All-in with any two cards?

People are saying that they're playing 'TAG' early on and doing OK, but are struggling to find playable hands late on in tournaments. An effective way around this is to use your stack and win pots via aggression. In this article I'm going to look specifically at what I call the '20-bet re-jam'.

With a chip-stack in the 15bb-30bbs range, I'm looking to be aggressive against loose openers. The later the position people are in, the less likely they are to have a strong hand, because the range of hands they play increases. So I'd treat an UTG raisor differently to a CO opener, simply because I have to assume the UTG player opens for much more strength. People are constantly trying to steal blinds, especially from late position, and the most effective way to defend yourself when short is to move all-in. Here's an example to think about:

The EV calculation

We're sitting with 20k chips in the bb, the button covers us and he raises to 3,000 at 500/1k/125 (we are 9-handed). So theres 5,625 in the pot, the sb folds and in the bb we're shoving 18,875 to win the pot.

We need to think about what range the button has, when opening for a raise here. The top 30% of hands includes all hands Ax, down to 98s+ and T9o+. It doesn't include A6o, A4o and pairs 22-44, so the range could really be even wider. For ease though, I assume the villain is opening 30% of hands, but only calling with 66+, A9s+, KQs, ATo+, KQo (which works out at around 10%). This is what we get from pokerstove [the online poker odds comparison tool which uniquely employs hand ranges]:

289,464,991,200 games 0.059 secs 4,906,186,291,525 games/sec

(equity - win - tie)
Hand 0: 68.059% 67.30% 00.76% { 66+, A9s+, KQs, ATo+, KQo }
Hand 1: 31.941% 31.18% 00.76% { random }

So, versus the villain's calling range, a random hand has 31.9% equity. Not bad! What use is this to us? Well at this point, we have a lot of data that can be used to run an EV calculation:

p - % he will fold
pot = size of the pot, before shoving
e - your equity if called in the final pot
s - amount of the shove

EV = (p*pot) + (1-p) * (e - s)

He raises 30% & calls with 10%
I have 31.9% with ATC

2/3 * 5625 + (1/3)* (.319(43375) - 18875) = 2070.6

So according to these estimations of the villain's opening range and his calling range, this is a +EV shove. Here the villain is opening wide relative to his calling range, and shoving works out with any two cards. Obviously rather than ATC, a strongish hand would be preferable for the sake of the times when we do get called. However this basic scenario is pretty common as stacks get shallower, and is definitely something to be aware of. Also, note that if you have the same fold equity, but a smaller stack, then your shove is even more profitable.

Finally, bear in mind history and metagame. If people are expecting you to shove wide, they will loosen their calling ranges and/or tighten their opening ranges, and so shoving any two cards may not be plausible. Playing around with pokerstove and thinking about specific situations you encounter will help with future 'in game' spots which are similar.

Andrew 'golfpro699' Teng is a Team PKR Pro and GUKPT champion. Chat with him in the PKR Forums!