but not on something that's based on an argument that compares the wrong things. It's smoke and mirrors for the sake of being cute. Your friend has pulled a sleight of hand on you.

There's nothing wrong with your list. It tells us what the lowest hands are, but it doesn't answer the original question: Why are they ranked that way?

If you don't know why 7 is lower than 8, do you have anything you're selling? Let's haggle... I'll cut you a deal you won't believe!

The original rank of cards, from highest to lowest was A23456789TJQK

Why is the alphabet in the order it's in? Is it because of that song?

Still in the box, Kitty.

Really? Aside from the fact that we're talking poker, not card ranks, do you have a source you can cite?