You are viewing a single comment's thread from:
RE: Preflop Poker Guide: Hands to Raise From Each Position
I used to hate it too, and against many players that overfold it's better to 3x a very wide range. Maybe I've been clouded because it's even better in tournaments (ante games almost require limping continuing range over raising), but I've found it's the better GTO strategy for small blind than raising a low % of hands and folding away your sb too frequently.
I suspect the real solution is a mixed strategy of raising/limping, but it's really hard to balance those ranges this deep OOP, so I chose to limp everything instead.
Against players that over-defend it is better as well! When you 3x a good range from the SB, you will be in a great position against many players. A lot of players will defend far too wide, because they always suspect a steal and they don't want to be bullied. They defend against SB opens as if the SB opening range was as wide as a button opening range... and then they continue on far too many flops.
But please, let me just say how glad I am to be able to discuss poker on this level here on Steemit. It is a breath of fresh air! I'm very grateful for your contributions.
Of course, constructing ranges for out of position flop play is a difficult task. But if you have a better range to begin with, you have a little bit of a greater margin of error.
If you choose to raise from the SB, I think the range should be wider than the button open -- assuming the SB 3bets his continuing range vs a button open, the button is already at a disadvantage of only getting through 85% to the bb, and the positional advantage likely isnt enough to make up for that. Plus, the SB risks the same amount by 3x'ing as the button does by 2.5x'ing.
I included my SB raising range above as it probably is more profitable against many opponents and is the better exploitative strategy, but I still prefer the limping strategy from a GTO sense, and still think the real GTO solution probably contains some mixed strategy where you limp hands like AA, AK, small pairs, suited gappers, and enough junk that is balanced by the limp reraises and limp folds, and raise the rest of your continuing range.
Open limping your entire range would certainly be more profitable than raising wider than a reasonable button range... But that's because it's simply awful to raise wider from the SB than you would from the button.
I'm not sure if this is influenced by a focus on tournament play, or what... But it's a tough sell for a nlhe ring player. I have a hard time seeing it as anything but a technical error.
The more I work through this, the more I'm convinced that this is a significant deviation from GTO play. I think it's highly exploitable... I think it's likely exploitive in the event you are playing against weak opponents who will misplay hands post-flop... It makes sense in that light, because although you'd be out of position your opponents' range would literally be any two.
But in general, I think this is actually very bad.
Doug Polk (widely thought best HUNL and among best overall NL players in world) disagrees. Independently checked his preflop guide that he released in March and he had a tighter range from the button (40%) than me and a looser range from the sb (48.5%). Seemed insane that he had an even larger disparity between his sb raise and btn raise, but you can defend wide enough vs BB 3bets to make it work.
Janda prefers the tighter sb range of 36% and btn range of 47.5%, but his advice is from early 2013, don't think I trust it over the top HUNL player in 2016.
Pokersnowie gives the answer I expected: it recommends a mixed strategy. It advises to raise 38% from the button, and recommends playing a ~58% range from the sb but lots of hands are limped with others raised. This is about what I thought would be the GTO solution, but it's complicated as fuck:
Green is raise, yellow is limp, yellow with writing is to raise that % of the time.
Looks like it wants us to raise roughly 20-25% and to limp roughly 30-35%. Overall can play 10-20% more hands by playing a mixed strategy than raising only, but 10% less hands than limping only, but is clearly the optimal way to approach sb play. Just hard for humans to follow that chart even when using suits to randomize play.
I'll have to defer to your expertise, then. I think maybe the game has passed me by.
But yeah, theoretically optimal ranges are bizarre and impossible for humans to adhere to in real time.
Please don't defer to what I say. I provided 2 ranges, then looked up 3 other ranges and all 5 were very different. I suspect the mixed strategy is what we want, but the ranges (all made by very smart people) are so dissimilar that it's clear lots more work needs to be done.
I'm relatively sure the snowie button range is too tight, so I am treating its ranges as not infallible, which means their sb range could be wrong. That said, from a game theory perspective, HU play is much simpler than multiplayer games so very likely this range is significantly more accurate than their other ranges.
This all sort of loops back to the calling question I made above. Open raise ranges are generally easy to make, same with 3bet/4bet/5bet etc, but calling requires much more guesswork. And since sb vs bb has a ton of calling....the opening range is harder to pin down.
I'm going to opt out of the sb limp explanation blog as I'm convinced the mix strat is better, but maybe I'll break down that range to see what happens when both a raise and a limp are raised or called by the BB
@daut44: responding here due to nesting limit...
"Defer to you" is probably a poor choice in words. I guess I just made my case, and I recognize that I could be wrong.
A mixed strategy is almost certainly better. And that would require weighting parts of your range... in other words, having certain holdings in both your limp and raise range. How to randomize effectively would be a good complement to the discussion.