ICM-ish idea: Calling ranges

[Slim Pickens]

So to elaborate, this is going to take an impossibly-large sample to collect any useful information. It's impossibly-large because game conditions are going to change faster than you can collect and analyze it. You might be able to reduce this by binning either your opponents or their hands or both. The problem with this is you have to bin them correctly for it work. Do you bin them by S-C ranking? That only works if that's what you're opponents are doing, which not all opponents will be. Put a hand in the wrong bin and it skews the probability of seeing a hand in that range. Fail to distinguish between a complete nit and a tight-early/pushbot-late player and you'll mis-estimate the probability of both types of players' actions.

It is further complicated by the fact that people's ranges are not absolute. What you really need to determine is the probability of them playing a particular hand. It's not a problem to assume 100% with AA, but around the hands where it really matters anyway, that probability won't be 100%, so it will take even more sample hands to determine that extra parameter.

[Slim Pickens]

[ QUOTE ]
At 3-4 players, position becomes fairly irrelevant for calling a shove. A shove from UTG/CO isn't much tighter than a shove from the BU or SB.

[/ QUOTE ]

? and !, and maybe [img]/images/graemlins/frown.gif[/img]

[fluorescenthippo]

you could also just do % of time called from each position and stack size. this would take over 10k hands im sure but it would be fairly accurate and maybe faster than looking at individual hands idk. just throwing another idea out there that ive had for 2 years.

you would then just convert that % to KS ranges. id imagine that would work

[Slim Pickens]

Position is definitely a good parameter. The problem I see with the S-C rankings is that people have some idea about how certain hands fare against people's typical calling ranges and adjust to that. In other words, correct pushing/calling hands don't necessarily follow the S-C rankings. My guess is you'll see a lot more (correct) shoves with JTs than with A2o, even though A2o has a higher S-C ranking. I actually like this eastbay modified set of hands people like, where JTs (0.414) ranks well above A2o (0.326) against {22+,A2s+,A6o+,KTs+,QJs}.

[Q..]

Absolute range problem is pretty serious - some times ppl would make really hopeless calls after lets say being pushed 6 times in a row from sb on some J4o because they had <9high cards previous 6 times, and this hand would be in the range along with AA, so youll need like huge(infinte?) amount of data to analyze to correctly compute range and adjust value of each possible call hand with probability to hold it and probability to call when its held. And none of programs on the market afaik deal with such probable ranges so youd have to create a pretty sophisticated soft to use the data collected, if someone would somehow gather it.
One can of-course analyze large history db and analyze calls of good players in various stacks - ppl remaining situations, such average results might be applied if you are dealing with lets say a good regular player, or a random guy. With all that done and computations performed on the fly you could probably exploit something new to usual icm model & absolute call range, but it all sounds too complicated, and most likely wont be worth the effort since that additional edge would be really small, because of possible difference of concrete player and average one.

[bogey1]

[ QUOTE ]
Let's say you put me on 22+,A2+,K7+,Q8s+,Q9o+,JT+ but I know my range is really 66+,AQ+,KQ+.

[/ QUOTE ]

Your example is backwards from what I'm trying to discover. I'm not looking to find the range you're pushing, but the range you'd call my push.

Though, the point is the same, if you're vastly different than the average, then my shove is wrong. However, if you're vastly different than the average, my hand range on you is probably wrong anyway. If you're so tight/loose it's obvious, then I'm deviating from the average in my decision making regardless.

[bogey1]

[ QUOTE ]
[ QUOTE ]
At 3-4 players, position becomes fairly irrelevant for calling a shove. A shove from UTG/CO isn't much tighter than a shove from the BU or SB.

[/ QUOTE ]

? and !, and maybe [img]/images/graemlins/frown.gif[/img]

[/ QUOTE ]

A frequent 6-man situation:
100/200 blinds
BB 2400
SB 3700
BU 1300
CO 1600

The CO shoves. Do you really think the BU is going to alter his call range much because he's on the BU vs. the SB or BB? Is the CO really shoving significantly different from the CO than he would from the BU?

If it was 7-handed, that's a different creature entirely.

[bogey1]

[ QUOTE ]
Absolute range problem is pretty serious - some times ppl would make really hopeless calls

[/ QUOTE ]

True, but irrelevant. Those hopeless calls will pale in number compared to the more reasonable calls and then get thrown out when I take the standard deviations off the frequency curve.

I'm far more concerned with relative stack sizes tainting the results. Big and tiny stacks call more liberally than medium stacks.

[bogey1]

[ QUOTE ]
modified set of hands people like

[/ QUOTE ]

I'd like to look at that, but the link is broken. Can you relink it?

[Slim Pickens]

[ QUOTE ]
[ QUOTE ]
modified set of hands people like

[/ QUOTE ]

I'd like to look at that, but the link is broken. Can you relink it?

[/ QUOTE ]

http://seoblackhat.com/texas-hold-em-poker-statistics/