I think a lot of people in SSMTT try to use SnGWiz, plug in a hand, either don't like what the see (or get bad results), get confused, throw up their hands in disgust and don't touch the program again. It's a very good tool, but you have to give it realistic ranges to get realistic results. By setting proper ranges, I take an AQo hand that SnGWiz thinks is a blunder to call and convince it that the hand really is okay to push.

Plus, I haven't seen any videos analyzing this sort of thing in SNGMTTs (not that I've really looked ).

The first half of the video reviews SnGWiz's interface. If you want to go straight to the analytic part, skip to 4:12.

So, without further ado...



(Also in glorious 720p on YouTube proper.)

Cliffs/additional points:

• If you want to analyze single hands, it's easier to copy/paste them into SnGWiz than trying to load an entire tournament's worth of hands.
• SnGWiz often has poor pushing/calling ranges for your opponents. Use your reads and your intuition to set more accurate ranges.
• When using custom ranges, be sure to set ranges for when you fold for your opponents yet to act. How your opponents play affects you in ICM situations when it doesn't in cEV situations.
• You gain when two other people are all in; if someone is playing too loose, step aside and let him take himself or others out.
• It's easy to underestimate your current equity in these FTs. With 8 players left, the average payout of a 12/45 on Stars is $61.88, so the equity of a near-average stack should be around that value. Since that's 40% of the 1st place payout, it's also easy to overestimate the value of a double-up.

Here's the analyzed hand:

Poker Stars $11+$1 No Limit Hold'em Tournament - t800/t1600 Blinds + t75 - 8 players
The Official 2+2 Hand Converter By DeucesCracked Poker Videos

UTG: t3915 M = 1.30
UTG+1: t3981 M = 1.33
MP1: t8340 M = 2.78
MP2: t10340 M = 3.45
CO: t12200 M = 4.07
BTN: t9664 M = 3.22
Hero: t7485 M = 2.50
BB: t11575 M = 3.86

Pre Flop: (t3000) Hero is SB with A Q
4 folds, CO raises to t12125 all in, 1 fold, Hero?

firstttttttt

2nd. I actually reviewed your hand history and made half the video lost the 2nd half and got really frustrated. sorry sopo.

No problem.

Making a video is harder than it looks (not a level, btw.)

really really great post

havent watched the video yet but im sure its good.. one question tho..

u said in ur post that its easy to overestimate the value of a double up and strictly according to ICM its understandable why that is. but how, if at all, do u value the ability to use a CL to force others into folds with shorter stacks still around? this isnt really much of a brag bc im assuming its std but i have multiple hh's from 6/45 ft's where ill gain the cl and take down like 16/21 pots, with like 13 of those taken down preflop. how do u factor this into ur ICM calcs?

I took a stab at estimating this once in the last half of a FT, and decided that stealing 1 M from everyone 5-handed when you already have half the tournament's chips was worth about 2/3 of a buy-in. I'd cut that in half to account for you get very bad hands that can't raise and the times when your opponents do call. As for the hand I analyze ITT, hero would have a smaller chip lead, so he's risking a larger portion of his equity with his shoves, and probably can't shove more than 20% from EP 7- or 8-handed. I'd guess doubling up would be worth less than 1/4 buy-in more than ICM (though that is 0.5% in Edge% terms).

What I thought was interesting is that when a big stack is doing this, he's mostly stealing equity from the smaller stacks and little equity from 2nd place.

well done bro! this is very nice! now if only they made a mac version of SnG

Wow great. Many thanks.

Can't wait to watch this when I get home. Thanks OP.



Jitxpert- use bootcamp or parallels.

ty ty Sir.

Well done sopo!

nh

Very educative...Thanks sir

nice sir

Two thumbs up. Do you mind if I link to your video on the SitNGo Wizard web site?

Creeepy, it's like you were right there ! !

But very awesome post, thank a ton. Always wanted to know how to use this bad boy.

icm dont work

I think this video and discussion are both awesome.

I consider myself a bit more ICM-abiding than other posters on here, but I am certainly aware that it has its shortcomings. You spelled out a few of the shortcomings in much more detail here than I usually think about them, and I feel like by understanding the shortcomings of ICM, it makes it much more useful.

Thanks for putting it together.

A question: now that you did that analysis and found that you do indeed like your push there, but you see that it is close, would you be happily folding if you felt that CO was on the tighter side? Or are you still looking for a spot to get it in +cEV since you are nearing the point where you blind out.

ty vm bro!

Go for it. Thanks for making a useful program.

At the table, I'd plan on calling unless I thought villain was a demonstrable nit (a "nit" in this case would mean stats like 14/11 with 20% ATSB). Even then, I wouldn't happily fold it; I'd say "Really?," try to rationalize calling, sigh, and fold (or mistakenly decide shoving is still +$EV and call).

I'm not particularly worried about the cEV-ness of a play at an FT, since there's a better framework for making decisions. Ideally, we should be always making +$EV plays, but early in a tournament, we use cEV as an approximation because there isn't anything better.

Are you sure, that most Regs shove 60%+ there?In most cases you have some players behind that dont know what icm is and call with Ax and stuff. And then it gets massively -ev to shove so wide. Its only +ev, if you have competent players behind.



Uploaded with ImageShack.us

Thats if everybody knows that he should fold AJo. So its already very conservative.

I wasn't saying most regs were. I felt this player would shove that much due to his aggressive stats.

It looks like copying the hand from the converted HH makes SnGWiz ignore antes.

Changing Edge% from +.1% to 0% changes CO's suggested pushing range from 51% to 76%. Further, if I go through the remaining 3 players and change their ranges to 40% if it's folded to them, 6% if any single player is all in, and 4% when 2+ players are all in, CO can actually push ATC for profit at 0% edge and about 74% at .1% edge.

Here's the raw HH, for posterity.

PokerStars Game #46828258099: Tournament #292078662, $11+$1 USD Hold'em No Limit - Level XII (800/1600) - 2010/07/15 20:24:49 MT [2010/07/15 22:24:49 ET]
Table '292078662 4' 9-max Seat #5 is the button
Seat 1: UTG+1 (3981 in chips)
Seat 2: MP1 (8340 in chips)
Seat 3: MP2 (10340 in chips)
Seat 4: CO (12200 in chips)
Seat 5: BTN (9664 in chips)
Seat 6: Hero (7485 in chips)
Seat 8: BB (11575 in chips)
Seat 9: UTG (3915 in chips)
UTG+1: posts the ante 75
MP1: posts the ante 75
MP2: posts the ante 75
CO: posts the ante 75
BTN: posts the ante 75
Hero: posts the ante 75
BB: posts the ante 75
UTG: posts the ante 75
Hero: posts small blind 800
BB: posts big blind 1600
*** HOLE CARDS ***
Dealt to Hero [Ac Qh]
UTG: folds
UTG+1: folds
MP1: folds
MP2: folds
CO: raises 1600 to 12125 and is all-in
BTN: folds
Hero: calls 7410 and is all-in
BB: folds
*** FLOP *** [6c 5d 9s]
*** TURN *** [6c 5d 9s] [Th]
*** RIVER *** [6c 5d 9s Th] [2s]
*** SHOW DOWN ***
Hero: shows [Ac Qh] (high card Ace)
CO: shows [Qc Qd] (a pair of Queens)
CO collected 17020 from pot
*** SUMMARY ***
Total pot 17020 | Rake 0
Board [6c 5d 9s Th 2s]
Seat 1: UTG+1 folded before Flop (didn't bet)
Seat 2: MP1 folded before Flop (didn't bet)
Seat 3: MP2 folded before Flop (didn't bet)
Seat 4: CO showed [Qc Qd] and won (17020) with a pair of Queens
Seat 5: BTN (button) folded before Flop (didn't bet)
Seat 6: Hero (small blind) showed [Ac Qh] and lost with high card Ace
Seat 8: BB (big blind) folded before Flop
Seat 9: UTG folded before Flop (didn't bet)

Solid thread.

And bump it