Note: Not at the old Poker1 site. A version of this entry was first published (2011) in Poker Player newspaper.
Two things barge into our brains when we hear the words “poker” and “rankings” in the same sentence. How do the possible hands rank, based on profit? And how do the players rank, when deciding who’s best?
Fine. We’ll use those two types of rankings as the basis for today’s self-interview. And that means there will be only two questions.
Question 1: What would be a correct list of top 25 hold ’em starting hands?
Actually, the ranks are different depending on whether you’re in a loose or tight game, whether the table is full-handed or short-handed, your image, the traits of your opponents, and more.
Here are my top-25 rankings for fixed-limit games, based on whether you’re facing few, many, or an average number of opponents. The latter (average number of opponents – the “Main” line) is a compromise, addressing all situations. If you don’t have any other information to go by, that’s the list you should use.
In no-limit hold ’em games, the rankings are, strangely, pretty much the same – although I’ve made some modifications that we can talk about another day.
The chart
|
Rank |
Main list |
Many-opponents list |
Few-opponents list |
|
1 |
A-A |
A-A |
A-A |
|
2 |
K-K |
K-K |
K-K |
|
3 |
Q-Q |
Q-Q |
Q-Q |
|
4 |
J-J |
J-J |
J-J |
|
5 |
A-K suited |
A-K suited |
A-K suited |
|
6 |
10-10 |
A-Q suited |
A-K mixed suits |
|
7 |
A-K mixed suits |
10-10 |
10-10 |
|
8 |
A-Q suited |
A-K mixed suits |
A-Q suited |
|
9 |
K-Q suited |
K-Q suited |
A-J suited |
|
10 |
A-J suited |
A-J suited |
A-10 suited |
|
11 |
A-10 suited |
A-10 suited |
A-Q mixed suits |
|
12 |
A-Q mixed suits |
K-J suited |
K-Q suited |
|
13 |
9-9 |
9-9 |
9-9 |
|
14 |
K-J suited |
K-Q mixed suits |
K-J suited |
|
15 |
K-Q mixed suits |
A-Q mixed suits |
A-J mixed suits |
|
16 |
K-10 suited |
Q-J suited |
A-9 suited |
|
17 |
A-9 suited |
K-10 suited |
K-Q mixed suits |
|
18 |
A-J mixed suits |
J-10 suited |
8-8 |
|
19 |
8-8 |
A-9 suited |
A-10 mixed suits |
|
20 |
Q-J suited |
Q-10 suited |
K-J mixed suits |
|
21 |
K-J mixed suits |
8-8 |
K-10 suited |
|
22 |
A-8 suited |
A-J mixed |
A-8 suited |
|
23 |
A-10 mixed suits |
K-J mixed suits |
A-5 suited |
|
24 |
Q-10 suited |
A-8 suited |
A-4 suited |
|
25 |
K-9 suited |
Q-9 suited |
Q-J suited |
No dispute
Notice that the top five hands are not in dispute, no matter which list you choose. It’s only beginning with the sixth-ranking hand that there are differences.
When I first published these guidelines in 1994, I realized that the choices – though based on careful analysis and computer simulation – would seem peculiar. That’s why I wrote this…
Since the “Main” column is a compromise between the other two (“Many” and “Few”), maybe you should look at it first. This would be the column you’d use to settle traditional arguments about which are the top hold ’em hands, assuming no other information is provided.
How come?
How can it be a compromise, you ask, when a pair of tens is shown as 6th in rank on the “Main” list, but 7th on both the “Many” and “Few” lists?
Simple. Peculiar, but simple. Notice that the 6th-ranking hands are different on those latter two lists, and neither one of those was quite as profitable on average as 10-10.
Notice that J-10 suited, a hand once thought to be hold ’em’s most powerful (especially in no-limit), due to its moderately high ranks and excellent chances of converting to a straight or flush, isn’t even on the list of top 25 main-list hands!
From 26th to 42nd
That J-10 suited is actually ranked 26th (just off the chart) on the “Main” list, but only 42nd on the “Few” list.
You might also be curious about why A-4 and A-5 suited are included on the “Few” list, but not the “Many” where you’d think they would have more value, because making a straight is more-often necessary against multiple opponents. Good point, but that theory doesn’t quite compute in practice, because other hands are elevated even more in importance against many opponents. Just so you know, A-5 suited is 27th on the “Main” list and 32nd on the “Many” list.
Can you use the rankings I’ve just presented to resolve arguments? Sure. Why not?
Question 2: Let’s move to the second part of what you said “barges into our brains” when we think of rankings and poker. Are the rankings of poker players accurate?
No.
Well, let me think. Maybe. They’re accurate in that they measure what they claim to measure. For example, both Bluff and Card Player magazines publish poker rankings, tracking players’ tournament progress through the year. Unfortunately, these and all other rankings fail to reflect who the best players are.
First, it’s extremely unlikely that the best players will perform near the top among thousands of competitors in a year. There are too few events and too much luck. But the main problem is that players don’t enter the same number of events.
Not compelled
Take me, for instance. I seldom play more than six events a year. Many years I play only one event or none at all. Why? It’s because I don’t feel compelled to gobble up my life scurrying around the world, entering 300 tournaments each year.
Yet, people judge me by my results. I’m measured by the fact that I’ve only won a few hundred thousand dollars in tournament play, but not on the fact that I have a really good “batting average” for the relatively few events I’ve entered. I’m not alone.
So, what’s the solution? It’s a simple one. Only players who pre-register to be monitored in an event should be included. That way you’d know how many times a player entered an event and failed to make the money. You could get an accurate measurement, and I wouldn’t be penalized for not playing.
Question 3: You said there would only be two questions, but I have another one. If players enter tournaments and don’t finish in the money, how will you know what positions they finished.
You won’t.
But you don’t need to. If they make the money, then you know what their finish was, so you can use that to determine an overall average. But when they fall short of the money, they typically just slink away, and nobody knows the exact place of their elimination. It’s unrecorded. But that’s okay. Then you’d use the mathematical assumption that they finished right in the middle of the non-money players.
So, if there are 600 entrants and 60 places were paid, then we know the identities of only 60 competitors and how they placed. That leaves 540 we don’t know about.
In the interest of fairness, any of those 540 who pre-registered to be in the best-player race and be monitored would be assumed to have finished at the average of the non-money field. That’s 270.5 away from the lowest-paid winner. Since 60 players made the money, 330.5 would be entered for a non-paid performance. You can get the same result by adding 61 to 600 and dividing by two. It would be averaged in with the times that player finished in the money at 7th, 14th, 28th, or whatever.
The point is, players could no longer try to buy their way into the top of the rankings by playing all the events and sometimes rating high on the list when they’re not even covering expenses year after year.
Is mine a perfect solution? No. But it works for me. — MC
I don't understand the overcards (esp unsuited) coming ahead of pairs. For instance, 99 is better than AJ to me. There r2 reasons. 1. The pair is a made hand. Overcards have 6 outs to hit, whereas the pair doesn't have to hit. 2. Overcards are a 55-45 dog in a race, I believe. Now if I am in early position, I would limp an AQ if I thought I could c a cheap flop; the reason being that from low position I would need to hit high cards. It would b harder to hit the pair when there r only 2 left in the deck. &if I have agressive players to my left, I would not even limp the AQ because I wouldn't want to face a raise on an unmade hand. So depending upon the position & the style of the opponents, I would occasionally take the overcards but in most cases, I believe the pairs r better.
Hi, Warren —
Thanks for making your first comment and joining our Poker1 family.
Your arguments make sense and are well presented.
However, the rankings are based on real-world database results and on computer simulation. They tend to be accurate, except that individual play could slide some hands above or below others that are very closely ranked.
Sometimes careful analysis shows results that aren't immediately apparent when "thinking it through" the old-fashioned way. That doesn't mean that customary thought process is less valuable. It just means that when it conflicts with a broader mathematical analysis that has been conducted correctly, it's usually prudent to trust the results and modify the original speculation.
Still, it's interesting to debate, and we learn a lot about poker from these discussions.
By the way, the chart does rank 9-9 above A-J unsuited, so I'm a little confused by the example.
Straight Flushes,
Mike Caro
Thank u Mr. Caro. Actually I joined your website early on and have been posting in your "TALK" section. There u&i had a similar discussion over 77 and AQ off. I thought you wrote 77 was slightly better. I apologize for my bad example here, as I meant the AJ suited shows as higher than 99. I understand your response here to mean to basically trust the numbers. It really goes against the grain for me not to take the made hand but your other chart on betting depending upon how many after u r left to act (in an unraised pot) helped a bit. I look forward to your next chart on playable hands after a pot has been raised. This is a great website. Keep up the great work!
3. J-10 is easy to get away from if you miss the flop big time.
It is #3 that makes playing J-10 attractive to me. You aren’t going to call off a large piece of our stack preflop with it and you can make some really big hands with it. But if you miss, you have garbage. It’s an easy fold post flop.
————–
I was going to say this, but Mike got there first, nice post. I’m thinking at some point we need a real conversation – maybe a book, or something, about “playablity” as it relates to the newfound (relatively new) ability to run computer sims that only tell us if we all stay to the end who wins.
I like your starting hands, but the J 10 suited has consistently won money for me, depending on my position and how I play it, so that is the only one that I would think about ??
Hi, Moon7777 —
Thanks for leaving your first comment and joining our Poker1 family.
Decades ago, J-10 suited was considered to be the most-profitable no-limit hold ’em starting hand. And it is, in fact, a much better-than-average hand.
Only using modern analysis, computer simulation, and probing into millions of hands from online poker databases were we able to put J-10 suited in its place. Unfortunately for it, that meant a large drop in stature.
Straight Flushes,
Mike Caro
It’s a large drop, but that being said:
1. It’s still better than top 50 no matter how you look at it, based on your numbers.
2. Moon qualified it with “position and how I play it”.
3. J-10 is easy to get away from if you miss the flop big time.
It is #3 that makes playing J-10 attractive to me. You aren’t going to call off a large piece of our stack preflop with it and you can make some really big hands with it. But if you miss, you have garbage. It’s an easy fold post flop. Compare that to 10-10 or J-J. There are a wide variety of flops that hand is hard to get away from because you look like you have the best hand but can in fact be drawing thin. In my experience it is hard to win a big pot with J-J or 10-10 unless you flop a set. I’ve lost count of the number of times where I’ve pushed with a garbage flop with J-J (yay, I have an overpair!) and gotten called by a set or Q-Q or better.
Most of the people I know (myself included) have a love-hate relationship with J-J. Yes, it is the 4th best starting hand, but you are close to a coinflip with AK, AQ, KQ. You’re only about 2:1 preflop against A-rag suited. It’s certainly not what I would call a monster in a real game, since once somebody puts in a raise, you really aren’t playing against random hands anymore.
I certainly like the hand ranking charts and knowing where different hands stand, but some people put too much stock in these things.
If I were a gambling man, I would be willing to bet that I’ve made more money with J-10 than J-J because I haven’t lost nearly as much with it. It’s not a matter of losing often. It’s a matter of losing big pots. Now I’m sure that part of this is simply me playing J-J incorrectly, but I would guess that over a lifetime, if people kept track of results down to the hand, they would find a similar outcome.
I wonder if you can probe those millions of online poker hands and see not just how J-J faired on a win vs. loss basis, but if you could tally up the money won vs. lost with J-J. I think that would be interesting to see. Also, would it change limit vs. no-limit? In limit you can at least minimize your loss by check-calling all the way down. Check-calling in NL can mean blowing your whole stack anyway. And J-J is a perfect hand to do that with in many instances. I’m guessing in real life there is a big difference there.
I for one happen to like pocket Jacks…. I do well with them simply because I am able to fold them… I dont get married to jacks like a lot of people I know… they say they hate pocket jacks but when they get them they always play them to the river and lose a bunch… what it all comes down to is the whole picture… chip stacks, players (tight or loose), position, etc… by using all the information and not just a part of it u can make a informed decision and that way u can make or save money with Pocket Jacke…
Very interesting rankings, Mr. Caro. Will we get to see the rankings for NLHE as well?
Yes, James —
Eventually almost all my statistical tables, summaries of research, and more will be right here at Poker1.
The process will take time, though.
Straight Flushes,
Mike Caro
Hey Mike, wouldn’t the average out of the money finish be 330.5, not 270.5? i believe you forgot to had 60-the number of money finishers-to the average out of the money finish
Hi, Stan —
Yes! Absolutely. The 270.5 is the average finish among the group of 540 that failed to make the money. It isn’t the average finish for all tournament entrants. That should be between 61 and 600 — with the average being 330.5, as you correctly state. Thanks for pointing this out. I never noticed. I’m clarifying the text.
Straight Flushes,
Mike Caro
Speaking of hand rankings. I like to play a trick on friends of mine, that are a bit more passive about poker. I will ask them which starting hand is worse, 2-7os or 2-3os? Undoubtedly, they always answer 2-7os. Then I offer them a wager in which I take 2-7, and they take 2-3 and we run it 5 times! : )
Good proposition bet, Lord BK. You’d have about a 2-to-1 advantage, per hand, with the precise advantage depending on which suits were or weren’t duplicated.
However, if you match each of those (one at a time) against random hands — with no ranks guaranteed to be duplicated — and play to the showdown, the difference between 7-2 and 3-2 offsuit is small.
Straight Flushes,
Mike Caro