Which hands do you NEED?

[Duke]

The point of this post, and possibly thread, is to ask what hands you really need to show a profit with over time to be a winning player?

Why do I ask? Well, I play a lot of hands, and after 7700 of my most recent hands in a certain limit game on a certain site, I've had kings 39 times and I've won once. I'm just barely ahead over that time, and I'm down over a grand with my KK hands. But I'm still winning, albeit barely.

So... what hands does one really need to show a profit with long term?

I don't want this to turn into a discussion of how I play the kings, since I don't need to defend my play. I'm showing profit with all other pairs down to 22. If even one of the hands were questionable I might consider posting it. But... if someone always lost with Kings what sort of earn rate should they expect in the long run? What if Aces were a negative hand in your logbook?

It's not like I can stop playing hand #2 - I know it's just a short term anomaly, but it got me thinking. I once read that in certain games one could show a profit playing only AA, KK, and AKs if the players were retarded enough. But I'm sure that that would be impossible if instead of winning X amount with KK, you were a net loser of X/3 or so.

So, once again... what hands do you NEED to win with?

~D

[symphonic]

Pocket Aces, every other hand. /images/graemlins/tongue.gif

[Nottom]

After 17k hands, my winnings break down something like this.

Hand -- % of total profit
AA - 35.4%
QQ - 30.7%
KK - 25.3%
JJ - 16.7%
KQo - 16.5%
99 - 15.7%
AKo - 13.1%

Some quick math will show this is over 150%

[Duke]

We must play a little differently, since in my 43,000 hand database the kings are a slight winner (they hold up their fair share of the time in bigger games and smaller games, but are obviously held back because of that "bad" limit), Aces are very good, but I show a definite profit with all pairs, AK, AQ, KQ, QJ, 78s, 89s.

~D

[Nottom]

I tend to think that my "liberal" raising standards with lesser hands help me get payed off quite well on my big ones to make up for any lost EV.

[DonWaade]

Hey Nottom or anyone else, how do you keep a log and is it just for online playing? I am curious on how I would do it? Suggestions. . . .

[Duke]

Depending on the game you're talking about, I may disagree. If it's anything below 10-20 on Party, the players are so bad that you don't have to mix up your play.

Like... I agree with the concept completely, and you're right. But in those games (if they're the games you're talking about), you don't get any value from mixing the play up. They're TFR's in general.

~D

[Duke]

As you can see, I play ubertight, and I'm not hurting for people to chase me down with any 2 when I got hand #2. And that's all I can really ask for out of a fish.

~D

[Duke]

Just for online playing for me, and I use PokerTracker.

www.pokertracker.com (http://www.pokertracker.com)

Aside from it being depressing that kings win 1/39 times, it's good to see what hands you're doing well with, and it makes it easier to step through hands that you might question your play on. You might look for collusion.

The only bad thing is when, like today, Party decides to never send a hand history that you request, so I'm SOL.

~D

[DonWaade]

TYVM

[daryn]

i think it has been said that a good winning player will make most of his profit from his big hands, AA, KK, etc.. and just break even on the rest overall.

it's the big hands that make the difference i think

[Sheriff Fatman]

Interesting that you should mention this as I was thinking something along these lines today after looking at my PT stats.

AA and KK account for 153% of my net profit after 12,000 tracked hands, which means that the other 167 hands are net losers! The next biggest winner (strangely 77!) wins only a third as much as AA and KK which are about even (Kings are slightly better at present).

Was surprised to see QQ way down in 21st place but this looks to be due to the games being at different limits as I've been building bankroll and then moving up. At the highest limit (3/6 at present) its currently a net loser but the sample is small only 12 hands. When I rank by BB/hand it falls into 3rd place where I'd expect it.

In terms of BB/hand, AA is 2.75, KK is 2.73 and QQ is 1.18 so the difference is significant.

Quite a surprising statistic to see in black and white!

[Nate tha' Great]

[ QUOTE ]
After 17k hands, my winnings break down something like this.

Hand -- % of total profit
AA - 35.4%
QQ - 30.7%
KK - 25.3%
JJ - 16.7%
KQo - 16.5%
99 - 15.7%
AKo - 13.1%

Some quick math will show this is over 150%

[/ QUOTE ]

Just a nitpick, but I think a more useful benchmark is "gross" profit, e.g. profit net of your blinds.

For example, if you play 10-handed 10/20, then it will cost you $1.50 to play each hand ($15 in blinds per orbit divided by ten).

Let's say that you've played a stretch of 10,000 hands and made a total of $4,000, e.g. 2 BB per 100 hands. You've also paid $15,000 in blinds to play those hands, so your gross profit is $19,000.

If you've been dealt KK 40 times during that stretch and pokertracker shows that you've made $1,000 from it, you'd calculate the percentage of profit that it was responsible for thusly:

$1,000 + (40 x $1.50) = $1,060 (that is, you also pay to play KK, even though it's a trivial amount compared to your earnings)

divided by $19,000 =

5.6% of your gross profit

Which is quite nice, but much different than what you get if you divide your net profit from KK by your net profit after subtracting the blinds.

By doing it this way, you can avoid the conundrum that Nottom gets into when he says that hands X, Y, and Z are responsible for 150% of his profit. It also avoids perpetuating the (incorrect, IMHO) idea that limit h/e is more or less about about catching big hands and knowing how to play them right, since groups of hands like suited connectors will be recognized as providing quite a bit of marginal profit to a skilled player if you evaluate them in a way that recognizes that you *do* pay for the privilege of being dealt two cards.

[Mike]

Standard disclaimers apply

Just on the point of your KK question, if you are playing card loose players with lower standards you should not expect your big pairs to perform as they would against a different type of player. KK winning once in forty times should be no more of a surpise than say AXs winning more than it really should because of loose players.

It seems possible a database of hands that gives a reasonable indicator for a tight game, would not give the same reliable indicator for a rather loose game because the hands in a loose game are tested against more random hands than the same hands in a tight game.

For your final question guessing your time is spent in loose games, look to Omahaha - where straights and flushes rule.

[Zele]

Why would you ever voluntarily enter a pot with a hand that doesn't show a long-term profit?

Also I think your data sample is far too small. Even though 38:1 may seem extraordinary, there is a selection bias since you did not go into the 7700 hands wondering if KK specifically would not show a profit.

P.S.: Even your win rate figure is suspect after only 7700 hands. Example: a typical SD is 10BB/40 hands. That translates into a 10*sqrt(7700/40) = 138 BB SD overall. If you're making 1BB/40hands (i.e. expert-level win rate), you'd expect a win of 192 BB. But within 1 SD, you could be up only 54BB, i.e an observed win rate of 0.28BB/40hands. And 1 SD isn't a whole lot, even if you assume normality has kicked in.

[fluff]

Losing with KK 38 out of 39 times is strange. I'm not sure but maybe someone can estimate how likely this is statistically?

KK is my 2nd largest profit maker right behind AA in Party 0.5/1 games, which are easily the loosest thing outside of play money. KK holds up 66.7% of the time in my database (10,000+ hands now).

[Mike]

Hmmm, maybe its how the two of you play KK - and your other hands of course?

[illunious]

Nottom,

Would you please post or PM me your stats on KQo? Also, would you tell me what kind of table conditions/positions/previous action you play it in? It's not a winning hand for me at $.50/$1.

After 113 times:
BB/Hand: -0.1
VP$IP: 34%
Won WSF: 18%

Thanks.

[Duke]

[ QUOTE ]

Why would you ever voluntarily enter a pot with a hand that doesn't show a long-term profit?


[/ QUOTE ]

I don't. The hands that are net losers are negative mainly due to folding them in the blinds, and then never playing them "with position", since i don't expect that hand to perform well.

[ QUOTE ]
P.S.: Even your win rate figure is suspect after only 7700 hands

[/ QUOTE ]

Who quoted any win rate, aside from losing a grand overall on specifically kings. If you mean that a win rate is still suspect if it's only supported by 7700 hands, then yes I agree.

And over 43k hands total, you'd be surprised at my win rate, despite the underperforming kings.

~D

[Duke]

All I'm expecting is that KK holds up aganist N opponents over time just as much as it mathematically should. Given that a typical hand of online poker sees 3 or 4 to showdown, after 5 or 6 see the flop, I figure my kings should work out about 1/3 of the time or better.

Even if you play them in a braindead manner, a hand that strong sort of has to show a profit... or so I thought.

~D

[Duke]

If you just take the hand to showdown against 9 opponents every hand you should win more than 1/39 times. That's why I didn't want to get into that... it's not a matter of how they play. The only question I was trying to pose here was more along the lines of... "If you are doomed by the poker gods to lose with certain strong hand: X, every time you play it, can you still beat the game?"

I have an $8 profit that says I can do it with kings being a great big net loser.

However, to steal a line from Something About Mary... "I work with retards."

~D

[Nottom]

After 164 times:
BB/Hand: 0.64
VP$IP: 81%
Won WSF: 47%

I know KQ is out of place, I have no real reason to think I'm not just running good with it. Its also only my 12th most profitable hand but is up there as a money earner just because I've had it 164 times.

As for how I play it, I raise it pretty much all the time from any position outside the blinds unless there are already a bunch of limpers. If the table isn't right to raise KQ from UTG, then I find another table.

[Bozeman]

I believe our Pooh-Bah David said that in a tough game, a winning player would be about breakeven if he was never dealt AA.

Craig

[Monkeyslacks]

1/39 huh? The poker gods obviously require a sacrifice. Fold your kings pre-flop as soon as possible. While doing so, visual your opponents like this: http://www.nonsenselit.org/Lear/MN/nr_f.gif

[Duke]

That picture is great!
~D

[RydenStoompala]

39 trips to the plate with pocket cowboys and won once? Wow. You really are a statistical anomoly. The correct answer to your question, which hands do you need to be profitable, is "it depends." Since you are getting clobbered with one of the top three hands in hold'em, I would say you need to start raising with 7-2 o and see if the stats baloon in the other direction.

[Zele]

[ QUOTE ]
Who quoted any win rate, aside from losing a grand overall on specifically kings. If you mean that a win rate is still suspect if it's only supported by 7700 hands, then yes I agree.

And over 43k hands total, you'd be surprised at my win rate, despite the underperforming kings.


[/ QUOTE ]

That's what I meant. Win rate is one of the most general measures of performance that you have, and even that is nowhere near precise after only 7700 hands. So data about how specific hands perform after so few hands is basically useless. I don't think you should even be thinking about this KK issue.

Why would I be surprised? I wasn't trying to imply anything about your play, only the irrelevance of this particular statistic. Unless your rate is less than -20 or more than +4 BB/40 hands, I wouldn't be surprised.

[Duke]

[ QUOTE ]

Why would I be surprised?

[/ QUOTE ]

It's not negative. /images/graemlins/smile.gif

~D

[BaronVonCP]

Shouldn't you want to show a profit with every hand you play?

[Duke]

[ QUOTE ]
Shouldn't you want to show a profit with every hand you play?

[/ QUOTE ]

The way my stats are stored in PokerTracker, no.

See, if I'm playing 10-20, and I get 72o in the BB and the pot is raised, I likely fold. That goes in as 1 time with 72o and a net loss of $10.

When is that $10 made up? Never. I never voluntarily enter the pot with that hand. I'm doomed to "lose" money with it over time.

The fact that I'm showing an earn with every pair is likely a short term anomaly as well. For instance... 33o (yeah I kinow it HAS to be unsuited) isn't much of a hand, and I fold it a lot in the blinds for a double bet, but since the situations to play the hand also come up in position with the right odds and dynamic, I'm able to recoup those losses.

I don't look for spots to make up what I'm losing with 72o in the blinds - at least not with the same hand.

I hope this makes it clear.

As a related note... I don't make money in the blinds. I recoup a very high percentage of the dead money I put out there, but I do not show a net earn. This is possible since I actually do play some hands from the blinds, and I show earns with the hands that I am playing there to counter the dead money aspect of it.

But 72o (among other hands) is destined to lose over time.

Something else that won't affect you until you move into higher limits is that you need to mix up your play at times against good opponents to buy action on your real hands. There is so much that goes into this that it's worthy of not only it's own thread, but possibly it's own book. This is unnecessary at the lower limits, though. The basic idea is that you give up a couple bucks with 67s to make a few bucks more with AA and KK. That's more than overly simplistic, but it's another reason why players would voluntarily enter pots with a long term loser of a hand.

If you're a rock, and your opponents are Ted Forrest, Phil Ivey, and Danny Robison in the big stud game at the Hustler... they'll always know exactly where you're at. If you never play KT8 or a semi-dead hand, they will eat you alive, since they'll be able to put you on exactly what you have when you're in there trying to win back the last 10 antes you lost. And when they can play as if they know the cards that you are holding... we all know what David says about that in TOP, don't we class?

~D