|
Robb
|
08-25-2008, 09:24 PM
Post subject: Poker Math Question - Geek Squad Please Help
|
#1 (permalink)
|
|
4-of-a-Kind
Join Date: Aug 2007
Posts: 3,072
|
|
I'm trying to calculate the probability of flopping a straight or SD with various connectors. I started w/ T6. Here's my problem. I'm counting 2,816 possibilities, or a 14.37% chance. I think I must be too high.
Questions
1. Does anyone know the actual probability or odds? or where to get them?
2. I'm unsure of #4 below. Can you help?
3. I'm pretty sure of #1 - #3, but if I'm wrong point it out please.
19,600 possible flops
64 straights - 0.00327
512 OESD's - 0.02612
2,240 Gutshots - .11429
Check my thoughts here:
1. The only straight possible is 987, which can happen in 64 ways (4*4*4)
2. OESD's can only happen with 3 specific cards (also 64 ways each). Here are the 8 combo's:
KQJ - QJ9 - Q98 - J98 - 875 - 874 - 754 - 543
3. Gutshots can happen as 3 card combo's (64 ways each). Here are the 8 combo's:
AKQ - KQ9 - QJ8 - J97 - 975 - 854 - 743 - 543
4. Gutshots can happen as 2 card combo's
98 but no QJ7 (already counted). There are 4 9's, 4 8's then one card chosen from the 12 remaining T986's and the 24 cards from the 6 other values, so 16 * 36 = 576.
97 but no J85. Same calculation = 576.
87 but no 954. Same calculation = 576.
Any thoughts? BTW, unsure of where to post this, so I came here.
|
|
|
Play for FREE and practice your game at...
Join the FTR Poker Forum to disable these banners and start posting!
|
|
oskar
|
|
4-of-a-Kind
Join Date: Apr 2008
Location: in ur accounts... confiscating ur funz
Posts: 2,452
|
|
I would go at it slightly different.
The possibility of flopping 987 for example - in all permutations would be:
12/50, 8/49, 7/48
Are you sure you're adding probabilities the right way... because I'm getting crap right here too...
you would do something like 1-[(1-12/50)x(1-8/49)...]=x right?
but that doesn't turn out right.
|
|
|
|
overflow
|
|
Flush
Join Date: May 2007
Posts: 356
|
|
Since you've pointed out all the boards that give us straight draws, this is a bit easier. As you've said, the only way we can flop a straight is 987, so:
(12/50)(8/49)(4/48) = 0.00327, or 0.3% chance of flopping a straight.
QJ9, 457, 345, 875, 89J, JQK, Q98, 874 give us open ended draws:
(8)(0.00327) = 0.02612 = ~2.6% we flop open ended
3card Gutshots:
AKQ, AKJ, AQJ, KQ9, QJ8, J97, 975, 854, 743 (543 makes us open ended)
(9)(0.00327) == 0.02943 = ~2.9%
2 card Gutshots:
98x but no QJ7, 97x but no J85, 87x but no 954.
(3)((12/50)(8/49)(36/48) + (12/50)(37/49)(8/48) + (38/50)(12/49)(8/48) ) = 0.2718 = ~27.2%
So I get 33%
|
|
|
|
overflow
|
|
Flush
Join Date: May 2007
Posts: 356
|
|
Quote:
|
Originally Posted by oskar
I would go at it slightly different.
The possibility of flopping 987 for example - in all permutations would be:
12/50, 8/49, 7/48
Are you sure you're adding probabilities the right way... because I'm getting crap right here too...
you would do something like 1-[(1-12/50)x(1-8/49)...]=x right?
but that doesn't turn out right.
|
Nope, it's 12/50 * 8/49 * 4/48
|
|
|
|
oskar
|
|
4-of-a-Kind
Join Date: Apr 2008
Location: in ur accounts... confiscating ur funz
Posts: 2,452
|
|
Well D'OH!
but how do you add it up the right way?
|
|
The strengh of a hero is defined by the weakness of his villains.
|
|
overflow
|
|
Flush
Join Date: May 2007
Posts: 356
|
|
Quote:
|
Originally Posted by oskar
Well D'OH!
but how do you add it up the right way?
|
Like I did in the previous post.
|
|
|
|
oskar
|
|
4-of-a-Kind
Join Date: Apr 2008
Location: in ur accounts... confiscating ur funz
Posts: 2,452
|
|
D'OH again... you're correct. I was calculating OR and not AND probabilities.
|
|
The strengh of a hero is defined by the weakness of his villains.
|
|
spoonitnow
|
|
Straight Flush
Join Date: Sep 2005
Location: IRC Chat Room
Posts: 5,406
|
|
Some similar stuff on flop odds is listed here http://www.flopturnriver.com/Common-Flop-Odds.html
|
Quote:
|
Originally Posted by Ripptyde
I only have 2 simple rules when I am coaching a new student.
Rule # 1: don't ask questions
Rule # 2: don't ask questions
I have no interest in discussing strategy with a protege'. Your job is to remain quiet and listen. I have a very systematic approach that I will share with the right candidate and I promise that I will turn you into a force of nature and show you elements of the game of poker that you never knew existed.
|
|
|
Robb
|
|
4-of-a-Kind
Join Date: Aug 2007
Posts: 3,072
|
|
Quote:
|
Originally Posted by spoonitnow
|
Thanks, I don't mind reinventing the wheel if I have some way to see if I'm in the ballpark. Nice to find that I got the same answer as Pyroxene on both the number of straights and OESD's. Maybe that means my gutshot calculations are correct.
|
|
|
|
Robb
|
|
4-of-a-Kind
Join Date: Aug 2007
Posts: 3,072
|
|
Quote:
|
Originally Posted by overflow
2 card Gutshots:
98x but no QJ7, 97x but no J85, 87x but no 954.
(3)((12/50)(8/49)(36/48) + (12/50)(37/49)(8/48) + (38/50)(12/49)(8/48) ) = 0.2718 = ~27.2%
So I get 33%
|
You're double-counting possibilities here, not excluding enough cards to keep the sets disjoint.
|
|
|
|
overflow
|
|
Flush
Join Date: May 2007
Posts: 356
|
|
Quote:
|
Originally Posted by Robb
Quote:
|
Originally Posted by overflow
2 card Gutshots:
98x but no QJ7, 97x but no J85, 87x but no 954.
(3)((12/50)(8/49)(36/48) + (12/50)(37/49)(8/48) + (38/50)(12/49)(8/48) ) = 0.2718 = ~27.2%
So I get 33%
|
You're double-counting possibilities here, not excluding enough cards to keep the sets disjoint.
|
I see exactly what you mean. The correct calculation should be:
(3)((8/50)(4/49)(34/48) + (8/50)(36/49)(4/48) + (38/50)(8/49)(4/48)) = 0.08816 = 8.82%
So a shave over 14.62%?
|
|
|
|
overflow
|
|
Flush
Join Date: May 2007
Posts: 356
|
|
I keep second guessing myself .
Last time, this one should be right:
(3)((8/50)(4/49)(36/48) + (8/50)(37/49)((4/48)*(47/48)+(3/48)(1/48)) + (38/50)((8/49)(47/49) + (7/49)(2/49))((4/48)(47/48) + (3/48)(1/48)) = 0.090135 = 9.01%
|
|
|
|
swiggidy
|
|
4-of-a-Kind
Join Date: Sep 2005
Location: Waiting in the shadows ...
Posts: 3,777
|
|
I didn't dig through all the math, but I think your percentage is high because of gutshots, it looks like you might flop more gutshots with T6 then anyother hand.
#4 looks right to me
|
|
(\__/)
(='.'=)
(")_(")
|
|
Robb
|
|
4-of-a-Kind
Join Date: Aug 2007
Posts: 3,072
|
|
Quote:
|
Originally Posted by swiggidy
I didn't dig through all the math, but I think your percentage is high because of gutshots, it looks like you might flop more gutshots with T6 then any other hand.
|
Yeah, I thought the gutshot % was too high, but the more I think about it, it makes sense. Gutshots just aren't that valuable. I found a few minor errors. I did the same calculation for 2-gappers, 1-gappers and connectors. Next, I summarized how many of possible straights are "nut" straight draws. Finally, I'm going to count how many are flush draws, how many are combo hands, and so forth. I hope to have the complete list of "flop odds" for connectors and gappers (sooted and otherwise) soon.
I will report back when I get through it all, and have had a chance to triple and quadruple check everything. The patterns were really cool, but it was near midnight when I hit pay dirt. So maybe I was just smoking crack.
|
|
|
|
overflow
|
|
Flush
Join Date: May 2007
Posts: 356
|
|
Quote:
|
Originally Posted by Robb
Quote:
|
Originally Posted by swiggidy
I didn't dig through all the math, but I think your percentage is high because of gutshots, it looks like you might flop more gutshots with T6 then any other hand.
|
Yeah, I thought the gutshot % was too high, but the more I think about it, it makes sense. Gutshots just aren't that valuable. I found a few minor errors. I did the same calculation for 2-gappers, 1-gappers and connectors. Next, I summarized how many of possible straights are "nut" straight draws. Finally, I'm going to count how many are flush draws, how many are combo hands, and so forth. I hope to have the complete list of "flop odds" for connectors and gappers (sooted and otherwise) soon.
I will report back when I get through it all, and have had a chance to triple and quadruple check everything. The patterns were really cool, but it was near midnight when I hit pay dirt. So maybe I was just smoking crack.
|
If villain is willing to stack off with one pair often enough, a gutshot and two overs on the flop has similar equity to a flush draw.
|
|
|
|
Robb
|
|
4-of-a-Kind
Join Date: Aug 2007
Posts: 3,072
|
|
Here is a brief update. The list is Straights, 8-card straight draws, and 4-card straight draws.
3-Gap: S = 64 ; 8SD = 512 ; 4SD = 2624
2-Gap: S = 128 ; 8SD = 1024 ; 4SD = 3072
1-Gap: S = 192 ; 8SD = 1536 ; 4SD = 3648
0-Gap: S = 256 ; 8SD = 2048 ; 4SD = 4224
You can get each probability by dividing by 19600, the total number of flops possible.
Also, I'm working now on "nut" straights. Unless you have AT, it's impossible to have the nut straight draw. With a connector like 98, 75% of straights are the nuts, 5/8's of the 8-card straight draws are the nuts, and 2/3's of 4-card straight draws are the nuts.
I've always wondered why connectors are so much more valuable than gappers (understood the theory, but wanted to know the details). I'll start a new thread with the results when I'm (a) done and (b) sure they are correct.
For those math geeks interested, the number of straights and 8-card straight draws make a nice linear progression. 2-gappers have twice as many straights and 8-card draws as 3-gappers, and 1-gappers have 3x as many, and so on. It seems the 4-card ones should have a pattern, too, so I might have some errors. I'm rechecking everything.
|
|
|