AK basics for advanced players

[basementproject]

[img]/images/graemlins/tongue.gif[/img]

[PantsOnFire]

[ QUOTE ]
[ QUOTE ]
If you think your opponent is unlikely to play this way with AQ or AJ, you might count 4 wins out of 11, which would let you call getting 7:4 odds, slightly worse than 2:1.


[/ QUOTE ]

ok, i have a few questions:

1) could you walk me through how you got to this "4 wins out of 11" statement? if i think my opponent is unlikely to play this way with AQ or AJ, then he has either AK or JJ right? how do i win 4 out of 11 if that's true?

2) are you able to do these kinds of hand combination calculations quickly and accurately while playing multiple tables and timing down, etc? any tips/links on shortcuts for doing this? i feel like i'd have a hard time adding up all the combos even playing live with plenty of time.

3) when i went back and tried to assign a range, i came up with some additional hands.

the first range you mentioned is JJ,AJs+,AQ+. vs that range i'm basically a 60/40 favorite. but i feel that's not a likely range.

"Holdem Ranger" is like poker stove but can process weighted ranges. i ended up settling on this:

AA(25),JJ,33(50),AKs,AJs-AQs(25),JcTc(50),AKo,AJo-AQo(10)

the (50)/(25)/(10) indicate that the opponent would likely play those hands this way that % of the time. not sure if i'm doing the weighting correctly, but i'm trying to account for the hands that opponent would be less likely to show up with here. certainly in addition to JJ & AJ+, almost every decent opponent will play AA this way sometimes, and 33 & JcTc this way quite a bit. vs this range, i'm about a 60/40 dog.

if an opponent would never CRAI in here with AQ, then i'm a 70/30 dog.

so the problem i have is that by making fairly subtle changes in my assumption about my opponent's range, the conclusion of my analysis changes from a clear call, to a coinflip, to a clear fold. and subtle differences from one board texture to the next will have similar effects, i'm sure. i mean, it took me a bit of time for this specific situation to go through and formulate what i feel is the correct range, so how can i possibly expect to come to the correct decision each time when i have only 30 seconds to do this type of analysis?

i understand the idea of doing these range/equity calculations to build a kind of sixth sense that takes over when you encounter situations, so that you just kind of naturally make good decisions without relying on doing any calculations on the spot. but i have a hard time feeling like it's going to prepare me to handle AK TPTK spots; instead, i'm going to be panicking, trying to figure out if i'm a 60/40 favorite in the specific situation or a 60/40 dog (which is going to be determined by one specific/subtle assumption relating to the board texture and the particular opponent...)

so yeah, any info you can share about how you can practically use hand combination or range/equity analysis IN REAL TIME would be much appreciated.

also, suppose i can get to the point where i can actually identify something like "OK, if my opponent would never CRAI with AQ here, then it won't be profitable for me to call". well, any specific tips on what kind of logic to use to decide whether a specific opponent would or would not make a marginal play like that? again, should i just go ahead and go with "all aggressive opponents would, and all passive opponents wouldn't"? or do i really need to be more specific than that, and rely on specific stats, etc? and if i just don't have enough information to know whether i'm likely to be a 60/40 dog or a 70/30 dog to a particular opponents range, which should i default to, and why?

[/ QUOTE ]
Seriously, by the way you are putting these (good) questions, you really should get the book Professional No Limit Hold'em.

The entire book is exactly about what you are asking.

[rrrorrim]

1) You made the pot too big for your own comfort. Play AKo a bit weaker. For one, it's devious. For two, it protects you from losing huge with the 2nd best hand.

2) Seems like you're playing TPTK strong because "that's what you're supposed to do." Not the best idea... Good players don't follow guidelines!! (So, don't follow what I'm saying all the time, either.)

3) Villain knows way more about your holding than you do his. He can have JJ, AJ, AK, straight+flush draw, etc... But does he know what you have?

Replay both situations in your head, but this time imagine you are the Villain. Very easy to figure out what this ryan guy has.

[tarheeljks]

some pretty nitty advice in this thread. generally you should be stacking off here unless you know that you are up against an extremely nitty player. yes you will be shown JJ sometimes and AJ occasionally but you will see AQ also.

[PantsOnFire]

[ QUOTE ]
some pretty nitty advice in this thread. generally you should be stacking off here unless you know that you are up against an extremely nitty player. yes you will be shown JJ sometimes and AJ occasionally but you will see AQ also.

[/ QUOTE ]
I may be nitty but there's no way I can put a player who is unknown to me (4 hands) on AQ in this spot. That would be one bad player and this is .25/.50 not ulimits. If he had a draw to go with it, then maybe but given the cards we see he would have no draw with AQ.

[tarheeljks]

i'm not saying this is never a fold, but i think calling is definitely the default. he probably shouldn't be check-raising AQ here, but having been 3bet after opening utg+1 he probably shouldn't be calling w/AJs pf. AJ is more likely to be check raised here; however, he's getting here w/AQ more often given pf and once he's there he's unlikely to fold.

i will agree that we have a tough decision in hand 1. we need ~35% equity to call and w/all available AQ combos we have ~33% against AQ+ . i didn't stove this so it is probably off, but it should be close. so we are getting nearly correct odds against this very tight range, not accounting for the times the villain will be making a move. it's not a fistpump call by any means, but i think we just have to make it and suck it up the times we've been outdrawn/coolered.

in scenario two we only need ~29% thanks to the dead money, so i think it is a standard call.

obv reads change this thought process

[pzhon]

[ QUOTE ]
[ QUOTE ]
If you think your opponent is unlikely to play this way with AQ or AJ, you might count 4 wins out of 11, which would let you call getting 7:4 odds, slightly worse than 2:1.


[/ QUOTE ]

ok, i have a few questions:

1) could you walk me through how you got to this "4 wins out of 11" statement? if i think my opponent is unlikely to play this way with AQ or AJ, then he has either AK or JJ right? how do i win 4 out of 11 if that's true?


[/ QUOTE ]
That should have been 3 wins out of 9. I used the count of AQ hands (8) instead of the count of AK hands (6). Your equity is close to 0/3 against a set, although you have some backdoor draws, and about 50% = 3/6 against AK, although other AK hands do not have the backdoor flush draw. That totals 3 wins out of 9.

[ QUOTE ]

2) are you able to do these kinds of hand combination calculations quickly and accurately while playing multiple tables and timing down, etc?


[/ QUOTE ]
Not all of the time. I miscalculated slightly here, away from the table. However, once you run through a few of these, you can remember common calculations.

It's not just a matter of counting, since you may have different amounts of equity against different hands, and you have to decide whether your opponent is just as likely to have each hand, which means you have to decide how likely it is for your opponent to follow that betting line with quite a few hands.

Since you have studied this hand, you will be more comfortable at the table when a similar hand arises.

[ QUOTE ]

3) when i went back and tried to assign a range, i came up with some additional hands.

...
by making fairly subtle changes in my assumption about my opponent's range, the conclusion of my analysis changes from a clear call, to a coinflip, to a clear fold.


[/ QUOTE ]
Well, I don't think those were subtle changes. You can get very different answers based on whether the players are loose or tight, whether they seem to bluff a lot, etc. It is interesting that in this hand, you may be in better shape against a tight player than against someone who will call a reraise out of position with AJo.

If you haven't been paying attention to this particular opponent, you can try to take an average over the tendencies of all of the players you have seen at that level. Sometimes that still isn't good enough; you might run into a very strange action, or stack sizes which are unfamiliar to you. You make a decision, and then think about it later. One of the advantages of Chris Ferguson is supposed to be that he does a lot more of that type of analysis away from the table than other well-known players do.

[ QUOTE ]
how can i possibly expect to come to the correct decision each time when i have only 30 seconds to do this type of analysis?


[/ QUOTE ]
You can't expect to come up with the right answer each time. The best players in the world don't come up with the right answer each time. Great players make more accurate estimates than merely good players, and make better decisions on average, but they still make plenty of mistakes.

In backgammon, there is objective feedback from backgammon programs which are good at estimating the absolute equities of positions. An error rate of 6 millipoints per move would make you a typical competitive player at a major tournament. An error rate of 3 would put you in the top few players in the world. That suggests that the top players make about half as many errors as the merely competitive players, which is a lot.

It is common to size a bet incorrectly, or to call a marginal fold, or vice versa, or to miss a marginal bet or raise. Many winning players often make large errors, too. All it takes to win is for your opponents to play worse. However tough these decisions are for you, they are much tougher for a casual player who doesn't take poker seriously.

[Pvnuts]

Unless you've been able to properly observe this player, I would interprete the check/raise as a loosing call. You have to assume that this unknown or Tight player has you beat. It sucks when the player is unknown and you have to fold just to find out that he's a loose bluffer, but it's cheaper to wait and find out for sure... and besides, you will take note of this player for the rest of the session simply because YOU WANT TO KNOW.... But no doubt the continuation pre-flop bet is good since you want to keep it high so you don't yeild implied odds to the other players and the continuation bet on the flop is a must uless you want to trap an agressor on the turn.

[rrrorrim]

OP, just saw that you have a few hundred thousand hands of poker. Here is my suggestion:

Use instinct a lot more. You have a valuable reservoir of experience... try using it in an unconventional way (instinct) because all your deep analysis may be more harmful than helpful.

Ie, in your AK situations, ask yourself, "Does he have me beat? Yes or no?" Then fold or cal.. [img]/images/graemlins/smile.gif[/img]

[ryanj247]

[ QUOTE ]


Ie, in your AK situations, ask yourself, "Does he have me beat? Yes or no?" Then fold or cal.. [img]/images/graemlins/smile.gif[/img]

[/ QUOTE ]

sounds great, but doesn't work in practice. no way to know if i'm beat or not. just have to try to estimate their range and my equity vs that range.

fwiw, after spending basically all day thinking about this and doing a lot of calculations and such, my conclusion is basically that it will rarely (if ever) be a big mistake to call the CRAI with TPTK in these spots.