Thoughts on c-betting percentages - rough start

[fnord_too]

I have some thoughts I am hammering on about how often you should c-bet against a good opponent. The first thing to consider is how often you will hit the flop. I think the better way to attack this problem is by determining how often the flop has hit your hand (i.e. you have a range of hands you will raise, given the flop how often will THAT flop hit your raising range). What I started doing was attacking a different problem and determining how often a given hand will hit a flop. I started with 6789ds, and categorized the good flops and combinations of them. Some of these 'good' flops are actually bad, specifically, the monotone flops in your off suits. Monotone flops in your suit may not particularly be good, either, but if you only have one opponent (or maybe even two) they should probably be considered good for the purposes of continuation betting. so, here are the numbers I got for the give hand...

6789ds Ways Bad Net Net Cum Total Flops 17296
Quads 4 0 4 4 0.02%
Full House 108 0 108 112 0.65%
Nut Straight 472 22 450 562 3.25%
3,4,5 64 2 62
4,5,6 48 2 46
4,5,7 48 2 46
4,5,8 48 2 46
5,6,7 36 2 34
5,6,8 36 2 34
5,7,8 36 2 34
6,7,T 36 2 34
6,8,T 36 2 34
6,9,T 36 2 34
7,T,J 48 2 46
Big Draws 2496 112 2384 2946 17.03%
4,5,{A,2,9-K} 448 14 434
5,6,{A-3,10-K} 336 14 322
5,7,T 64 2 62
5,7,{A-3,J-K} 288 12 276
5,8,T 64 2 62
5,8,{A-3,J-K} 288 12 276
6,7,{A-4,J-K} 252 14 238
6,8,{A-4,J-K} 252 14 238
6,9,{A-4,J-K} 252 14 238
7,9,{A-4,J-K} 252 14 238
Other 2p/trip hands 636 78 558 3504 20.26%
8,9,{A-4,J-K} 252 14 238
6,6,{A-4,T-K} 96 16 80
7,7,{A-4,T-K} 96 16 80
8,8,{A-4,T-K} 96 16 80
9,9,{A-4,T-K} 96 16 80
Other 648 28 620 4124 23.84%
8,T,{A-4,K} 240 10 230
7,T,{A-4,Q-K} 288 12 276
7,8,T 36 2 34
7,9,T 36 2 34
8,T,J 48 2 46


What this tells us, amongst other things, that we are going to solidly hit the flop a little over 20% of the time. I did not count the other flushes we flop.

So, where do we go from here? Well, given that we will hit about 1/5 of the time, we have several possibilities:
Case 1: We hit they didn't
Case 2: We hit they hit
Case 3: They hit we didn't
Case 4: They didn't we didn't

Lets, just for the sake of moving things forwars, say there is a 15% chance they hit and we are HU on the flop. Then we have for percentage break down of cases:
1: 17%
2: 3%
3: 12%
4: 68%

Now, it get's even harder. There are 2 cases for each of these, namely when we are in position and oop. There is a lot of math to do to come up with an unexploitable continuation bet % here, but since I really think that we should be considering flops and how likely THEY are to have hit our raising hand, we should just do some rough figurin' and guessin'. But I am heading home from work now, so I will noodle some more later (but please feel free to add analysis or make observations like "You idiot, this has all been done before and here is a link..."

Note, the chart will probably look better if you cut and paste it into excel, though you may have to past it into notepad and global replace tabs for spaces firsst, or do some other trick.

Note 2: the formatting wont be as nice and the columns may not be as obvious as in my spread sheet. The categories below full house are broken down into the specific cases and summed up in the category row.

Note 3: x,y,{a-b} means one card that is x, one that is y, and one from the range a-b. So 8,T,{A-4,J-K} means an 8, a ten, and one of A,2,3,4,J,K

Note 4: now I am late getting out of work, so no more editing of this post [img]/images/graemlins/blush.gif[/img]

[Troll_Inc]

Interesting analysis albeit a bit difficult to read.

A couple of questions:

What are you defining as hitting "hard"? I would say >60% equity would be pretty good. And with some other combinations in there, I'd agree that you are just a bit above 20-25% of flops.

What do you want to do with the hands that you are a slight 50-60% favorite in? For example, say you hit top pair plus flush draw, you are still a favorite vs an average big pair:
http://www.propokertools.com/simulator/s...amp;h4=&h5=

If you are in a loose game where 5 people see an average flop, you just need to pick up 1 in 5 of the flops that you enter. If you are just playing hands such as these and everyone knows it, you'd just have to pick up a few more than 1 in 5 pots. So maybe just continuation bet like half the time? I would imagine you want it to be just slightly more than you should be hitting the flop and being a favorite. Just a guess though.

[fnord_too]

Hard was big wraps with food nut outs and maybe flush/fh outs.

I have not done things like top pair + flush draw yet, I probably should add those in.

My first concern is not for multi way pots, but for HU because that should give insight to how often one should be continuation betting which also will give an idea about how to exploit people who don't c-b enough or c-bet too often. The analysis gets really tricky though when multiple streets and decent stacks are involved.

[RoundTower]

I'm too drunk to read your analysis. Tomorrow I will read it, praise your thoroughness, but say that you just can't break it down this way, etc.

For now I'll just say it's really hard to get spreadsheets and tables to display in ubb even if you can make the code tags work. I prefer to take a screenshot, host on photobucket, and post it as an image. It's not much more work and gives you a lot of flexibility.