Open-shoving short stacks in MTT play; guide and discussion

As multi-table tournament players, I'm sure we've all found ourselves shortstacked in some very borderline pre-flop situations where we weren't sure what to do. Should we push all-in now or fold and wait? How do we know? These push-or-fold situations are clearly more straight-forward than more complicated endeavors like post-flop play, but they still confound the best of us now and then.

Well, when in doubt, why not use the power of math? You don't actually have to do the math part. I'm providing a very comprehensive chart made by some brilliant individual. It outlines when you're recommended to push all-in with a shortstack with certain cards based on your position. Aside from just giving you the chart (the link is a few paragraphs below), I'm also going to discuss it some, because understanding why certain parts of it make sense is arguably more important than just having it to reference and look at.

The chart is based on three central concepts: M, first-in vigorish, and +cEV.

M is a concept discussed in Harrington on Hold 'Em Volume 2. Your M is essentially the number of orbits you can survive around the table before you're blinded off completely and have no more chips. To calculate your M, add up the small and big blinds, and all the antes (the ante multiplied by the number of players at the table). It's essentially the total number of chips in the pot before the first player has acted on his or her hand.

First-in vigorish is another concept mentioned in HoH2. The guidelines of the chart stipulate that you have first-in vigorish; that is, that everyone has folded to you and you are first to act. Clearly you want this because of the fold equity it affords you. You'd much rather be the one pushing all your chips in with AJo than calling them off to a player who shoved before you. When players have entered the hand before you, you must make your own mental revisions to the chart's recommendations. Consider that, while there are now more chips in the pot to be won, you are more likely to be called if you shove.

+cEV simply means that it gives you a positive expected value in tournament chips. That is, if you were to make the same push in the same situation an infinite number of times, on average you would have more chips afterwards if you pushed all-in than you would if you merely folded the hand. Now, I realize we're talking about a tournament, so +cEV is not necessarily the end-all be-all. Survival is paramount as well, right? Well, yes, but you must also consider that accumulating more chips in the present can greatly increase your chances of survival in the future. Add to this the top-heavy nature of tournament payouts in general, and pressing any sort of edges you can find becomes that much more important. You want to win 1st, not 48th! When you find yourself in a desperate situation with very little leverage, the type of situations we're discussing, I'd argue that making +cEV decisions is very much the way to go. The last thing you want to do is get so low that you're forced to put all your chips at risk when your expected value is actually negative. That's like giving away chips (or money).

So that's what this chart is all about:

http://members.cox.net/wastrel333/Pushbot%20Charts1.pdf (credit to jcm4ccc for creating this chart)

Read the first page carefully to understand how to interpret and use it properly. It explains pretty well.

As mentioned, the chart is almost entirely mathematically based. Does this mean that, if we follow the chart, we can always make the correct decision as to whether to fold or push with our low M? Not exactly. Therein lies the beauty of poker. As mathematic as it is, there still aren't absolute answers, because it depends on you having somewhat accurate reads on the opponents left to act behind you-- that is, whether you judge, as a whole, their calling range to be closest to "tight", "average", or "loose" (see chart). Of course, there will be times when you've only recently been moved to a table with new players and you have no reads. Even without reads, you can make some educated guesses about calling ranges based on stack sizes (and other things such as bubble-play, which I'm not going to discuss for fear of making this post longer than Gone With the Wind).

Dan Harrington also talks about this in HoH2: the more average-sized stacks will often have tighter calling ranges than either the very big stacks or the short stacks. The big stacks may see themselves as having chips to burn and willing to take gambles in order to build even further, if calling an all-in will not damage them too badly. The short stacks may be pressing and looking for any chance to get their chips in with a decent hand and double up or go home. If you really don't have anything pointing you solidly in either direction, I'd recommend just assuming the calling ranges you're dealing with are "average".

The interesting effect of your opponent's calling ranges is that some very different types of hands are much better for pushing at a loose table than at a tight table, and vice versa.

Take a hand like A9o where you're UTG+1, at a particularly tight table. You shouldn't push this at a tight table unless you're really desperate (an M of under 3). Why? Because here, it's so often the case that you'll either take the blinds or get called and be dominated by a stronger ace. What we'd like is to take the blinds but also have a hand that, if called, does relatively well against our opponent's calling ranges. The dilemma, of course, is that once our M gets below certain points, we can't afford to wait for such luxuries. The key is getting a feel for where these certain points are.

Conversely, at a table with loose calling requirements, A9o becomes a much greater pushing hand because of how often we will be the ones in the dominating position, if called. Many of the weaker aces that would have folded at a tight table now become calling hands for our looser opponents. At a loose table we can actually profitably push A9o from UTG+1 with an M as high as 10; complete desperation is not required.

Now let's look at a hand that does much better pushing all-in against a tight table than a loose table. Take something like 87s. This is a very different type of hand, for a few reasons. Whereas having our all-in with A9 called could prove to put us in dominating situations sometimes, with 87s this will virtually never be the case. Generally, the best case scenario would be a pair 66 or smaller calling us, giving us a 50% chance to win. Otherwise we will almost always be behind. This is really a hand we just want to take the blinds with. We never really want to be called. Thus it makes perfect sense that 87s would do poorly against a loose table, a table that calls more. This is especially because a loose table is more capable of calling a shortstack's all-in with a hand like A8 or A7 that actually dominates. At a tight table, when we are called, it will usually be by two big cards which we don't fare poorly against at all. 87s is an impressive 42% to win against an offsuit AK! Of course, when we're called by a pair 77 or larger, we're in bad shape, but we can just chalk that up to poor luck.

Because of the danger of running into these larger pairs, we virtually never want to open push a small suited connector or otherwise small cards in early position. It only becomes profitable when we're opening in late position and unlikely to be called. Also note that, while suitedness certainly helps a bit, this all basically applies to offsuit connectors as well. We're not going to be called and then hit a winning flush often enough for it to make a gigantic difference.

There's really more I'd like to mention, but this may be intimidatingly long, (I apologize!) so I'll just leave it up for discussion. I'm going to leave you with a little test, to follow in the next post.

Test: How good of a pushbot are you!? (answers in next post)

Without looking at the pushbot chart, see whether your instincts tell you to push or fold in the following situations. Then check with the answers in the next post. If you're pushing when you should be folding, this may indicate you're playing too loose and pressing too much. If you're folding when you should be pushing (probably the more common problem), this indicates you're playing too tightly with your shortstack.

(assume all situations are full 9-player tables, and that pushing or folding are the only options)

1. Blinds are 1,500-3,000, with an ante of 150. Your stack is 28,000, and you have an M of just under 5. It's folded to you on the button and you look down to see QTo. The players in the blinds are tight. Push or fold?

2. Blinds are 400-800, with an ante of 50. Your stack is 6,600, and you have an M of 4. You're UTG+1 and UTG folds to you. You look down at 55. You judge the players behind you to be mostly loose. Push or fold?

3. Blinds are 200-400, with an ante of 25. Your stack is 2,500, and you have an M of 3. You're fifth to act (on the hijack) holding T8s. The players behind you are loose. Push or fold?

4. Blinds are 2,000-4,000, with an ante of 200. Your stack is 50,000, and you have an M of about 6.5. You're on the CO with ATo. The players behind you are neither particularly loose nor tight. Push or fold?

5. Blinds are 400-800, with an ante of 50. Your stack is 7,500, and you have an M of about 4.5. You're UTG+2 with 33. The players behind you are tight. Push or fold?

6. Blinds are 6,000-12,000, with an ante of 600. Your stack is 96,000, an M of approximately 4. You're UTG with A7s, at a loose table. Push or fold?

ANSWERS:

...


1. Push
2. Push
3. Fold
4. Push
5. Fold
6. Fold

Feel free to discuss if you disagree with these answers and why, or anything else you'd like to add or debate. Thank you and goodnight.

Great Thread.

Thanks!

I rarely start pushing quick enough, but, when I answered the questions I agreed with everyone of them. I guess that just proves that I need to start my M calculations sooner. Thanks again, you helped me realize a leak in my game.

Great Article, where is part 2?
The test;
First run through was push em all.
Then I looked at the answers, but still not the chart, and realized why the answers are more correct than my gut feeling.
Chart has finished printing and I think it will have a prominant place under the other prominant pieces of stuff on my desk.

This article now moves atop ChukT's late entry into a new improved unauthorized post article contest article contest.

notes in bold.

I also thought #2 seemed the most questionable for the reasons you state, but the chart, which is seemingly widely supported by math and successful tournament players, stipulates that a push is +cEV here. I imagine it's largely because the blinds are coming up right behind us and our fold equity will be really badly damaged once they go through us that time, but I agree it's open to discussion.

vv interesting stuff, and I'll take a peek at the pdf and try it out. I'm a very slow reader so expect comments in about a month

Bump

ok i open shove any stack but that is probably because i am playing a lot of turbos at the minute great read + rep

That was the coolest information I have read in a while. Thanks and i printed it up

this is my answers to this ?
1 push
2 fold
3 fold
4 push
5 push
6 fold

thats what i think about this

Very nice thread. FTR, I agreed with all the answers as you gave them, with some reservations, but a fascinating read, an I'm sure I'll refer back to it often.

Nice guide. Lately I've been using pokerstove and if Im in late position and my hand has over 60% equity in a hand against any two random cards I'll use that as an excuse to push. Only problem with that theory is, with some hands, if you're called you probably dont have 60% equity in the hand. For instance a suited KQ has about 63% equity against two random cards but if you're called it drops to probably around 45% at best

But I joined a 180 person SnG on pokerstars and opened this guide and referenced it. Didnt really use its advice until my last hand which unfortunately my QQ didnt hold up against a 55 instacall. Darn river card

PokerStars Game #9520614268: Tournament #48323750, $10+$1 Hold'em No Limit - Level VI (100/200) - 2007/04/20 - 15:57:27 (ET)
Table '48323750 20' 9-max Seat #5 is the button
Seat 1: Bytor_269 (1710 in chips)
Seat 2: BigBuffet (2541 in chips)
Seat 3: chipsos (1645 in chips)
Seat 4: PK_Buck (6231 in chips)
Seat 5: cotorra1 (3309 in chips)
Seat 6: _Marklar_22 (1475 in chips)
Seat 7: Quiet Man 75 (6865 in chips)
Seat 8: guillote (4350 in chips)
Seat 9: rockstar68 (4874 in chips)
_Marklar_22: posts small blind 100
Quiet Man 75: posts big blind 200
*** HOLE CARDS ***
Dealt to _Marklar_22 [Qd Qc]
guillote: folds
rockstar68: folds
Bytor_269: folds
BigBuffet: folds
chipsos: folds
PK_Buck: folds
cotorra1: folds
_Marklar_22: raises 1275 to 1475 and is all-in
Quiet Man 75: calls 1275
*** FLOP *** [Kh Td 9d]
*** TURN *** [Kh Td 9d] 7♥
*** RIVER *** [Kh Td 9d 7h] 5♥
*** SHOW DOWN ***
_Marklar_22: shows [Qd Qc] (a pair of Queens)
Quiet Man 75: shows [5s 5d] (three of a kind, Fives)
BigBuffet said, ".lol"
Quiet Man 75 collected 2950 from pot
*** SUMMARY ***
Total pot 2950 | Rake 0
Board [Kh Td 9d 7h 5h]
Seat 1: Bytor_269 folded before Flop (didn't bet)
Seat 2: BigBuffet folded before Flop (didn't bet)
Seat 3: chipsos folded before Flop (didn't bet)
Seat 4: PK_Buck folded before Flop (didn't bet)
Seat 5: cotorra1 (button) folded before Flop (didn't bet)
Seat 6: _Marklar_22 (small blind) showed [Qd Qc] and lost with a pair of Queens
Seat 7: Quiet Man 75 (big blind) showed [5s 5d] and won (2950) with three of a kind, Fives
Seat 8: guillote folded before Flop (didn't bet)
Seat 9: rockstar68 folded before Flop (didn't bet)

A few hands later with his stack of about 7000 he shoves preflop with 33 and gets called by AQs and flop comes Q36 turn is Q and river is 6 and he is eliminated.

I feel so robbed when someone who bad beats me doesnt take my chips to the final table or even make the money.

I feel the pain, Marklar. Literally. I actually got knocked out of a tournament in an eerily similar situation earlier; my QQ vs 55 AIPF and the 5 hit on the turn. Pocket 5s are rigged?

Thank you. Excellent post.
Very helpful, especially the low pair chart. Seems I've been pushing with them too early. It's nice to have something to reference.

I know this is an old thread, but i am looking for a copy of the Pushbot Charts, but all the links I have found a no longer any good. Does anyone have a copy they can post somewhere?

archived~