2-7 TD Heads Up

[Hosayif]

In Holdem for Advanced Players, Sklansky says to do this against people who raise a lot:

Preflop:

Call at least 40% of raised pots.
Reraise 25% of raised pots

On the Flop:

After reraising-
Check bottom 20% of hands
Check-raise top 20% of hands

After Calling-
Call more than 50% of hands


I was wondering, what are the top 40% of hands in TD with 3 draws to go and the top 25%? What are the top and bottom 20% with two draws and what are the top 50% (or slightly more) of hands with two draws to go?

Also, is there a free (preferrably Mac, but windows works too) program I can get that I can use to find out this stuff and to run simulations for 2-7 TD?

[randomstumbl]

I'm not sure that knowing what the top 25% of hands is is as useful in TDL as it is in hold'em.

In TDL, you're better able to estimate the relative value of your hand. If your opponent drew one card on the second draw and you drew 2 and made a 97432, you know you have the best hand a majority of the time. If your opponent has been pat, you're less likely to have the best hand.

Therefore, I think the real trick is estimating the strength of your hand relative to the range of hands your specific opponent is likely to hold. For example, in a ring game, I discount some opponents chances of winning because I know they love rough one card draws.

[Hosayif]

I'm fairly new to this game (ok, brand new, i just got SS2 two days ago) but ive been playing fake money sngs on UB and have gotten heads up on each one ive played in.

My questioon is, knowing nothing about what my opponent draws to or stands pat with, what are those percentages. I may be drawing 2 at a smooth 9 or something like that and he may get lucky with his 3 card draw to a rough 8 and beat me, so i need to know these percentages. A free program that does this would be nice. (I don't know if I made any sense with this post, so if you don't understand, maybe i can explain it better)

[Hosayif]

Like, say i have a 2479, if i draw for the 7, my hand is in a different percentil than if i draw for the 9, so knowing these percentages would help in making a decision to perhaps call with the 2 card draw or raise with a one card draw. (assuming the 2 card is a dog to the one card, once again, im brand new at thsi game)

[MarkGritter]

[ QUOTE ]
Like, say i have a 2479, if i draw for the 7, my hand is in a different percentil than if i draw for the 9, so knowing these percentages would help in making a decision to perhaps call with the 2 card draw or raise with a one card draw. (assuming the 2 card is a dog to the one card, once again, im brand new at thsi game)

[/ QUOTE ]

I'm not entirely sure that thinking in these terms is helpful for triple-draw, even HU. Position matters a lot, so even a relatively stronger hand before the first draw may not be worth reraising. For example, 732xx is going to be better than a lot of the 2-card draws, or 3-card draws, that your opponent is raising, but still may not be worth reraising from the big blind.

My rule of thumb is to reraise any 1-card draw or something like 73322 or 77732.

Another key rule of thumb is that any (decent, no straight or flush) 1-card draw is a favorite over any 2-card draw.

Hand percentiles do become more important when evaluating value betting on the river or standing pat vs. multiple opponents, but knowing your opponents is much more valuable in making most decisions.

Drawing to a 9 is a tricky case--- you should probably not ever be drawing 2 to a 9, of course. But in the situation you give I would not draw to the 9 on the first draw OOP, and not draw to the 9 against an opponent drawing 1. I might also try to draw more smoothly against an opponent who chases rather than folding when two cards behind.

Unfortunately I don't have access to the software I wrote which can illustrate the difference in expected outcome (in absolute rather than relative terms) between drawing to the 9 and to the 7 in this case.

[randomstumbl]

Mark's post says basically what I'm thinking. Your opponent's tendencies are often the most important factor in any decision.

I guess at this point, I'm probably the worst person to ask this question. I just have a set of drawing decisions that are basically memorized. That doesn't mean I always discard a 9 or always stay pat with a 9, it just means that I know what I normally do and can look for reasons not to do that.

For example, if my opponent stays pat with rough hands and pays off with them, I'm more willing to draw rough (but, not as rough as my opponent). If my opponent folds too much, I'm more willing to open the pot.

Basically, I look at how their play differs from mine. Then, I think about why I don't play like they do and think about a strategy to exploit that decision. Unfortunately, you have to know the basic strategy of the game to do that.

My rule of thumb is to throw a 9 away on the first draw. Also, you should realize that a decent 3 card draw is not that much of a dog against a 2 card draw. I'm not sure you're a favorite at all when you draw two cards to a "smooth" 9 against a decent 3 card draw.

[MarkGritter]

[ QUOTE ]

I was wondering, what are the top 40% of hands in TD with 3 draws to go and the top 25%?


[/ QUOTE ]

Here's a back-of-the-envelope estimation.

Pat 7's, 8's, and 9's are 2% of hands. (But note that a pat 9 may be a dog vs. 3 1-card draws to a 7.)

1-card draws to a 7 (including all the crappy ones) contitute 3.5% of the deck, or 5.6% cumulative.

1-card draws to an 8: 7% (12.7% total)

2-card draws to a 7: 15.5% (~28% total)

2-card draws to an 8: 23% (~51% total)

3-card draws to a 7 (again, including all the crappy ones you shouldn't play): 15% (~67% total)

(Pat Ts and Js occur 6% of the deals, but there is some overlap I didn't account for between these and the draws listed--- and these are not good hands to stand with.)

Note that even though all playable 2-card draws to a 7 lie within the top 25% on this ranking I think it is unwise to reraise these OOP. Hand values will fluctuate much more than in Hold'em, with the last draw often being a 70/30 or 60/40 situation (which would be unusual on the river in Hold'em.)

Anyone besides me think that playing like that is too mathmatical?

[timprov]

Note that this is against someone who raises a lot, and by a lot I think Sklansky means even more than I do. Certainly 2x7 ought to be three-bet OOP.

As a basic fairly-tight guideline, play any hand with a 2 and a 7 in it, 2x8 and 3x8. If you need to loosen up add 23, 24, 25, and semibad one-card draws like 3457. 3-bet 2x7, 2xx8, 3xx8 and better.

You won't run into many of these in TD though, because unlike holdem there are a lot of hands that just aren't playable, and nearly everyone realizes this. Against a typical player who almost never raises and draws 3, you'll want to be a little more passive. Against a nut who raises and draws 4 or 5, you'll want to 3-bet more.

[MarkGritter]

[ QUOTE ]
Note that this is against someone who raises a lot, and by a lot I think Sklansky means even more than I do. Certainly 2x7 ought to be three-bet OOP...

[/ QUOTE ]

I could get behind 3-betting 2x7 vs. those obviously raising junk (50%+ with the button) but I don't like it so much vs. a more reasonable but aggressive player. Is getting the extra 1 SB in here worth it? I think you can easily be at enough of a disadvantage (and villian is drawing 2 enough of the time) that you may be better off just calling.

Hmm... need to run some 853 vs. 237 simulations and see what happens.