The thing is that AK still 4-bets preflop maybe 50% of the time here, so if I count combos I lose to 4 combos of 44-66 and 2 combos of KQs and I win against 6 combos of AK.
That's a breakeven value-bet if he calls AK+ on the river.

Basically he needs to have and call KJs, KTs at high frequencies here, which has been traditionally unlikely in this spot.

So I take an edge case and hope that he still jams AK+ and now adds some AsJs, AsTs combos into the bluffing region -- now I win 8 times and lose 6.

By the way, if he's really a nit, check/fold is probably best here.

1st IP strategy vs shove 2nd IP vs check

IP need to defend every Kx and some Qx. If you assumptions correct than we can print money by shoving bluffs. ( In theory our hand 2bb better shove)
and 2nd - he actually doesnt have many bluffs (JT-98s mostly folding pre) thats why Im spectical about checking. Prefer shoving and get over with it, but I see your reasons and certainly could be higher EV

I agree completely Jeff. Basically, I don't think this river much resembles the PIO-sim in practice. Kxs is pure-folded here by a decent number of players.

You can print money with bluffs here :). People don't like bluffing or paying off in this spot. Marginal decision for $X,XXX dollars is not something people like doing.

You make an excellent point, especially on low cards like 3c, I'd block his folding hands of AhQh, AsQs, KsQs, KhQh, QhJh, KhJh, QsJs and KsJs.

However with the exception of AQ, all those other hands are now in his turn calling range, so my blockers have the opposite effect that they would on blank.

The triple barrel is much closer in value. I think PIO might end with a slight edge to this combo on a complete blank, but it would never be very valuable.

Thanks SoundSpeed!

I tend to simplify my flop play in most situations, because it's a small loss of value at equilibrium (like 1% of pot) and let's me play turn and river more correctly and with better understanding of my range. It's something I've done for many years and I think that the better understanding on turn and river makes up for the loss of EV on the flop.

18:35, I thought about that too, but I think that this requires him to fractionally call 9x and 88s on river and I don't think that most players make those calls. The data supports this, btw.

27:35: Basically no-one is folding an open ender + overcard in a 3-bet pot. PIO models the call at generally around +10bbs.

Hi Peter!

I appreciate your feedback. I do a variety of different videos for different subscribers and I have several more recent theory oriented videos: "Time to do some Hardwork" and "LucidGTO C-Bet strategies".

I feel like your principle criticism of my play is that it doesn't match the PIO sims that you've created. However, my job is to make the most amount of money playing poker, PIO doesn't guarantee this, it guarantees that I lose the least amount of money against the best possible counter strategy. My opponents aren't generally machines, so I can make more money by taking lines that maximize their mistakes rather than assuming that they are omniscient computer programs.. Here's some followup on this:

1) My flop play is a simplified strategy, I didn't discuss it, because I thought it was common knowledge. If I block bet 100% on every non-monotone board, then my strategy is less than 1% worse than the complicated strategy. The thing is that complicated strategy is murderously difficult to execute, so for 1% less value, I get a fighting chance to play the turn and river closer to optimally.

2) Preflop ranges here matter a lot! In the AQo hand, that caller might be playing a strategy of exclusively calling pocket pairs with the intention of folding any non-set hand to a triple barrel. Is it exploitable? Yes! but the exploitation is to do things like fractionally call AQo on the river, and barrel a lot of bluffs. PIO would never do this, because the correct answer is to now call all 99 and 88s on the river, but this happens rarely (never) in practice.

3) My preflop ranges are proprietary. If I gave everyone who watched this videos, my exact preflop strategy, then they could run perfect PIO sims against me or copy the ranges exactly, cutting down my edge. There's plenty of good resources available on preflop-play: Pokersnowie, Monkersolver and Peter Clarke's course are some examples.

4) Just because PIO calls a hand at pure frequency, like JTs, it doesn't make it a profitable call. In fact in the sim, I believe JTs is worth roughly 0 chips as the call. The JTs is actually being called to prevent some exploitation that the solver spotted. The thing is that it's not clear what the exploitation is on this board. Is it 3-barrelling 99 on a T turn or is it to take advantage of player who check-folds too often on blanks (Maybe AJ here can be pure folded without JTs-type hands in range)? Does that mistake exist in practice? If it does then JTs is profitable and if the player is computer it's breakeven and if he's overly aggressive (a common mistake in 3-bet pots) then it's -EV. I chose to mention that being overly aggressive in 3-bet pots is a common player pool mistake.

Finally the 54s hand, I discussed this with Jeff in the comments above. Yes PIO jams 100% with 54s here, but PIO also calls Qx fractionally and KJs-KTs at pure frequency. A common mistake is to fold Qx and call KJ-KTs fractionally. With those assumptions it switches to a check-call.

Yeah I agree 100% about the part of pio not being the best or optimal strategy vs population in most scenarios. I just think that we would have a way more productive discution and higher quality content making a more in depth analysis of the spots, taking a look at how it should play out in theory and most importantly how people will deviate from PIO responses and the best way for us to exploit that.

For example saying like you mentioned on the comment that you use a simplified strategy on ''x'' board and just bet high frequency small sizing instead of spliting between b75 and check like pio does because people will not defend properly vs that and our EV loss on this simplified strategy won't be significant enough for us to worry about, already generates a way higher quality content imo than just saying ''We go for a small bet here'' in a spot where small bets are not really used in theory and jumping into the turn.

For what is worth I nodelocked the 854r board and forced OOP to rangebet flop for 33% sizing. On the previous sim OOP EV was 117.33 and on the nodelocked version it droped down to 112.02, which is quite a huge EV loss for us to simplify the strategy in my opinion.

I did also watch both of the previous videos that you mentioned and enjoyed them btw!

Hopefuly I made my points clear and looking forward to more content :)

The intention of this video was to examine turn double barrel spots in 3 bet pots, which I'm sure is why Tyler speaks only briefly about preflop strategy, flop ranges, C betting strategy, and sizings in the majority of the hands reviewed. Running sims for every hand and examining preflop strategies, flop ranges in detail, and breaking down correct flop C betting strategies and sizes would have greatly limited the number of turn double barrel spots that could have been covered, thus defeating the purpose of the video.

That said, I agree with some of what you wrote. I was surprised to see that Tyler seemingly C bets entire range for 1/3 in 3 bet pots regardless of positions, ranges, and flop texture. I'd be really curious to hear Tyler's thoughts on this strategy and why he does it - exploit vs pool, thinks the difference in EV is small and wants to simplify his strategy, likes to keep ranges extra wide to create more bluff/barreling spots on later streets, or? I was definitely surprised by some of the turn double barrels as well as some clearly weren't good. That said, some of those loose/-EV turn barrels became good bluff spots when the river blanked and Tyler did take advantage by firing the third barrel all in.

All in all I liked the format, concept, and the spots covered. Having the turn solutions - as you posted above - would definitely have added value to the video and allowed viewers, and Tyler, the opportunity to compare and contrast Tyler's assumptions and the way the hands were played to what solver says is optimal.

4) Just because PIO calls a hand at pure frequency, like JTs, it
doesn't make it a profitable call. In fact in the sim, I believe JTs
is worth roughly 0 chips as the call. The JTs is actually being called
to prevent some exploitation that the solver spotted. The thing is
that it's not clear what the exploitation is on this board. Is it
3-barrelling 99 on a T turn or is it to take advantage of player who
check-folds too often on blanks (Maybe AJ here can be pure folded
without JTs-type hands in range)? Does that mistake exist in practice?
If it does then JTs is profitable and if the player is computer it's
breakeven and if he's overly aggressive (a common mistake in 3-bet
pots) then it's -EV. I chose to mention that being overly aggressive
in 3-bet pots is a common player pool mistake.

Hi Tyler Forrester This response really helped me and I'm HAPPY you took the time to respond to the long comment. This is a common error of mine where I'll call 3bets with a lot of suited broadways, then I'll float BDSD BDFD and hope they have AK and don't barrel the turn. There are been many of times where I floated the flop and turned a T, only to stick around even further and end up losing to QQ+ region and talk myself into calling, like how can I call flop and fold top pair on the turn? Maybe they are bluffing with AK? I think 80% of the time though it ends up being QQ+ because the pool also has a flat call range with AK and are not 3 betting AK at 100%.

Anyways just wanted to thank you for this response. Really helpful! I was going to respond to his comment as well, basically what you said because it does seem he wants to mimic PIO as much as possible and not trying to make the most money via exploits. Your break down much better than mine :-)

If you can execute the flop mixed strategy correctly while multitabling (including the 18-20 nodes on each of 1100 turn and rivers (some are isomorphic)) , then yes, you'll make an extra 5 dollars on this board and I also suspect we'll be seeing you battling at Stars 5K and refusing to make videos because you don't want to give away your secrets.

On this point, a lot of the reason I don't dive super deeply into some of the PIO sims, ranges, my choice of approximations is that I still make a living player poker. My player pool watches these videos and if I'm consistently giving away my entire strategy, then my player pool is going to play very well against me. "Oh Tyler Forrester likes to bet "XX and XX here" I know exactly how he plays here, so I'll just make hero call or hero fold.

Worm

Thanks for the feedback! I always appreciate your input into how I could make better videos :).

Basically you can approximate the game with a simpler flop strategy without a major loss in value. the simplification cuts my game tree from roughly 2 gigabytes worth of data to 1 gigabyte. It's still gigantic, but it's twice as easy to understand. I've done some work on the best approximation trade-off strategies in 3-bet pots and small bets I believe are some of the least costly. However, on certain boards like monotone or 3-straights (especially low) then I need to change the approximation, but on many boards I believe the tradeoff between betting small and playing a complicated mix strategy is a reasonable choice.

RunItTw1ce
Your comment was great! I completely understand the frustration of making our hand then losing. Hopefully, you can find some situations where you can figure out how to make a 0 EV call profitable :).

https://www.youtube.com/watch?v=P-YxF4S86Tc

Tyler Forrester
This video by Dnegs analyzing some showdown hands reminds me of floating JTs on 8xx and turning a pair LOL. He definitely needs to pay Eliot for some mental game coaching.

Really depends on your edge -- You need bigger edges to generate the same winrate at each stakes because of rake, It's 6bb/100 at 1/2NL, 4.5bb/100 at 2-5, and 2.4bb/100 at 5/10, so to maintain a constant winrate you need to win substantially less money at 5/10 per 100 against the player pool.

I suspect 15-20bb/100 is the upperbound for winrate, but most players end up between 1-4bb/100.

Was thinking 100NL - 200NL 10-15bb on ignition is possible. At $2/$5 I don't have experience online there but Imagine something like 3-7bb/100. Then $5/10 makes sense for it to be 1-3bb/100. If you are able to 4 table, then it's roughly 250 hands an hour.

Hourly 200NL @ 10bb ($50/hr)
Hourly 500NL @ 5bb ($62 / hr)
Hourly 1KNL @ 3bb ($75 / hr)

These are roughly my goals going forward. As long as mental game can match the lower win rate being higher variance.

I'd check the board again. He had a double gutter with an 8 also making him a straight. My 54 combo was actually a gutter :).

Hahahaha I love it!

oh man! This would of been great if you put like JcTc on Sub Zero outfit LOL It's still great, but you know what I mean.

lmao

There's a lot of noise around the edges of this range based on the model (PokerSnowie, MonkerSolver, PioSolver, or SimplyPostflop) and opponent's 3-betting frequency and size.

On the T8s regions, I'd still play them, because the computer thinks they are super close and that if the opponent plays perfectly OOP that they will make money. I'm skeptical that the OOP is going to perform as well as GTO, because the OOP strategy is 2x-times harder here than the IP strategy.

I think PokerSnowie's ranges are definitely usable and you shouldn't make many large mistakes against a typical SB and BB 3-bet strategy. Be aware that if your opponent is significantly tighter (8-9%) or significantly looser (17-20%) that these ranges are going to be too wide or too tight.

I think you're right that the turn fold frequency here is going be low. I'm not sure that this precludes some barrels that bet/fold though, because then our opponent basically has perfect information about our hand -- JJ+, Axss, Add, 65. I'd prefer to leave some doubt in his mind. The other thing to think about is that if my range is that tight, then he's going to correctly overfold the roughly 30% of blank rivers.

I haven't looked specifically at 50%, but I've looked at larger sizings like 3/4 pot and 100% pot. It's pretty board texture and play dependent on who plays the best against each sizing. If you've got players who hate to fold bigger will likely make more money. If you've got nittier players who are looking for an excuse to fold the smaller sizings generally will out perform.

Honestly, I don’t think it much matters what flop approximation you choose as long you play the rest of the tree well. The 1/3rd is safe and not terribly exploitable.

Thx @gogols_nose. Do you believe this also applies to SRP both in and out of position (ie btn vs bb, ep vs lp flat)

Thanks Truepower!

It's probably okay with the 3-straight on the bottom with 550 back. I am priced in with my nut flush draws here, so his equity is going to be pretty reasonable against the calling ranges. Solver pure flats thought.

Yeah, that is just moving money back and forth. We both lose the same here.