$200 Stipend to Expand A8 Result - Page 4

Phat Mack · #31 10-26-2003, 08:51 PM

Divide by 2 if you want the ratio of the bet to the dead money.

I think I'm missing something. There's $3 in the pot: in the A8o example, moving in with $70 would be 70/3 the dead money. In a $5-$10 game with $15 dead money, I would move in 70/3 * $15 = $350. Correct?

karlson · #32 10-26-2003, 10:23 PM

Yeah, that's right.
Just a question of terminology....to me, there's only $2 dead money in the pot and there's a $1 bet to you.
I think you have the idea.

magic_man · #33 10-27-2003, 02:18 PM

In Mr. Sklansky's articles on "The System" and "Improving the System", he challenges others to improve upon his "groupings" even more...can we use these results to do just that?

~Magic_Man

Nottom · #34 10-27-2003, 02:26 PM

Does anyone happen to have a link to the articles on "the system"?

magic_man · #35 10-27-2003, 02:32 PM

It used to be at the link below, but it's not working now for some reason. Maybe it will come back:

http://www.cardplayer.com/?sec=afeat...p;art_id=13194

~Magic_Man

tewall · #36 10-27-2003, 03:27 PM

So shouldn't it be the Sklansky-Chubukov No Limit Rankings? By the way, it's a good thing "karlson" is with a "k" and not a "c"! I'm not sure David could stomach that.

Nottom · #37 10-27-2003, 03:49 PM

I had checked there before I posted and couldn't find it. I also noticed that there were none of Mason or Davids articles from before a couple months ago listed in the archives. [img]/images/graemlins/frown.gif[/img]

I did however find this thread on these boards.

Summary of Davids Bellagio Seminar

rockoon · #38 10-29-2003, 09:02 PM

[ QUOTE ]
With a one and two dollar blind we now know that A8offsuit in the SB, is better off moving in up to about 70 dollars more (even if the big blind saw his cards) than he would be folding. If we thus say A8 has a rating of 70 or so, what are the ratings for all other hands?

To make sure you understand, notice that 32 has a rating of one while slightly better hands have a rating of two. Thats because you are getting 3-1 if you put in one and 4-2 if you put in two.(Obviously the big blind would always call in this case). Fairly poor hands would be rated three, getting 5-3 odds. At the other extreme, two kings would have a rating of about 1000. Since it will pick up the pot unless the big blind has kings or aces and will win some of those hands too.

If someone can tell me the "rating" for all hands (I would assume with the help of a computer program), I'll send them $200 and give them credit whenever I write about those results.

[/ QUOTE ]

It is interresting to note that the "optimal" (maximizing EV) all-in amount is different from the "maximum" (break-even) all-in amount and that the two functions are not linear to each other. Considering the pocket pairs:

(all values in units of the big blind)

22 - 3.1, 24.6
33 - 3.4, 33.3
44 - 3.8, 41.5
55 - 5.0, 49.9 <- the "opimtal" amount peaks here at 5*BB
66 - 4.8, 58.2
77 - 3.6, 68.0
88 - 3.5, 80.2
99 - 3.2, 96.3
TT - 3.4, 120.3 <- smaller peak here, TT kills str8s
JJ - 3.0, 160.2
QQ - 3.0, 239.6
KK - 3.0, 477.5
AA - INF, INF

While the "maximum" amount keeps rising, the "optimal" amount rises then falls and does a dance near the end. The values for pocket aces are undefined. Aces EV is the money in the pot.

These values were derived from the A8o puzzle where the small blind hold's one of these pocket pairs instead.

JonCooke · #39 10-30-2003, 04:53 PM

Next job, David.
And this is even more crucial for the heads up stage of one table satellites.

With a stack size of n, blinds 1-2.

Lets constrain betting to fold or all-in.

For various values of n:
What range of hands should go all in?
Given that range of hands going all in, what should the opponent call with?

This is a game theory excercise. I've got some approximations for n=20 using TTH. Does anyone have a spreadsheet with Hand A/Hand B/Odds that they could send me? Then I'll tackle the whole problem.

At what value of n is the allin or fold strategy no longer optimal?

bigpooch · #40 11-01-2003, 06:14 AM

David:

Karlson's numbers don't seem correct. Consider KK:
The number quoted is 1290+ but look at this calculation:

Assume (with a benefit to the KK!) that KK wins about
19% of the time vs AA:

There are 50C2 = 1225 possible hands:

AA (6): EV = (x+2)(0.19)-x(0.81)
KK (1): EV = (+3)(1/2)
other (1218): EV = +3

Thus, for the EV to be >0,

6((x+2)(0.19) - x(0.81)) + 1(1.5) + 1218 (+3) > 0
solving yields x as approximately 983. It should
be a tiny bit less since KK doesn't quite win 19%
of the share of pots versus AA.

Cheers,

"bigpooch" a.k.a. "mangler"