Optimal VPIP Out of Position

I have played nearly a million hands of online poker in my career. There are a lot of benefits to having this much experience, but one of the most overlooked is that you can actually have a meaningful sample size to look at for how you play so many different situations. Not sure if you're making money by playing certain hands in certain ways? Look it up! A recent debate on a poker forum ended very quickly when a student of mine bragged about obtaining a certain expectation by limping in an unconventional situation. I posted that over a significant sample size, my expectation had been nearly a full big blind better. Upon hearing this, he quickly conceded that his position was probably incorrect, punctuating the conclusion by saying “more debates with numbers to back them up, please”. One big check to make sure you do not get too lost in theory-land is the actual results of your plays: The proof is in the pudding, as the phrase goes. If you can show that a strategy makes the most money, then you can imply the strategy makes the most sense. I have found falling back on my statistics to answer questions about my own play to be an invaluable tool in getting better.

After a small downswing, I once got paranoid that I was playing too many hands out of position in superturbos. Calling all these marginal 84s, 96o, K5o type holdings – was I actually making money 20-25bb deep, or was I bleeding it away? I did a HEM query for the range shown in the image below at this stack depth facing a minraise, and here were my results:

Negative \$24,000 in equity! Yikes! Remember, though, that we're always comparing against the next best option - if we think the next best option is not playing these hands, the expectation from folding is -1bb from the start of the hand, every time this happens. That's the same thing as -100bb per 100 hands. And as we can see from the stats, my expectation from calling is far, far better than this, at around -59bb per 100 hands. Thus, in general, folding one of these hands is about 0.4bb worse than calling, and that's over 2079 hands. If I had folded all of them, I'd be about 850 big blinds poorer! That's about 38 buy-ins in equity, thrown away out of a desire not to play pots out of position with marginal holdings.

Surely, the expectation is helped by the inclusion of some stronger hands – perhaps you were never considering folding 76s. But even when you take out some of the best hands from this range, calling has still done significantly better than folding.

Holding nothing back, I include all of my stats. I show EV-adjusted numbers, so you know the numbers are not influenced by all-in luck. More importantly, though, I show other stats that look rather tame. There is a myth that if you play these sorts of hands, you have to be Mr. All-Star postflop, check/raise bluffing a ton and making your opponent fold the best hand. In reality, my flop play was pretty pedestrian: Check/folding over half the time with a low check/raise percentage, only winning the pot postflop around 31% of the time. That means a large majority of the time I called and lost postflop. It sure can really feel like we're just bleeding chips, and should stick to stronger holdings preflop.

The problem is, the math on that doesn't really add up. It's OK that we are losing these pots a lot. It's OK that we only go to showdown a little over a quarter of the time, and win the pot just a little over half the time there. Finding good spots to bluff turn when the flop is checked through and other reasonable contest situations, along with the equity from when we hit, is enough with these hands.

That's not to say you shouldn't check/raise bluff or put your opponents in tough situations - I think you can improve on my expectation by playing better postflop than my passive 5-tabling auto-pilot that likely characterizes a lot of these hands. My point is that the belief that you need to be a hero postflop to play 60% of hands is misguided.

First hand in a super turbo, readless, I think it's generally best to play about 55-60% of hands against a minraise, something like this:

You can tinker with this and argue for a few more calls and/or a few more folds. However, what I'm mostly concerned about is the general sense of the borderline. In strategy threads, I often see posters saying that they would fold Q7o to a minraise, or K5o, or 74s, or Q4s, or J7o, or 95s. I think these are all significant mistakes.

This conclusion is not just for endgame play, either. First hand in a turbo or a reg speed against a minraise, you should not be playing less than 50% of hands. Here are my numbers with this same range, this time expanded for 20-48bb deep:

The results are more of the same, and actually even more pronounced that we should be playing these hands. Again, the story isn't phenomenal postflop play - my stats look thoroughly boring and like I should be getting run over. Again, though, while I am down -\$62,000 in equity from these situations, I'd be down almost another \$60,000 by NOT playing these hands. In summary, if it feels like you are bleeding money when you call wider OOP, maybe you are - but bleeding money by folding too much can be a hell of a lot worse.

I have one more point to make on this topic. Throughout, I've said that we should be playing these hands - that doesn't necessarily mean we should be simply calling them (although that's going to be our most frequent play). Many of these hands, and the hands just a little bit worse than them, are really good 3-bet bluff hands, particularly 84s, J5s, 96o, etc. My expectation from 3-betting this range is actually much better than calling, but my sample size is a bit too small to make too much of it. Optimal play is far from settled – see if you can improve on my numbers. Perhaps you already have. At the least, though, we can conclude that folding more is not likely to be the answer.