|
Here is an excellent discussion about playing middle pairs, when Tournament Indicator shows you are Orange MZoned. The variable involved are seriously endless, but Phaedrus and JSquared try to put this all into numbers... The problem with orange M is that you can very easily get to pot commitment without betting all that much. When you have small to medium pocket pair (PP) you are either dominated or mostly in a race. Occasionally you will have a hand like Ace rag dominated as well.
From shortish stack, the value of PPs and AK increase while the value of suited connectors decreases. Generally we don't want to be calling raises with a small PP, but raising them first in can be a good steal especially if there are few players left to act where it stands a greater chance that your pair of 3s is not dominated.
So I have spent a bit of time thinking about how PPs can be played to maximum value from Orange M and below.
We know that against a hand like AK the PP is a small favorite over 5 community cards, and that AK is a 70% favorite to miss on the flop.
Now, lets consider a 8M stack of 2400 chips with blinds at 100/200.
If you raise 3xBB (more or less standard raise) to 600 and it is called, the pot will be 1500, and you will have 1800 left. Given your single oponent is 70% favourite to have missed the flop, can you shove any flop with positive EV from this raise?
Answer is no. If you play that approach (raise and shove any flop) from a 3X raise, then AK is getting 180 expressed odds (70% of 600) and implied odds to the value of your whole stack.
That means after 100 hands, you will win 60K (assuming AK correctly folds to a shove when it misses the flop - if it doesn't you actually make more..) But you lose 2400, 30% of the time, for a total loss of 72K. Meaning your shove any flop routine has a negative EV.
What if you make the preflop raise bigger though?
I played around with 4X and 5X, but ultimately it seems to come to a percentage of your stack, not multiple of the BB.
If you raise 1/2 your stack pre flop and shove the rest on the flop, this is what happens when AK or 2 over cards call,
You win 1200x70% and lose 2400x30% of the time, which over 100 trials translates to 84K win and 72K loss.
So that is positive EV, but it is also seems to be about the sweet spot because it small enough to allow AK to fold after the flop.
Going back to the 2400 stack size with 100/200 blinds, your raise of 1200 plus a call, creates a pot of 2700. When you shove the flop for your remaining 1200, you are laying odds of 3.25:1. I.e the pot will be 3900 and it costs AK 1200 to call. Those 30% pot odds are not quite enough to make calling a positive EV play with only 6 outs (of course he might flop more than 6 outs, but those times will be off set when you flop more than 2 outs or flop a set to have AK drawing dead).
So based on the above, it seems to me raising 1/2 your stack with a small to medium PP is about right if your table position is such that you are a solid favourite not to be dominated (i.e earlier the position, the higher the PP you need).
-----
Thanks, this is really good stuff.
Hands like 22 - 55 does the same rule still apply or is it more of a fold.
Also how should you play AK if you suspect they have pair? orange m
-----
I may have picked this up from Harringtons book but if I am orange Mzone im pretty much folding anything under 77 in early position, and if its later then perhaps trying to get in cheap or fold to a raise.
-----
Yeah, i neeed to get around to reading that book
----
Ok, couple of ways to look at this.
Harrington says that Suited Connectors and small pocket pairs are no longer playable from yellow and orange M. The reason he says that is that you do not have the implied odds to chase hitting a set or flush/straight profitable.
IMO whilst that is true, it does not mean that you cannot steal with these hands from yellow or orange M. If you stop playing PPs and SCs al together, you are just not playing enough hands. Remember, from yellow and orange M the correct approach is actually to play MORE hands, not less. (In SNGs - including MTT SNGs there can be some bubble and stack size dynamics which influence this a bit too though).
Anyway, I mentioned at the start there were a couple of ways to look at it...
The second is Ryan Fisler's approach of using orange M as a re-shove zone. With an M of between 12 and 8 you usually have re-shove fold equity. Instead of making a play for the blinds only, now your re-shove is making a play for the blinds plus the open raise, so its normally aiming for 4.5BBs instead of 1.5. Accordingly it needs to work only half as often to be profitable. If called you will be guaranteed for a race or be dominated, but now the reward is also higher being an immediate ticket back to high yellow or green M. So the Fisler approach has loads of merit too. It does take some tones to re-shove with pocket 5s though and profiling comes into whether or not you should do it in specific situations.
The last thing to consider is the chances of being up against a higher pocket pair, which is the disaster scenario.
For this, Phill Gordon's PP rule comes into play.
The rule is C = (N x R) / 2
Where C is the % chance you are facing a higher PP, N is the number of players still to act and R is the number of possible PPs higher in rank than yours.
If you have 33 UTG at a 9 player table, then there are 11 higher PPs, and 8 players left to act. The chances someone has you dominated are therefore 45%. Notice however that lots of these PPs may not be able to call an all in bet on the flop. If you called with 66 and the flop came 9QA would you call all-in on the flop?
However if you wanted to accept a 25% chance that you were dominated, then you can make this play with 88 UTG (there are 6 higher PPs, 8 players to act to making it a 24% chance of domination) Is a 25% chance +ve EV?
75% of the time you will be up against big slick or similar
25% of the time you will be dominated.
Lets start with the 8M 2400 stack with blinds of 100/200
There are 2 events to consider. 1 is you are dominated, 2 is you are facing unpaired high cards.
Event 1. Assume the higher PP will always call your all-in on the flop. This is pretty unrealistic since 99 or TT will often fold to a big fat scary flop, but lets make 100% calls a worse case scenario.
If your 88 is up against something like JJ, you will be playing for 2400 chips and you will lose about 80% of the time. So over 100 hands you win 48000 and lose 192000 making for a deficit of 144K
Event 2 is big slick and AQ type hands
In this event there are two main sub events..
2(a) is AK misses and folds. This will happen 70% of the time, giving you positive 105000 over 100 hands.
2(b) is AK hits on the flop and you also flop a set or later improve to a set. The chances of flopping a set are about 7.5:1 or 12%. The odds of improving to a set after missing the flop are about 8%. Flopping or later improving are equally as good so we can add them together for 20% chance of beating TPTK.
Out of our 30 hands left, 6 of them fit this category, where we win 2700 each time or 16200 over 6 hands.
The remaining 24 hands, we lose to TPTK for 2400 each time or 57600 total loss.
Putting that all together,
In the we are dominated situation we lose 144K in 100 hands, which is 1440 per hand or 7.2BBs .
In the "we are facing big slick" situation, we win 63600 or 636 each hand or 3.18 BBs per hand.
If we stand to lose 7.2BBs in event 1 (being dominated) and win 3.2 in event 2 (facing AK), what is our acceptable risk of being dominated?
The minimum acceptable risk must be break even, so
3.2Y = 7.2 x (100-Y)
where y = acceptable % risk
Y = 31%
Which mean a 25% risk tolerance threshold is somewhat conservative but probably about right given that the less we push marginal situations, the more time you can wait for genuinely good situations.
Accordingly, from Orange M, you can make this play from UTG with 88 plus.
From UTG+1, 77 achieves 24.5% risk of domination
From EMP (6 players left) 66 works
From MP with 5 players left, 44 works
Let me know if anyone disagrees with any of this. I am not suggesting it is gospel, rather I'm working it out as I go
------
Interesting discussion, but I have always leaned to shoving it in, because it is rather difficult to play a flop if you get called. What might be missing here is who, not how many are behind you, how close to the money, and what profiles are in the blinds.
I think I have played this situation more intuitively up until now, but this math is def something to ponder David - requires a closer look by me to be sure.
------
The benefit of raising and shoving is that it IMO it tends to give you slightly more fold equity on the flop. If you are resigned to be playing for your whole stack anyway, its doesn't really matter how you get it all in, other than if one way creates more fold equity than another, then the higher fold equity approach must be higher (I would think)
Having said that, if AK does fold on the flop, then you have only won 1500 chips whereas if you shoved and won the race to the end, you would win 2700. That 1200 difference may turn out to be significant in terms of how long it will be before you are back in high risk waters...
The next task is to look at what happens from an M of 6 and then less than 5.
I'll also give a little more thought to comparing equity between raise and shove versus open shoving from an M of 8.
Those two tasks will however have to wait till a few more homely duties are attended to Laughing
------
OK, M of 8, stack of 2400, blinds of 100/200....
Instead of 1/2 stack raise then shove, how does open shoving compare?
Our 1/2 pot raise was 1200 (6BBs) Are there any hands that can call a 6BB bet but fold to a shove for 12BBs? Probably not. So we are still getting called by bigger PPs and AK / AQ type hands and still getting called at the same frequency
Against the bigger PPs nothing changes. We still lose 144K in 100 trials
Againt AK type hands, we are 52% to win with 88 against AKs, and 55% against AKo. Average it out at say 53% to win.
53 times we win 2700 (143100) and 47 times we lose 2400 (112800) for a net +ve expectation of 30300 over 100 trials or 303 per hand or 1.5BBs.
That compares to 3.2BBs if we got aggressive at two stages rather than 1.
If all of the assumptions are correct then clearly the 2 stage approach is a big winner.
Next is what happens when M drops?
I'll get to that shortly
-----
Same blind level (100/200), this time you have an M of 6 or a stack of 1800 (9BBs).
When you raise 1/2 your stack you are raising 900 (being 4.5BBs) pot is 1200. A call makes it 2100 and you have 900 left behind you
When you shove the flop you are laying 2.3:1 odds or 43%. Two over cards can profitably call with 24% pot odds but not 43% so you still retain theoretical fold equity over AK. Nothing changes for the higher PPs situation.
So from an M of 6 the 1/2 pot raise and shove still works. However with only a 4.3BB raise, you may get more callers and so the objective of getting it HU might justify shoving from EP.
What about M of 3?
3M stack gives you 900 chips. Half stack raise is only 2.5BBs. Now you are definitely better off shoving.
Seems to be the case then, that from an M of 6 its borderline and somewhat position depended whether you open shove or 1/2 stack raise then shove.
From an M of 8, it seems to be better to 1/2 stack raise and shove
From an M of less than 5, just open shove.
------
great thread, great read!!
phaedrus, could i ask you to elaborate on your idea of the "raising half your stack- stop and go".
doesn't that lead to many a situation where you have to call all-in (ugly with 33) or fold away half your stack?
------
That's a good question Jumping and relates to what Marty said about PPs often being difficult to play after the flop.
Ideally you would prefer to be OOP after the flop for this play, so that when you shove the flop, you are shoving not calling. That won't always be the case. This is especially so when you are raising something like 33 which will be from LP and the most likely callers will be from the blinds.
It think if you start this play pre flop, you have to be prepared to finish it, which means you will have to call that all-in bet post flop. There are a couple of reasons for this.
Firstly, when you have an Orange M AND you are fairly short stacked compared to the table and/or the pre-flop caller, then you are an obvious vulnerable stack. Most players know that slow playing works best against short stacks and big stacks. Pre flop with an M of 8 you will mostly fit the medium stack category. After you raise half your stack your are then getting pretty close to a short stack after the flop.
Accordingly, there is a higher than normal chance that TPTK will slow play against you rather than shove.
Technically, bluffing works best against a medium stack and after you have invested half your stack you should be viewed as a poor candidate to be bluffed. However, because you started as a medium stack, lots of times, big stack and short stack players will call your raise when they are OOP, with a predetermined view to running a stop and go, themselves committing to shove any flop. Big stack players do this because its does not hurt them that much if you call them and you double up. Short stack players do this because they are deparate and have to be making moves.
So to summarise, there are two factors at work which make the all-in after the flop (that you are being asked to call) not necessarily represent a strong a hand as otherwise might be the case. 1. is that a hand that hit the flop hard will very often slow play because of your stack size, rather than shove and 2 is position which makes a stop and go bluff by the OOP player more likely.
These factors, together with the fact that you really are pot committed anyway, mean that one you start the play, you pretty much have to finish it and call. Yes you will look foolish quite often and you would definitely prefer to be shoving than calling. But I really can't see that you have the option to fold.
What the above highlights, is that the play should not be an automatic option every time you are orange M and pick up a reasonably PP for your position. You have to be aware that by starting the play, you are committing your tournament life to it.
I have said in previous posts that M is a somewhat fluid concept and the value of an Orange M as an expression of your tournament equity can change significantly depending on your standing in the tournament.
The purpose of this play is to gain maximum value from the combination of your hand and your stack size. It is not designed to destroy an otherwise good tournament position, where stacks, bubbles and other factors might suggest a fold is more appropriate.
-----
and preflop? i mean if i see someone raising half his/her stack i will never just call, unless i feel really in the mood to trap and hold kings, i guess. shouldn't you expect to be pushed all-in pre-flop still rather than get called?
and if that happens i suppose you just have to hold your breath and hope for a race?
------
Another good point. Quite often you do end up all-in pre flop. I have been playing this kind of a tactic for some time, but have not previously really investigated the varying EV from different stack and positions. However I would estimate that maybe a little over half the time you end up all-in pre flop anyway, and you obviously have to call.
When you don't end up all-in pre flop, it is normally against someone who have you out-stacked by a considerable margin and their pre flop call is something of a gamble that they can afford to make.
-------
LOL i was going to post this link up, but then I see you have already done it.
I think you may be doing some incorrect math in the first post though...
when you say you lose 2400 if you are called, I think this is wrong.
Remember, the bets you make before the flop are old news once you have a new betting street. You have to keep this scenario separate.
I am not sure if you are right or wrong in your example, but it could be something wrong with it.
-------
I lose 2400 because I am going to put my whole stack in irrespective of the flop. So because of the nature of the play, pre flop bets are not old news in this case
-------
Would it be more correct to say something like.
Preflop investment:
PP VS AK, When AK hits flop
55% EV * 1200
+ (10% EV * 1800) * 30% this % represents the A or K hitting
+ 1200 (assumming AK folds on the flop.
I have no idea if that is right, but it breaks the EV up a little bit.
-------
I don't think so. I am shoving or calling all-in irrespective of the flop (so even if an A or K flops I still shove).
I run the risk of shoving into a paired AK which is obviously very negative EV, but this is all factored into the 2 stage approach. The 30% of the time when this happens is off-set by the 70% of the time AK misses and folds, or misses and calls without correct odds.
I'll try to explain a little better a bit later, right now, I have to go out for a while.
-------
Nah I fully get your explanation, I just didn't know whether it was correct mathematically or not, cuz on the flop, you could say you are losing 1800 30% of the time, but winning 600 70% of the time.
I think changing losing 2400 70%, to losing 1800 70% may change the EV of the situation, because you will have 1800 post flop.
I do get what you mean though. If you raise with a PP and shove ANY flop and get called by any pair above you, or achieve a fold to any hand with overs, you either lose 2400, or gain 600.
--------
JSquared wrote: "I just didn't know whether it was correct mathematically or not, cuz on the flop, you could say you are losing 1800 30% of the time, but winning 600 70% of the time.".
I don't think I follow you. The calculations above only assume you are called pre-flop. The exercise is to compare a 2 stage aggression with a single all-in pre flop bet.
I'm not sure where the 600 fits in. Blinds were 100/200 so that's 300. From a stack of 1800 the pre flop bet would be 900....not 600. You would only win 600 if the blinds were 300 and your pre flop be was 300, but then you would be shoving instead of 2 staging....
------
Sorry, I was a little brief .
Assuming that you have 2400, make a standard raise preflop of 600, and assuming you are called. I forgot to to assume that you ar called by someone on the button, and the blinds fold, making the pot 1200 + 300 = 1500.
So now assuming the guy had AK and you shove any flop with your remaining 1800, you will win 1500 70% of the time where he folds his overs, and lose or run into when he hits his over card 30% of the time, your whole stack, the orignal 2400.
Thats what didn't really roll with me.
I think you should say on the flop you win 1500 70% of the time and lose 1800 30% of the time, because you have 1800 left in your stack on the flop, because you are counting the 1500 as a win, even though it has 600 of "your" chips.
I think that last part is a bit confusing, but my main point is that it should count as an 1800 loss on the flop, not a 2400 loss even though the end result will be a 2400 loss.
I don't really think this really matter in the whole purpose of the discussion though,
-----
I tried to cover that in the first post where I pointed out that a standard raise and shove any flop does not work. Rather it has to be a raise of 1/2 your stack. Hence you would be raising to 1200 not 600.
-----
nice read, thanks for the input. Info on Pocket Pairs being played in orange mzone is something that I think all of us could benefit from because it can be a tricky decision based on the blinds, your stack and your position at the table.
|
Blog 
