The video was pretty good. I'm just curious why you didn't just make it a HU game instead of CO vs BTN. Is there any reason for solving a 3b IP situation to be more interesting than a 3b OOP one? It seems like there would be smaller error in the calculations (cuz of fewer variables), or at least it would be simpler to have the blinds be a part of the players' stacks that we're dealing with. I don't know I'm not too sure about either of my assumptions, just wanted to know if there's a particular reason you chose this situation.
Also, I think you could have briefly mentioned the [0,1] distribution for people who haven't read MoP. It probably wouldn't take too long to explain it, and other than that I don't really think you have to be familiar with any part of the book to understand this video.
BTW, I would be interested in both more GT videos and being in a MoP study group.
The video was pretty good. I'm just curious why you didn't just make it a HU game instead of CO vs BTN. Is there any reason for solving a 3b IP situation to be more interesting than a 3b OOP one? It seems like there would be smaller error in the calculations (cuz of fewer variables), or at least it would be simpler to have the blinds be a part of the players' stacks that we're dealing with. I don't know I'm not too sure about either of my assumptions, just wanted to know if there's a particular reason you chose this situation.
Thanks. I can't remember any particular reason for choosing CO vs. BTN, I guess it's a spot where both players ranges should be pretty wide. I was looking at 6max play rather than HU so that's why I didn't do HU.
Also, I think you could have briefly mentioned the [0,1] distribution for people who haven't read MoP. It probably wouldn't take too long to explain it, and other than that I don't really think you have to be familiar with any part of the book to understand this video
I mentioned it in my last two videos, they were designed as an intro to game theory so I wouldn't need to cover too many basics in future vids. I think it's pretty important to have read MoP, although perhaps there is a subset of people who don't have the book but would still get some use out of this video. It might be a little difficult to understand why I'm doing what I did without seeing the same process in toy games and, if somebody didn't have enough interest in game theory to get the book, I'd be surprised if they weren't put off by the amount of mathematics in the video.
BTW, I would be interested in both more GT videos and being in a MoP study group.
I wasn't suggesting starting a MoP study group, I was thinking of doing a video series based on the book for people who haven't been able to get through it by themselves. Just a bit of brainstorming and interest gauging, it may not happen. If you were looking to get into a MoP study group anyway, there's one going on [url=http://forumserver.twoplustwo.com/33/books-publications/mathematics-poker-study-group-another-kick-can-811251/]here[/url].
I had a request for an excel spreadsheet where you could change the inputs and it would spit out the new ranges.
Cells that are highlighted in yellow are where you can change the inputs. Blue cells are the new ranges and the chart will automatically update whenever you change something. It's pretty straight forward to use, the formulas for calculating the outputs are a bit long but you don't need to go at them.
Your link doesn't work. RedJoker's does though so no real harm done.
@RedJoker
Great video - would like to see more GTO videos from you. MoP is a really excellent poker read for the mathmatically inclined. But honestly I am not very interested in a series basicly walking through a book I have studied quite a bit. I think it would be much more interesting with new material and ideas (I know this is really hard work and probably too time consuming to make it worthwhile for you). Anyways, thanks for the video.
Your link doesn't work. RedJoker's does though so no real harm done.
There's been extensive testing done on those links, everything except the .zip file works.
@RedJoker
Great video - would like to see more GTO videos from you. MoP is a really excellent poker read for the mathmatically inclined. But honestly I am not very interested in a series basicly walking through a book I have studied quite a bit. I think it would be much more interesting with new material and ideas (I know this is really hard work and probably too time consuming to make it worthwhile for you). Anyways, thanks for the video.
Thanks. Yeah, it wouldn't be aimed at people who've already studied the book.
Thanks for explaining.
Fwiw, I would be interested in videos about MoP too.
Edit: I just read WaterBear's post and I think he's right. But you could make videos that kinda make use of MoP to try to solve real life poker situations (nlh or plo). Kind of like this video I guess.
Is there a way in the spreadsheet to narrow X's calling 3b oop range, since to model it more realistically, calling 1/3 of the hands you are opening (so about 8% of total hands - and that is not even the top 8%) OOP is just not going to end well
Is there a way in the spreadsheet to narrow X's calling 3b oop range, since to model it more realistically, calling 1/3 of the hands you are opening (so about 8% of total hands - and that is not even the top 8%) OOP is just not going to end well
Yeah, you can change cell C15. By increasing the amount of equity you assume X has (decreasing the value of cell C15), you'll reduce the size of the calling range. It seems counter intuitive that assuming a higher expectation will reduce how often you call but it's a common result in game theory that the higher the EV of an option/line, the less you take it. There's an example of this on page 106 of MOP.
Alternatively, you could create a different model which assumes no 3bet calling range and solve that.
Really really nice vid Phillip, ty. i wd b interested to see a vid series on the optimal play chapter in MOP.
Yeah, you can change cell C15. By increasing the amount of equity you assume X has (decreasing the value of cell C15), you'll reduce the size of the calling range. It seems counter intuitive that assuming a higher expectation will reduce how often you call but it's a common result in game theory that the higher the EV of an option/line, the less you take it. There's an example of this on page 106 of MOP.
yeh but that doesn't increase X1 and the 4b/fold range at all (and hence any other ranges)
I was more concerned with reducing X's calling 3b range while still widening his 4b-value and 4b:bluff range to make it more realistic
yeh but that doesn't increase X1 and the 4b/fold range at all (and hence any other ranges)
I was more concerned with reducing X's calling 3b range while still widening his 4b-value and 4b:bluff range to make it more realistic
Ah, o.k. If you increase cell C22 slightly, which is the equity assumuption for the hand Y1 vs. the range 0-X1 it'll tighten the calling range. I chose 30% based on:
equity win tie pots won pots tied
Hand 0: 71.849% 70.71% 01.14% 435898404 6998250.00 { JJ+, AKs, AKo }
Hand 1: 28.151% 27.02% 01.14% 166534536 6998250.00 { AQo }
equity win tie pots won pots tied
Hand 0: 66.832% 66.63% 00.20% 273837216 811740.00 { JJ+, AKs, AKo }
Hand 1: 33.168% 32.97% 00.20% 135492264 811740.00 { 99 }
equity win tie pots won pots tied
Hand 0: 70.469% 69.36% 01.10% 142527048 2269746.00 { JJ+, AKs, AKo }
Hand 1: 29.531% 28.43% 01.10% 58409940 2269746.00 { AJs }
equity win tie pots won pots tied
Hand 0: 70.670% 70.24% 00.43% 144321864 889020.00 { JJ+, AKs, AKo }
Hand 1: 29.330% 28.90% 00.43% 59376576 889020.00 { KQs }
but if you widen the 4bet/call range than an assumption of 32% - 33% might be reasonable:
equity win tie pots won pots tied
Hand 0: 66.963% 66.77% 00.20% 342971688 1009050.00 { TT+, AQs+, AKo }
Hand 1: 33.037% 32.84% 00.20% 168701412 1009050.00 { 88 }
equity win tie pots won pots tied
Hand 0: 67.995% 66.91% 01.08% 178733520 2894472.00 { TT+, AQs+, AKo }
Hand 1: 32.005% 30.92% 01.08% 82596960 2894472.00 { AJs }
equity win tie pots won pots tied
Hand 0: 68.116% 67.71% 00.41% 180860688 1090542.00 { TT+, AQs+, AKo }
Hand 1: 31.884% 31.48% 00.41% 84077652 1090542.00 { KQs }
Alternatively, you can assume that X never calls and solve the resulting problem which is pretty straight forward. To make Y indifferent to 3bet bluffing we need the EV of a 3bet to be equal to 0:
(Z% x 4.5bb) + (1-Z%) x -10 = 0
Z% = 0.6896
So X needs to 4bet 1 - 0.6896 = 31.03% of the time. Using the 0.8096:1 ratio we get that the 4bet value range must be (1 * 31.03%)/1.8096 = 17.15% of the opening range. This would be the maximum the 4bet/calling range could get to.
There are a couple of adjustments we could make to this. The first is the 4bet value:bluff ratio which, as discussed in the video, may be a little high. Widening the range, as we saw, will reduce the expectation for SCs and PPs but increase it for suited aces so we may want to reduce the ratio slightly. This increases the amount of 4bet value hands needed. Another adjustment is that, with blinds still to act, we don't actually need to make villain indifferent to 3bet bluffing by ourselves. The blinds will have both value and bluff 4bet ranges which will help us out and make us less exploitable. Because of this we can reduce the 4bet value range slightly.
