OK, I have time to add some thoughts of my own.
This, to me, is the key part of the post that I linked and recommended:
Quote:
Originally Posted by SABR42
When everyone has deep stacks, AA gains more value, because, in a multi-way pot, if you flop top set, you can win a ton of money. Also, you can get tons of money in pre vs KK/AK. And there's more room to bet/fold and get away from AA in single-raised pots.
AA plays best when either short or very deep (where making the nuts becomes more important, and AA is the nuts pre-flop, and can make top set on the flop which is often the nuts). AA can play awkwardly with medium stacks, where it's easy to get outflopped but tough to get away from your hand because of low SPR's and the fact that villains can be shoving worse (but usually have better).
I think the bolded basically is the explanation for what this debate is trying to be about.
Some people are claiming that the reason we want to use small raise sizing is, to paraphrase, because it keeps our "postflop maneuverability" high. We don't get in spots where we feel like we have to double people up when we have a pair (forget AA specifically, this idea is for all hands that can flop good 1 pair hands).
The thing a lot of these arguments are missing is that this depends on stack size!
So my understanding is that, in theory, there are 3 important reasons for small raise sizing:
1. We lay ourselves a good price to win the blinds.
2. We are more profitable when we get 3bet.
3. We keep the SPR high for postflop, which helps us.
The first two reasons drop in importance in most LLSNL games, where getting no action preflop is extremely rare and people call too much and don't 3bet as much as they should. That makes things more about playing postflop.
So let's imagine that if we were playing 1/2 with $200 stacks, you think the optimal raise size is a min-raise to $4. That's 2% of the stacks.
Now let's imagine that instead of $200, stacks are $600. Is the optimal raise sizing still $4, which is now 0.67% of the stacks? I don't think it is.
Let's say that when stacks were $200, you were getting stacks in at some point in the hand X% of the time. If you triple the stack sizes, you're going to really lower the value of X! It won't be 1/3 of what it was, it'll probably be more than that, but only slightly more. Now when you have a big hand and you wish you could play for stacks, you won't always get it all. That means your "theoretically correct" small open sizes could still be profitable for you, but they're not allowing you to get the full advantage of playing deep.
The way to fix this problem is to start raising bigger preflop.
Just as a thought experiment, imagine that we're sitting $1k deep with 2 people, we raise to $15, they both call, we flop a boat and we extract about $500 each from our opponents. (This is obviously not an original thought experiment--check the link in the OP.) Without knowing anything about how the hand was played postflop, we probably could figure that doubling our preflop raise size, and letting the hand play out the same way otherwise, would have let us stack them instead of leaving them with money left over.
In short, it seems to me like in a typical LLSNL lineup, when stacks are so deep that it's hard to get all-in either way, increasing our preflop raise sizing would roughly proportionally increase our profitability from the hand.
Now the first part of that is important. You can call it an "exploit" if you want. It's tailored for games where people will call our raise, but not 3bet; otherwise we have to figure in the cost of getting blown off our hand, or the fact that now the blinds are a smaller percentage of our raise size. But in games where we don't care about those things, I don't think the theoretical reasoning for why you're supposed to raise small fully applies.