blonde poker forum

Is there a case to be made for meta-folding an any two shove?



Here is the situation. 4 players left on the bubble of an STT

[SB] HERO(1360)
[BB] SuperHungry (1260)
[UTG] JERKIE93 (8170)
[But] Aguilar13 (2710)

The blinds are posted, HERO had  4d 9s

JERKIE93 folds
Aguilar13 folds
Hero to act???




Big Blind is a pretty decent player, but perhaps slightly too tight. I would have him on calling me less that 20% but I have raised his blind for the last three rounds. (Twice with a 2.5BB raise when I had more chips. On the last round with shove)

For this one hand, I have the big blind calling with about 20-25% of hands. Probably nearer the 25% mark.

It's a standard any two shove that gains us equity equal to just over 0.57 of the prize pool. You would make the play with any two cards.



I wonder if a case can be made that folding here is not such a bad play?

If I shove and he folds (3 times out of 4), he is further convinced as to my stealing, and this will loosen his range on future rounds.
If I don't shove (in an any two spot like this) he takes notice and might tighten up his calling range on future rounds.

As it's obviously better for us if his calling range is small (given the stack sizes and positions, it is very likely I will be in a position to steal his blind again.)


The shove is pretty clear in this example.
+0.7% when he calls with 25% of hands
+1.1% when he calls with 20% of hands


What I'm wondering though, is how much value you would put on folding here as far as it would affect SuperHungry's future calling ranges?

In the actual hand I folded, (which I'm pretty sure was a mistake)

I picked up SuperHungry's blind on the subsequent two rounds before he bust out against one of the bigger stacks. I had pretty decent hands both times (66 and J8s)
I'm wondering if he might have called one of them had I nshoved 94o instead?
The J8s hand might have been the 6th consecutive round that I took his blind, and he would have had less chips.

Hence the question about meta-folding. There must be some value to it, but the question is how much?

I think thats the key part Tank, IF he does notice and note it i think its fine to fold super weak holdings here occassionally as every time you shove his range will widen...  If he is a decent player he WILL notice, but most dont... 

It was late when I posted this. Forgot to mention that the blinds are 100/200

Shoving 23o is still +0.6% of the prize pool though. That's $2.80 in a $50 STT, so is this adequate compensation?




I know what you're saying here but  I'm not sure it's only the good players. A lot of pretty bad players will still twig on to the fact that the same guy is beating up on their blind every single round. It's maybe the only thing in the game that they will notice.

The adjustment both the good and the bad player take is the same, they loosen up their calling standards. The bad player just doesn't understand what he's doing.

Some break even players make no adjustent in their calling standards, and just have a moan in the chat box, truly the worst play. (love them)

Oh, surprised by that tbh, so can we shove atc every single time here unless we give him a massive calling range?

its always a shove.

The point is he needs something to call us

There must come a point that we've opened his calling range up so much by continually shoving, that shoving here is -EV surely?  (generally i mean, not necessaryaly for the hand in this thread)

If he starts calling us with any king for example, we can't shove ATC anymore. (close to it though)

exactly, so its not always a shove, history dependant

If he's an expert who knows we're shoving any two he can profitably call us with over 50% of hands!
In this instance we would have to adjust and only shove with 40% of hands 22+,A2+,K8o+,K4s+,QTo+,Q9s+,JTs

Of course if the expert then adjusts again to our new range and only calls with 20% of his hands, we could and should shove any two again. :)

with 10/11bb;s yes not with 6/7

So you shove here with 23o knowing that any A, K, pair, decent connectors call?

It's all about putting people on calling ranges, and yes they will be affected by the history.

In this example, I have him not quite calling with the likes of K4o yet, but he might very well do so soon.



I want to know the value i should assign to easing up on him a bit so he doesn't think I am shoving with ATC. Probably not enough to give away 0.7% but what about 0.3%?

If the big blinds range was something like that.

22+,A2+,K2+,Q2s+,Q5+,J5s+,J8+,T8s+,T9,98s

Then shoving any two would be incorrect. We would still shove a lot of hands though. (40%)


Rare that you'd get called with K4o and Q4o in this spot unless the player was either very very good or very very bad. Most of the time his range will make this an insta-shove.

I fold in spots like this fairly often, and think I might well fold here, depending on the dynamics of the table. If the bigger stacks are playing fairly passively and waiting for someone to bust then I'll probably shove, as I'll worry that if I give the shorty a walk he'll then be able to pick up another set of blinds next hand, and then I'll be falling off the pace.

I kind of know when I do fold that it's probably a mistake. But at the same time, as Tank says, people start to open up their calling ranges whether or not they're really aware of what they're doing and why. They often seem to do it pretty dramatically as well, almost deciding that they're going to call with anything half decent next time you shove. I think this can happen especially if they're not picking up their share of blinds from other positions. I've had a few people in recent days snap call shoves in spots like this with stuff like J8 and Q7 (top 35-40%?), where obviously they've suddenly opened up their calling range to the extent that I can't actually be shoving junk any more. I guess you have to be aware of if and when they're about to go into panic mode and adjust accordingly.

Obviously it's pretty hard to estimate what the 'value' of a fold is here, but I'd say it's definitely one of those things that ICM doesn't take account of, and I suspect it's enough to make a shove that's breakeven or slightly +EV on a pure ICM basis into a shove that actually loses you money. So if in doubt, as in close situations like the 94o, I would guess there's nothing wrong with making the fold and giving everyone another couple of hands to get involved and make a mistake. (Particularly if you think you can shove into the medium stack next hand, when your raise will probably get more respect.)

I dunno, I always get a bit confused thinking about these spots, where ICM and common-sense strategy can appear to conflict. This is probably partly because I don't have that fantastic a grasp of ICM, but also I suppose because while ICM can provide you with an inexploitable strategy, that doesn't necessarily mean that it's the best possible strategy. Am I right in thinking that?

Good question anyway.

I snap shove this, even though you have been beating up his blind. Given he is semi comptent he will still fold enough to shove it.

Yes you have to take into account that your constant shoving will change his calling range but you will often find in this spot that you never end up having to open shove his bb again. Therefore the whole "meta game" is wasted, just estimate a range and if it is +ev shove it.


ICM is simply a mathematical model, which you give ranges to and it does the maths. Get his range wrong that is a poker mistake, not the models mistake. 

Yeah I realise that - I guess what I was trying to say (kind of in line with the point of the thread) was that doing a standard ICM calc for that one hand doesn't necessarily give you the best strategy... For example, it could be the case that even though an ICM calc says shoving this hand is +EV, you'd be better off folding the 94o this hand and shoving next hand, as the fact that you've just folded your SB could tighten up the table's calling ranges against you, and maybe the new BB will be less likely to call anyway. Obviously yes, you could work out the EV of this strategy using ICM, just in more convoluted form.