So I have noticed that almost all of the RIO coaches choose not to open much of the bottom of the GTO pre-flop ranges from UTG-CO. Why is this? It seems likes excellent players would be able to overachieve with these near zero EV hands since their opponents are going to play much worse than a solver. Some of the hands I am talking about:

UTG - A3s, A6s, Q9s, K8s, KJo, ATo
HJ - K6s, K7s, Q8s, KTo, QJo
CO - A8o, A5o, Q6s, Q5s, K9o, J9o, A5o

Side note - Some of them do tend to play the suited connectors 98s-54s even when they are mixed at near full frequency.

The explanation I usually hear is that top players have been unable to produce positive results with these hands over large database samples. This sounds like a good reason, but I wonder why top players are unable to generate a positive winrate with these hands when a solver can do it playing against itself?

Fwiw, I have chosen to stubbornly play these hands to see for myself if I can turn a profit, but I am also slightly losing with them over my last 200k hands.

Why does this happen? Are the hands I listed just too difficult for humans to play well? And do suited connectors tend to overperform against human opponents?

I would love to hear your thoughts.

Well looking at GTO Wizard, I believe most of those hands are listed at close to 0 EV to open. So I don’t think the solver is generating a big profit on them either. I’ve tried to not open hands or 3 bet hands that the solver has listed as essentially 0 EV. I figure if they really don’t make a profit with them I won’t be haha.

Most players at micros call opens way too wide IP, which decreases the EV of those mediocre opens.

Also, even if the open does actually make you some small profit on average, it is often still more profitable for you to just fold:

I've noticed this as well. Some of it I think is an old school mentality. The play you described, folding ATo UTG but opening 98s, is what pre-solver preflop play looked like. So for people who have been playing and winning for many years with that preflop strategy locked in, it might be hard to break. Solvers definitely seem to prefer big cards over board coverage hands like smaller suited connectors.

The other theory is that having slightly losing hands in your preflop range makes the rest of the hands your range more profitable vs. the right opponents. I have no idea how to test this theory, but it's definitely plausible. I think someone who opens the full range of preflop hands is going to end up with a VPIP at 6max of around 25. Someone who opts for a more traditional tighter approach may end up around 20%. If having the higher VPIP means the bad regs give you more action when you are strong, then it could offset the losses that come from playing some of the weaker hands. I haven't played with any preflop solvers, but are the ranges based on all hands that have a positive EV, or is is the hands that give the strongest range EV?

Also if you are multitabling, turning down a 0EV spot is probably worth it from a mental energy standpoint. It can be quite taxing to play multiple tables at once, so folding more in 0EV spots can allow you to shift your focus to a postflop spot at another table.

Other answers here are good.

It is still weird to me. I would expect zero EV hands in solverland to be quite profitable for good human players who are up against weaker players.

That is interesting! So bad players who cold-call too much in position are actually punishing the rock bottom our range when they do this. I guess these hands rely heavily on taking down the blinds a lot, but suffer when this doesn't happen. I imagine that multiway pots are especially bad for these hands. And yes, excellent point about these hands probably just being a waste of time... especially in fast fold games which I play lol.

As far as the suited connectors being played due to an old school mentality, I have found that they are performing well in my 200k sample. I feel like they might just have excellent anti-human properties. Like a solver is going to defend perfectly against semi-bluffs, but maybe this is difficult for humans?

Solvers never play any hands that are below zero EV. They never play hands that are -.02 BB in EV to add value to the rest of the range, for example. You can see this in the GTO Wizard ranges. But in practice, it seems like this would work okay. It surprises me that LAGATG and LAG play just isn't a thing anymore. It seems like an otherwise excellent player playing 28 VPIP could generate a lot of extra action for his premium hands. Maybe with everyone trying to play so GTO nowadays, this sort of strategy just isn't effective anymore.

I like the idea of trying to exhaust my human opponents by constantly wearing them down with LAGTAG play. It's what I always did in LHE. But maybe NL players are just too good for this type of thing. Not to mention, it is likely totally worthless in my anonymous fast fold pool lol.

It still weirds me out that these zero EV solver hands don't generate significant profits in the hands of a strong player who is playing in soft online games. Zkesic's explanation makes some sense to me, but it is still surprising.

It's mostly fish putting us in multiway pots with hands that do poorly multiway like 2 offsuit high cards.
Also, when fish is OOP relative to us, regs will CC wider IP, and playing HU vs a reg with a somewhat tight range while being OOP and holding a hand that plays bad post is also bad.

Offsuit high cards are good in environments where you face a bunch of 3bets and not many calls, since they block them

I think the idea is that those hands become -EV if population calls too much, but our stronger hands benefit from them calling too much, hence why it’s not exploitable. It’s similar to how bluffing vs someone without a fold button is -EV, but the EV of our overall strategy will increSe

I think it also depends on what kind of PF chart they follow. Depending on parameters of sim you can get different solutions esp for those borderline hands.
(I don't know much about PF sims so this next part is speculation) if you run sim with no cc ranges for MP-CO then big card are better because they block 3b ranges and SC unblock them. So if 67s and ATo have similar EVs, on table where ppl play 3b/fold ATo will face less 3bs.

That makes a lot of sense to me, thanks! That pretty much convinces me to stop trying to force a profit out of some of these hands.

Cool, that makes a lot of sense too. I am essentially making the mistake of trying to bluff a fish over and over when I keep opening these high card, blocker hands.

That sounds quite logical, thanks man! I only really know the GTO Wizard charts myself. But I notice that the RIO elite coaches play so similarly to the GTO Wizard charts that I have never really bothered to seek other sources. I would eventually like to see some 3x sims though. I think I might defend too tight in the BB against a 3x.

You could make a strong argument that those 0EV opens increase the value of the top of your range. It's similar to the idea that if you stop bluffing then the EV of your nutted hands decreases, as people can stop calling.

Here's the EV of a LJ open at 50NL according to GTO Wizard



About 1/3 of these hands are ~0EV opens. Here's what a no-frills LJ opening strat looks like:



Assuming your actions can be tracked, if you start opening 12% UTG people are gonna give you a lot less action. If you don't bluff your value doesn't get paid.

----

GTO Wizard has solutions (Simple) where every player outside the blinds uses a 3bet/fold strategy. Interestingly, LJ gets to open about 2% wider overall, and the EV of hands like ATo increases significantly.

Interestingly if you were to open at the approximate frequencies, the overall would be around 17.6%

If you just simplified and went 50/50 with all of the hands that mix. It only goes up to 18.3% or about an extra 10 combos.

You could also mix them 25/75, and the overall frequency would go down to 16.4% (-14 combos).

If you expect loose calls preflop you should cut some opens with small suited cards, but not with offsuit medium cards IMO, we want to open 67s against good player for board covrage and "good" bluff to balance our sets on drawy boards, but against weak players you want hands like ato, k9o, just to make top pair or good 2nd pair and get 2 streets of value, but those hands are useless against regs, as cant defend vs 3bets and play badly post against proper defending ranges. I cant imagine not opening kto or qjo on hj if i expect fish to call, even if fish have position, and vs blinds defending fish i would rather open wider than solver than tighter, so even junk like a9o, k9o qto goes into open, if regs start to punish this with 3bets i stop, but regs are usually handcuffed by loose fish or shortstacker too, so in practice you can get away with this, but this is of course experiance of usual nl25 table with at least 1 donkey, not sure this scale up with low stakes.

I dunno, call me old-fashioned but I'd much rather be opening hands like 87s and 55 which are either "go hard or go home" hands after the flop, than something marginal/junky like ATo or KJo where you're liable to either win a small pot or lose a big one. This whole argument seems to be a perfect example of why half the GTO arguments are flawed in any event - there are other things to think about as well, notably fatigue - if you're battling like crazy to break even in marginal spots ("cos balance") then you're probably not going to be playing optimally on your big/high-equity/high EV spots. Rake, too.