Hi Avoid,

The link you posted is to the results of the ACPC. Where did you find evidence that the bots are flatting SB a lot? By looking through the logs? Do we know how the best 3-handed bots stack up against the best humans?

Have you put much thought into constructing a flatting range that is superior to always 3!? I think it's quite possible that such a range does exist, but I find it really hard to imagine how to go about putting it together without suffering post-flop against a good hand-reader.

I wrote software to convert the hand histories + have spent several thousand hours reverse engineering their fundamentals. The hard/fun part is creating methodologies useful at game speed for a human to use. FWIW the bots are not flatting the SB a "lot" it's ~10% vs a btn open while they still 3bet ~20-25%.

The cold calling ranges have lower RvR EQ than the 3betting ranges, so you should never be flatting a stronger range than the opener... But adopting a mix of both plays with well constructed ranges makes you a huge pain in the ass to play against. I find this to be true for most plays in poker. IMO, always cbetting OR always checking AQ on J84 is vastly inferior to balancing both your cbetting range + checking range to contain the hand. The same concept applies to preflop in the SB, most of the bots adopt a mixed strategy... even preflop!

The ranges for cold calling do exist, and require solid postflop play to show profits... much of the value from these hands from from the ability to thin value bet + know when you have fold equity. (Basically vs fish/postflop nits) There are also a few preflop spots in full ring LHE i will always flat in, giving the BB odds to call.

You worked 8 hours a day for nearly a year on this?

I have obsessive tendencies with a capacity for hyper focus. Been doing it for many years, pretty much every time I had a losing session for the first few years of playing. Poker has never been about the $$ for me, always just about trying to find a real solution for the game.

Like. Agree

this is interesting as balanced play indicates that you necessarily have to cold call the sb with hands that beat the bu's opening range some of the time. i.e. if you never flat JJ or QQ for example, your opponent will know you can never hold those hands, which is what most people here are taking issue with: that flatting a specific range, makes you easier to read. i.e. your 3bs will not have 98s and your flatting will not have AA.

following this logically through, it means that you're a) calling with the top of your folding range in a 3b or fold situation, and b) calling with some of your 3bing range so that you aren't easily readable.

both of these things should cost $ in expectation. as a result, the rest of the range has to be seriously more ev than you'd get if you folded the hands that you're now calling with and calling some of the hands you would otherwise be raising with.

in addition to those, if you call, you're giving the BB 5:1 to call. no matter what your hand, you do better vs. the button and in expectation overall if you only have to face 1 other hand from a 35ish% range than if you have to face 1 hand from a 35% ish range and another hand from a much wider range (without doing the range analysis i'd guess you're calling basically anything > the worst 20-30% of your range. so 70-80% of hands getting 5:1).

this is especially bad if the bb plays well.

so your above statement aren't be sufficient as a strategy, other than possibly intimating that one may exist. as a result, can you elucidate something close to what these ranges would look like:

3bing from sb:
calling from sb:

ty.

but if you're a solid postflop player, and if some of the value is from thin vbetting, those are two things already solved by incorporating some weaker (but still playable) hands from the sb. you don't need to flat things in the sb that you would potentially fold in order to get that thin value.

also, it's harder to give credit for a sb flat than a sb 3bet, which also makes it harder to win pots you wouldn't have won otherwise. so both of your arguments have other sides.

in addition to all that, you're easier to read if your ranges aren't balanced.

such as? please give a few examples

Jeez 9to5 you are giving this stuff away?!

On 2nd thought, you're right. I'm going to ask bbb to delete this post. o.0

I try to stay away from getting too deep into stuff usually... sometimes I get carried away =P it's why i don't post so much tbch

great post 9to5. and ty for the detailed explanation. i play in a game w/ a few solid players, and generally hilarious people. so getting an opportunity to play more against the hilarious people is certainly attractive.

and you actually got to my next question a bit early with this:

i.e. if you're going to flat the sb for these reasons, then flatting in other positions should also be considered, right? and this extends to your other examples at the end where all of a sudden you're flatting 2 or 3 bets some decent % of the time to even a solid player's EP raise.

so the issue i have with this is that now we're talking about a lot of money to be giving up in order to make yourself hard to read. in the "never 4bing" example (assuming nobody behind you plays, so 3 players), you're giving up at least 1.5bets pf*your RvR equity. if you're considering 4bing, that's probably around .5-.8 bets. that's a LOT to give up so that you're harder to read when at least one person in the pot is very likely to not be a solid player.

and you give up the chance to act last on the flop, which also has some value.

that isn't to say you're wrong as i can clearly see why it's important to be hard to read. we have a 2p2er in our game who is incredibly easy to play against b/c he is the quintessential ubertag. no matter what postflop balancing he does, it's too easy to know where he's at. this does get the money vs. the hilarious players i mentioned above, but it leaves $ on the table for sure, esp in late position/blind situations. i LOVE having him in the CO/BU when i'm in the bb b/c he folds way too much, for example.

in terms of cold calling though, you said:

i tried to get to this and even w/ this range: 88+,A9s+,KQs,QJs,JTs,87s,76s,ATo+,KQo

it's only 10.5%, so i'm not sure how a pro would open 13% utg. ignoring that for now though, even if you flat that ENTIRE range (so you hold that same range and are calling even the low end of it so that you let other players in), you still have a fairly narrow range. even if you adjust a bit and tighten it slightly (while adding more suited hands) to something like: 88+,A7s+,KQs,QJs,JTs,T9s,87s,76s,AJo+, your range is narrow(ish). you can add more pairs i guess to try to get it wider again, but generally, that's still around a 10-11% range.

and you're again sacrificing the ability to win on the flop w/ a single bet (that definitely has value) by not 3betting, again with the goal of being harder to read.

so why would "never X+1b'ing" vs. either utgraise or ep raise + mp 3bet be optimal given the losses we're talking about both pf and otf? even balancing can't make that up though (which is what my gut says so i can be wrong)

How do you know whether the bot is any good?

lol i want to make a post so great that it has to be deleted

Hi armor32:

According to Equilab, if the CO opener has a 50 percent range, your jack-ten suited will only win 48.3 percent of the time against his range in a hot and cold simulation, and against a 30 percent range, the jack-ten suited only wins 43.6 percent of the time. Do you still want to play it?

Best wishes,
Mason

The problem with a mixed strategy here is that you'll frequently be giving the big blind a cheap play and there will now be many hands that he'll be correct to play. If you somehow knew that the big blind was going to fold, then a mixed strategy from the small blind makes more sense. I refer you to The Intelligent Poker Player by Philip Newall for more discussion.

Now don't get me wrong, I'm not saying that what you are suggesting is not the best strategy, but I agree with the thinking that playing a simpler strategy, which is almost as good as the best but more complex strategy, is often better.

Best wishes,
Mason

We often have the dead BB overlay (investing 3 sb in 7 sb pot w/ 43,6% equity is fine), plus the fold equity we have when behind is nice. I three bet this and don't feel bad in doing so.

yes

Hi jdr0317:

I'm going to back out of this to some extent. There's another reason to reraise with the JTs and that's even if the play has a small negative expectation, it can help you do better with your other stronger reraising hands. This would be true in this spot as well as many other situations that come up. So this may, in my mind, swing it to a reraise.

Best wishes,
Mason

Totally agreed. I think all considerations (improved expectation on strong hands, fold equity against a wide/weak range, ability to hit various board textures, discount on the three bet) make this a 3 bet.

Hi jdr0317:

There's something that needs to be pointed out for our less experienced readers and posters, and it's that I'm playing live, not at the highest limits which have a higher percentage of tough players, and many of the players I'm against have no familiarity with me. This means that the idea of "improved expectation on strong hands" is not that important in my play since many of my opponents will not be affected by these type of plays. On the other hand, if you're playing against tougher opponents and/or a small group of regular players, then this idea of "improved expectation on strong hands" becomes more worthwhile.

Playing against many players who have no familiarity with me may be more unique to Las Vegas than any place else due to the large number of visitors that come here. So the best strategy for me may be somewhat different than the best strategy for you.

Best wishes,
Mason

Hi Mason,

Read more. Pontificate less.

Best wishes,
thesilverbail

meh.

Whilst good advice, if followed by all posters these forums would die.