Discounted outs

[PantsOnFire]

I's like to discuss the concept of discounted outs.

A clear situation where you might discount outs is when you have an OESD on the flop but there are two suits on board. If you knew your opponent had a flush draw you would discount your outs from 8 to 6. If you didn't know your opponent had a flush draw you would have to assign a probability to him having one such as 50%. You could then discount your outs from 8 to 7.

So I'd like to ask about some other examples. One is where you have a flush draw but your cards are say 76s. Thus, there are other two card suited combinations that would result in a better flush than yours. So you can't really give yourself a full 9 outs.

In the example of a straight, if you have 76 and the board is K89, you have 8 non-discounted outs. However, a 10 on the turn might make somebody a higher straight with a hand like J7 or QJ. Further to that, there is the possibility of a backdoor flush hitting or the board pairing on the river filling up a boat.

There are tons of other examples like when you have a flush draw and an overcard. The overcard might be 3 extra outs or it might not be.

So how do you approach discounting outs? Or do you only really assess and apply it to close decisions? Is this more or less important in NL as compared to limit holdem?

[PantsOnFire]

OK, this is bascially a bring out post.

I'm not sure why there are no responses here. Is it in the wrong forum? No interest? Stupid question?

???

[AKQJ10]

No, it's a fine question.... I think I read it, meant to reply, but felt like you were asking so much and I couldn't do the topic justice so I procrastinated.

So I'll just answer one aspect. I think you need to discount outs in both NLHE and LHE, but for different reasons. In LHE (limit poker, really) you're often making very marginal decisions where your estimates will shade a decision one way or the other. In NLHE, I suppose the decisions are often marginal too, but they're made such by the huge implied odds. If you make a flush that turns out to be no good, you're probably going to lose a lot on that hand.

That said, I don't think you should discount the outs to a small flush very much unless the board's paired. Yes, you occasionally see flush vs. flush, and a lot of money changes hands, but most of the time in HE with four flush outs among the two hands and two on the board, you'll just have a draw that doesn't go anywhere.

[mvdgaag]

I agree with AKQJ10.

I like to add that I think it is better to think in terms of the difficultness and profitability of decisions in later streets: "I can catch the Ts, but it would also make a flush possible, so I how do I proceed when it happenes?" It is still a fine out, just one you don't like as much as others, because you might lose because of it and therefore might not want a too big pot when you catch it. So if you don't loose you've probably been smart and win a smaller pot than with the other outs that wouldn't have made a flush possible. In other words your implied odds are a lot smaller with these outs and you might have to fold to a bet and sometimes lose.

Backdoor flushes and straights are something you rarely encounter against players that had big flop desicions. There are very few hands possible that are worth calling a descent sized bet AND have a backdoor draw. It still happens though.

A pair on the board is mostly bad (unless it helped you to a fullhouse). Any twopair or set might have gotten a fullhouse. Rewind the hand and think how likely it is your opponent hit a set or two pair before the board paired. I would defend my hand against possible redraws, but I won't let them affect my choices too much. You don't want to be outguessing yourself.

If there has been betting I assume someone hit the flop and my overcard is more likely to be good, unless I have reason to believe it is someone's kicker. For example a preflop reraise and a K on the board could very well mean AK. If you are there with AQs and a flushdraw don't count your ace as overcard. It's a matter of judgement.

About the flush VS flush: The chances are really slim. Of the 13 cards there are 8 left that could be dealt to your opponent if you already have the flush. So the chance is 8/47 * 7/46 * 100% = 2.6% chance for your opponent to have been dealth two cards of your suit (if so, the lower flush could also be in his hands, depending on your cards). You can up this estimation a bit on later streets, because he's still there. On the other hand, would he have played Ts3s? There are many combinations he wouldn't have chosen to play preflop... Anyways, he's far more likely on some other hand/draw. That's why you can play small suited connecting cards for the flush as well. If someone has the bigger flush you probably have to pay him off, but most of the time he won't.

Just my ideas, hope they help.
GL

[Smelly Cat]

You need to think in pot equity, which counts all possibilities. Thinking in outs is just an approximation.

[smbruin22]

haven't read the responses very closely...

but these issues are MUCH more important in NL... it's a big way that good players make big money. good player has nut straight, you have non-nut, he may take you for alot of money when you probably think you're splitting the pot. LIMIT it would be one big bet (hopefully)

need to deeply discount in alot of those cases.... i don't completely understand play of 76s. what is relative value of nut straight vs. non-nut flush in playing that hand? you can't have too much confidence that your 76s flush is good, especially with good players potentially bluffing a flush.

very complex subject, but you're on the right track thinking of discounting the outs, and i think most people are too optimistic i.e. need to discount them even more

[mvdgaag]

I agree that you can lose very big pots when you make a big hand and face the nuts or a bigger one. But this is no reason to stop going for these big hands. It happens far more often that you win that even if you have to pay off like once in a while you are making a profit overall. You're not in there to win the pots, you are there to make the most money. And you don't make the most money by folding everytime you might be beat.

[Bad Lobster]

[ QUOTE ]
I's like to discuss the concept of discounted outs.

...

So how do you approach discounting outs? Or do you only really assess and apply it to close decisions? Is this more or less important in NL as compared to limit holdem?

[/ QUOTE ]

In principle, any out should be discounted by the possibility that it's not the nuts and upcounted (is that a word?) by the implied odds. In practice, I usually give 'em full value unless one of the following apply:

<ul type="square">[*] (hold'em) it only takes one card to beat my hoped-for hand, instead of two. For instance, if I hold overcards K Q, I could make my pair and still be beat--but it would take two cards to beat me; a pair of aces, two pair, etc. So I'd give these outs most of their full value.[*] For hold'em, more than two opponents. Even two opponents are more dangerous than one, of course, but if I'm up against three or more then any card that gives me less than the nuts has very little value. (I can't speak to the danger level of two opponents in other games such as Omaha.)[*]negative implied odds. In other words, will I really lose a lot of money if I make my hand and still get beat? If I'm in a limit game, the potential damage is modest and I don't worry about this to omuch. But if it's a no-limit game and stacks are large, then even a very unlikely enemy hand represents a serious threat. A common example is a draw to a flush that's not the nuts--it's unlikely that the other guy has two cards in your suit, but if he does, you could end up walking home in your underwear.

There's also a hidden effect working against you when you draw to a hand that's not the nuts: you won't make as much money when you do have the best hand, because you'll be afraid to bet as much.[*] Redraws. If I make my straight on the turn, will villain improve his set to a full house on the river? I think redraws have the least impact in discounting outs for reasons of implied odds: if the danger card hits, you're going to be very conscious of it and you won't lose very much more money.
[/list]
Sorry that I can't be more specific with numbers, but I don't want to go farther than my expertise justifies. I think I'll stop here.