Leading Flops In 3-Bet Pots
Posted by Richard Gryko
Video Player is loading.
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Resume Video
Start from Beginning
Watch Video
10
Loading 18 Comments...
Excellent debut and great addition to the RIO team! Really interesting topic you introduced and highly informative.
I‘d vote for a continuation of the series, especially dissecting turn play. My first instincts after watching the video were „ya , leading Flops is a great idea... but what the heck am I gonna do on most of the turns that don’t improve my marginal hands???”
Also diving into IP’s strategy facing the lead would be interesting, like how often do we have to calldown with AA or KK on the QQ6r board vs an optimal strategy.
As a last question:
Could you elaborate how an observant player could exploit in practice a strategy of not leading AQ/KQ on the QQ6 ?
I'm sure Richard has a good answer to the last question as I was pondering this for a few minutes myself, but I would imagine if we on the button know that our opponent isn't leading AQ/KQ for value, then we could use our blockers to his most likely QX hands and counter his lead with a float or raise with say KJT9ds or JT9Xds with backdoor straight and/or FD's since his CO range that contains QX will likely contain a lot of queen high rundowns. Purely speculating, but I would consider that kind of combo in my defending range in some manner.
Hey, thanks very much, glad you enjoyed. Re your questions (and this applies to every question that might get asked in-thread) if my response seems abrupt or incomplete, its likely bc I intend to address it more comprehensively in a future video. So, the turn issue is definitely a question thats gonna get an abrupt response :p, but in general id imagine we're going to polarise more on turns so our marginal hands likely do quite a lot of chking, since SPR is pretty low and we led a bunch of strong value combos on the flop, we can have a chking range that can counter an IP who bets at too high a frequency pretty easily. In the T96r case, on bricks our nut equity advantage increases (we led higher freq strong value than our oppt flatted a lead with) so we get to barrel quite a lot, on nut-changing turns our strategy will obv depend a ton on how equity distributions have shifted.
As to the AQ/KQ on QQ6 - Nick covered the flop side of it, also IP can opt to defend vs b/b/b based on blockers rather than linearly or using randomisation and convert some theoretically roughly breakeven bluff catchers into clearly (if slightly) +/-ev river calls (i.e. if riv bluff catching range is [TT,JJ,KK,AA] and OOP bets pot, he can use informational edge and call [JJ,TT] region.
Really good video, congrats. One question: Whats the EV difference when we employ a 100% oop check strategey?
Hey, thanks. For the handful of comparison sims I ran, EV loss in equilibrium environment with donk/no donk is relatively small (1-3% would be a reasonable range, I'm away at a live event atm and don't have access to my server so can't confirm, but I'm pretty sure most fit within those limits). Tbh I was slightly discouraged initially when I ran those comparison sims, but then I thought
-poker in 2018 is a game of small edges
-3bet pots are large enough that even a couple of % is meaningful over time
-value is predicated upon an optimal response to our strategy; in a real world environment our EVgain, at least at this time, will almost certainly exceed equilibrium value
Pretty good video. I appreciate how you use the solver to demonstrate principles many players might find counterintuitive such as leading with a reasonable frequency on paired boards. You also do a pretty good job finding reasons to explain those frequencies in human terms. Along those lines it would certainly be interesting if you continued these hands along future streets and demonstrate how villians’ potential misadjustments (e.g. raising the lead too frequently) might cost them EV.
Hi, thanks, I'm glad that you found the explanations clear. Yep, turn play having led seems to be an early favourite for continuation of the series, I'll obv keep that in mind.
Very nice video Richard. I found it very interesting as I am currently not working with a solver, but have tried to experiment with constructing leading ranges in these types of pots, so it may be time to start considering that as a new path. Thanks!
Hey, thanks - yep, solvers are, for better or worse, gonna be a huge part of the next stage of PLO's evolution. They can definitely lead you down a bunch of rabbit holes, but there are also a bunch of eureka moments just waiting to be uncovered.
Excellent video. I think oop player has too much information to prepare his range (he can face only 50% raise). Does oop leading frequency depends a lot on flop/turn sizings?
To an extent, but within a reasonable range frequency shifts in response to diff sizing options within a typical range are usually pretty minimal, changing sizings tends to impact range construction more than overall frequencies (i.e. OOP might chk to chkcall hands that would be the bottom of their b/c range vs 50% raise if IP had larger raise sizings open to it, IP would likely trim some of the more depolarised raises in a 50% range like, on T96r, AATx bdnfd), also in many cases Monker as IP will pretty much ignore the larger sizing options even if they were to be made available.
Still looking for more feedback on structure of a potential pt2 - anything ppl want to see addressed/expanded upon other than potential turn spots?
Great debut! Very very good video, like really nothing to criticize at all and lots of insights.
I have a few questions:
- Did you run any of the sims with more accuracy (like ~ 50 I/N) and try to see, whether there are any huge shifts? 20 seems kind of not much
- Did you try and run it for any of the sims (QQ6 would be a good candidate obv) with even smaller sizings? If yes, where was the threshold?
- Did you run any sims with nodelocked ranges to see what monker would do against probable population tendency reactions to the leads (way too much raising with the nutted parts of range)?
I guess over the course of the next few months 3b pots will turn into a jungle :)
Hi, thanks, great to hear :)
-prior to running these sims i ran a few boards for each scenario with even more sizings (albeit to a lowish I/N count :p), when 25/33 were both available the smaller sizing was often preferred but the solver mixed between both, so I decided to simplify to 33 for the sims that I ran bc theoretical evloss would be zero and if solver was mixing between 25/33 in some cases and 33/50 in others I (and again, at the time this was me as a player rather than a coach) would rather keep my strategy simple to execute and use 33 across the board. def something to look into and include results in future parts if they yield anything of interest.
-im in the process of running some sims along these lines and as with the above will include any noteworthy results in future content, but tbh exploits along those lines are usually pretty intuitive (ie, in your example monker as OOP would barrel more aggressively on future streets vs IP flop calls than in the baseline sim).
haha, twas ever thus, going back a couple yrs it used to feel like half the regs on Tuesday would be trying to implement concepts from whatever video PG put out on Monday :p
Amazing. I would love to learn more about what the OOP check ranges look like on some of these boards that have a reasonably high lead %. Intuitively, is there a good way to think about these ranges? Obviously you’d expect a high cbet% from an observant villain and you’d need to respond with high cr range, but what does that range look like? Do you just occasionally mix some very strong hands into that range or is it more balanced?
Hi, those questions have a broad enough scope that they're definitely something to address in pt2 (coming soon to a computer near you) and they kinda take us into the realm of converting monker outputs to real world strategies, which obv has a subjective element and precision/simplicity trade-offs. In terms of what a chkraise range typically looks like - id say, value heavy but with more bluffs than the population current have, value component not particularly linear/equity driven - it kinda just sets a threshold above which every hand is strong enough to theoretically chkraise for value, and then picks its actual chkraise range based on other factors like unblocking IP continues and "playability", bluff component varies a ton by texture but usually some mix of blocks top of IP range/retains eq vs IP b/c/strongest hands that cant profitably chkcall.
This is my first day on RIO Elite, and this is the first video I've watched. It was excellent, easy to follow, and well presented. Learned tons.
Great to hear, thank you and welcome :)
Richard Gryko 26:50 it would be interesting to see the increased OOP turn barreling range given a lock on IP to raise nut hands otf as you discuss here.
i would just imagine on blanks it would be the same types of hands that are high freq flop leads on blanks, and then likely shut down with those same hands on high straightening cards, and limit our blufs to those interacting with KQJ on the nut straight changing cards
Be the first to add a comment
You must upgrade your account to leave a comment.