Drelza's avatar

Drelza

42 points

This month has been off to a somewhat rough start, but I’m at least happy with how I’ve played and how much I’ve improved. Results are breakeven in big blinds and dollars, but losing a bit in EV$.

and in bb

Even though the volume looks low, I’m happy with it being on track with what I was aiming for. I should be able to get 25k hands in this month. The red line obviously looks ridiculous, perhaps a little too ridiculous, though I’m generally pretty happy and confident in the changes I’ve made in my game which I think are causing it. Will keep an eye on it… my guess is it's also a product of my bluffcatching (which I haven't changed) and just runnning a bit bad there. My river call efficiency is so low I'd have been better off check-folding literally ever river this month!

Gym/other roles have also been on track.

Study has been good too, but it’s tricky to know what is and isn’t helpful. That said, even though some of the solver-work I’ve done hasn’t been particularly practical, I find it makes me more excited to play. Going to keep it up.

Feb. 17, 2021 | 7:29 a.m.

Hey Ben, @15:33 I just noticed the 96 is mixing between betting and checking back the river (on the 964Q4 board) in LLinus' spot. I was wondering if you understood what could be going on here?

Often I've seen Pio have these kind of mergey/thin value bets being mixed on the river as IP. I think I understand in the reverse positions that OOP has some incentive to mix thin value-bets as a response to IP stabbing/vbetting thinly, but I don't understand why this would be true for IP.

Could it be that it helps prevent OOP calling with weak hands in absolute terms but good blockers (JTo :D /s) while simultaneously getting OOP to fold better hands with bad blockers (like A6/K6 here). Alternatively, is it the case that OOP is doing something with its betting/checking range that retsrains IP from always betting 96s (though I'm not sure how this would/could work?).

Thanks!

Feb. 15, 2021 | 12:53 p.m.

Post | Drelza posted in Chatter: Pursuing winrate at midstakes plus

(This is cross-posted from 2+2 but the threads may end up going separate directions)

Learning how to try hard and maximising winrate at nl500-nl2k
I’m a mid-stakes grinder from Australia, playing mostly nl500-nl1k (with some nl2k mixed in recently) and also a full-time student with two part-time roles. There are a lot of legitimate and self-proclaimed sickos in this community who smash out 70k hands per month while running sims every day. That’s not me, and that’s not what this thread is gong to be. Instead, I’m going to be aiming for having the highest winrate I can achieve while managing my other responsibilities. Here are my results from nl500-nl2k last year:

And in BB

I lost (or at least can’t be bothered to find) my nl200-400 hands, as well as my 50k hand stint in PLO, when I changed to HM3. I think I had a ~4ev/bb winrate before rakeback over ~200k hands. This year, as I’m taking a lot more on outside of poker and won’t be able to get the same volume, hence the new focus on winrate.
It’s been a rough start so far, with breakeven results ev bb/100 but the losses predominantly coming from nl2k. I’m going to turn that around with:
- Each session including a warmup using the Elliott Roe tapes (highly recommend the mental game course!)
- More attention to each spot and marking hands to review
- Playing intentionally slowly and thoughtfully

I’ve never had a particularly stable or systematic approach to poker, with somewhat poor volume and inconsistent study habits. That’s gotta change this year, so with that in mind my goals are:
- Play 20k+ hands per month (4-6 tabling Ignition + Apps) with high focus and aiming for a high win-rate
- Studying at least 5 hours per week
- Gym three times per week
- Performing adequately at part-time roles
- First-class honours at Uni, with stretch goal of getting the medal for topping my cohort
Another thing I’ve thought about but always avoided is coaching. I had a lot of fear that strategy could spread, but I’m less concerned these days given the amount of information out there and the difficulty mostly coming from individuals’ ability to find the right info and executing on it. So, eventually I’ll look more into that.
Looking forward to the journey! Will post semi-regular updates with results, possibly also some productivity/volume freerolls. Might also add some rants on certain topics, maaaaybe a few hand histories if there are requests, but I’m still a little wary of sharing much strat.

Feb. 12, 2021 | 4:56 a.m.

Thanks for the content Saulo :D

I'm one of your old students (Michael, the Australian) and have been really enjoying your content. From your coaching I've been able to take the year off and now am beating nl1k on ignition/apps, taking shots at nl2k at the moment. The only thing I wish is that you'd made this video at the start of the year so I could've learned these lessons earlier on!

Cheers,
Michael

Dec. 28, 2020 | 7:32 a.m.

I heard they're releasing trips 2t boards first :(

Sept. 3, 2020 | 4:34 a.m.

Found this excellent. I also liked how you described how you would adjust vs different types of opponents. One type, however, I would be interested in hearing your thoughts on (as they are who I tend to play the most against) are players who have a sufficiently aggressive XR range, but who also overfold. I often feel the mixed incentives (to bet so they fold; to X so we don't face a XR) confusing. Are there any deviations you would make here? Or is this one of those spots where we exploit them by 'not exploiting' them, and just passively generating EV from their mistakes.

Aug. 22, 2020 | 7:49 a.m.

*not teacher.

I think in its current form, this question is a little bit vague.

On one interpretation, given you ask about unfamiliar players, you seem to be implying that there's potentially not going to be an EV gain of using a strategy recommended by Pio? While I don't think this is straightforwardly wrong, I imagine it reflects an understanding of game theory that'd be improved significantly if you watch some of Sauce's earlier theory videos.

Another interpretation of your question is: how much extra EV does there need to be in a particular strategy for us to implement it? If so, even this question isn't answerable unless you're clear about what strategy you're comparing it to. If you're comparing it to your current strategy, it's possible that you would generate a lot of extra EV/or possible you wouldn't generate much (and it's also possible there would be other parts of the game you'd be better off focusing on! It's not possible to answer this question without detailed knowledge of your game). If you mean: compared to a simplified strategy... then this question is more tractable but you probably want to be using Piosolver yourself and seeing what the EV differences is between the simplification and the higher accuracy but more complicated strategy, thinking about how realistic it is you can implement the complex strategy question, and making a decision based on that.

Hope that helps!

Aug. 17, 2020 | 3:47 a.m.

https://www.runitonce.com/poker-training/videos/phil-galfond-proview-25-plo-poker-making-you-a-winning-player-pt1/

Aug. 13, 2020 | 6:42 a.m.

Thanks :D!

Yep, that's it. My impression is that regs overfold to XR on (most) textures, whereas fish probably bet too linear and defend too much, so we want to just be going thinner for value.

Aug. 10, 2020 | 3:54 a.m.

Great vid :) like other commenters, I love the visuals.

One thing I'm curious about is how this would change in the scenario that: (a) BB does not have a leading range; (b) IP does not adjust to this (e.g. imagine if the btn cb strategy was nodelocked. I'm interested in this because that best represents the situation I'm in as I don't have a leading range + I'm 90% confident 90%+ of my won't or can't notice this.

My guess is that you'd see a greater number of bluffs being able to squeeze out a profit on the boards where you would typically expect a leading range to exist (and no effect on boards like A64 or TT2 where I assume you wouldn't) as there will just be more value combos and high equity hands in the range BB can defend with. I don't know whether it would go thinner/less thin for value though. Love to hear what you think and looking forward to the next one :)

Aug. 9, 2020 | 9:46 a.m.

Love that there are still theoretical videos to consume :D. In particular, I love that I now have a a more intimate theoretical understanding of: "when draws complete, overfold river vs badregs and fish" :P.

I've got a few questions I've always wanted to know the answer to but until now I never knew where it'd be appropriate to ask!

One very basic question that I've always been confused about: why are hands mixing when it says the bet EV > check EV? My understanding was that when hands mix, the EV bet = EV check, and that the frequency was a product of the fact that had we chose a different frequency, villain could respond in a way such that EV bet =/ EV check. As far as I remember, I think you may have said as much in previous videos // I'm quite esure Kanu said it in his Upswing course.

My second question may be inappropriate here but I'm curious how you think this would apply in PLO? As there are likely far more situations that crop up where you just can't have enough bluffs to support your range... but also, because hands generally have much more equity, there is less incentive to bluff airballs.

(*just realised the following is less of a question and more of a summary of my understanding the video, and am curious if I've 'got it').

Often in training videos, the language used to describe how people 'underbluff' often seems to imply that there are all these amazing, profitable, bluff spots that we/they should be taking! But, my understanding from this video is that generally the airball bluffs on the turn have very close to 0 EV difference between bet/check, and their function isn't really to generate much profit themselves (though they still generate some), but rather it's that they prevent our opponents from being able to exploit us by correctly overfolding and thereby lowering the EV of our value hands. My understanding is this is also generally true on rivers where we do have enough bluffs to support our value range (forcing them to call at MDF), but there are cool exceptions where we just have so many value hands that our bluffs actually get to generate profit as they begin to have to fold more than MDF. Is that right?

Aug. 8, 2020 | 2:22 a.m.

:) I'm incredibly excited! I think midstakes US games (Bodog/Ignition for example) might be the most appropriate*, as I suspect there are small winners who are elite subscribers (such as myself) who are still making fairly substantial mistakes. I get the impression those winning at 200z or 500z on stars have much more solid fundamentals.

*I also imagine you'd be even more keen to review someone playing on RIO.

Good luck in the lab! Fwiw I think you seem to underrate how your thought process, game knowledge and general poker ability, lead you to really valuable insights, even if you're not as versed in how solvers tend to play.

Aug. 5, 2020 | 1:07 a.m.

Hey Phil. I find the format of video really interesting, in particular because you distinguished between small, medium and large mistakes that were made by our hero. While I love RIO, one flaw I do see in the format is there is an abundance of information on the minutia of different spots, and it can be difficult to distinguish between those that are really important (in terms of our winrate) and less important.

This issue is potentially much more significant for essential members, but I think is still neglected for elite members, and therefore I'd love to see a video of a similar format being made analysing a midstakes midstakes player for elite members. Again, with the particular emphasis on not getting lost down rabbit holes of (interesting) but marginal situations, and as you put it, with an aggressive emphasis on plugging the most important leaks.

There was a review of Alien Slayer a while back, but after just reviewing it now, while it was a fun introduction to RIO, I think it may have been better if either/both a weaker player was chosen as the person to review and/or more tables of footage were watched with fewer pauses elaborating on the complexities of different spots, with greater focus on the largest mistakes (either actually made or potentially just those often made by regulars).

Aug. 3, 2020 | 12:28 p.m.

https://www.runitonce.com/vision/

Aug. 1, 2020 | 12:50 a.m.

Great video Phil :)!
For those curious @10:00 on AQ43, when you have A665 with the frontdoor flushdraw (both A high and 6 high variants) it prefers to XC with some XR mixed in, but it always XR with the backdoor flushdraw. Curious if anyone has any guesses as to why this is?

July 28, 2020 | 9:19 a.m.

Really enjoyed this video. Thanks!

You have a lot of intuition about population tendencies; I was wondering if you've ever done any database analysis to confirm any of these? If not, I'm curious if you think that's a useful avenue to develop exploitative lines further. (It seems highly plausible that it would be for others who have less experience with the games, but less so for you).

I also noticed in some hands your thought process emphasised things such as: "I wouldn't value bet JT here because you don't have enough bluffs to balance it, given you wouldn't bet enough JJ on the flop". I've always assumed, perhaps from a highly exploitative NLHE background, that such considerations aren't particularly relevant when playing in a moderately sized population of non-elite players (which it appeared you think Betonline is?).

More relevant are considerations like: what size will they make the biggest mistake against (in this case, calling too frequently). I'm curious if you have any heuristics which determine what situations you tend to think from a more game theory oriented mindset vs exploitative mindset? Just at a glance through this video, it appeared your earlier street (preflop especially, somewhat less so flop) considerations were very exploitative focused, whereas your turn and river thought process was much more cognisant of what you were representing/what bluffs you had. That said, sometimes you had a 'game theory' thought process which you seemed to be using to assist you in figuring out how your opponent's are thinking (e.g. with the A8 nut flush hand, table 3 51:00) you used considerations about what was being represented as highly relevant to how to exploit - so perhaps that's what's going on?

I now realise that was an abstract question, so may not generate particularly useful answers! Obviously in general wherever you have good reason to be exploitative, you ought to be, and when you don't have good knowledge of tendencies, defaulting to GTO makes sense. I suppose what I'm particularly curious about is how you tend to balance these considerations on different streets/situations.

Thanks so much! Also, really enjoyed your commentary in the challenge.

July 2, 2020 | 2:02 a.m.

Hey Phil, loving the videos! Been a huge fan of yours for a while and recently switched to PLO. Congrats on your recent successes.

I think I've noticed a thought process which you (at least occasionally have) which I'll try and describe, and I wonder if you could tell me whether you think it's a feature or a bug of how you think about PLO. The summary is that I think it seems when deciding whether to bet or check, you given significantly more weight to what a particular hand 'wants' (given its equity/blockers), than you do the base rate frequency of that particular spot. My own opinion is that this means you're likely to make more mistakes when toying with Vision, but may nevertheless be a better way of thinking about the game in play. I've expanded on this below, but that captures it (so please feel free to skip it!).

It seems your thought process works something like the following:
1) How does the range vs range dynamic of the particular situation determine what size and frequency you should be using overall?
2) Use the above as background information to inform a first principles analysis of what the hand wants to do given its qualities (blockers/equity). Take the action that this step leads to.

My own thought process while toying with Vision has been somewhat different, though in writing this, it's now apparent to me the difference is subtle, so I hope I can communicate it.

1) How does the range vs range dynamic of the particular situation determine what size and frequency you should be using overall?
2) Rather than having the above inform what action you should take in the background (for example, by using it to evaluate what the equity or blocker thresholds might be needed to bet), use it as a strong prior that your hand will take whatever action your range generally takes. I.e. if you are betting 90% of the time, unless you have exceptional reason to check this particular hand, bet it, and vice versa. When your range is closer to 50/50, give more weight to the merits of your particular hand.

I wonder if this potentially explains why, on the J535 board, you ended up playing quite differently from Vision. That said, it also seems to me that your approach might be a more useful one in game! This is because, in game, you don't have your range's overall frequencies right in front of you. My approach could potentially result in big mistakes whenever I incorrectly think my range should take a certain action at a high frequency and use that to form a strong opinion about whether my hand should bet or check (before even looking at its merits!).

Another thought is that you have had over a decade's experience of PLO; I've played about 30k hands (currently shot-taking plo1k on apps). Perhaps a good thought process for you would not be the same as it would be for me.

Thanks so much for the content!

July 1, 2020 | 2:01 a.m.

Great video, I'd be really interested in seeing a similar analysis of how turns are played out in these situations on diamond/non-diamond turns.

Sept. 22, 2018 | 12:38 a.m.

Hey Ben, enjoyed the vid - even though I don't play 2-7 Triple Draw it's still entertaining to watch.
With AA on 496K (4:30) after villain x/r you on the flop, you decided to mix between bet (presumably a bigger size) and check. What do you think about using a smaller bet of 1/3 pot with a large portion of your range? My understanding is against that sizing villain will need to do a lot of raising, which I imagine is going to be tricky to do.

Dec. 19, 2017 | 3:19 p.m.

Comment | Drelza commented on dafuq

this is such a massive overbet that you can feel fairly happy about just calling this when you have AQ yourself.. if this continues to happen though then AK is a decent hand to call down with. Also, completely separate thing but your name is familiar - did you go to grammar (if you dont know what that means just ignore this)

June 11, 2016 | 12:34 p.m.

Comment | Drelza commented on AK 3rd barrel shove?

I think when I wrote the above I assumed you shoved bc of the title

June 11, 2016 | 12:31 p.m.

Comment | Drelza commented on AK 3rd barrel shove?

my bad. Lol. I think a check is best but a shove isn't really that bad. Main issue is you block hearts so it's less likely villain has a fushdraw.

June 11, 2016 | 12:31 p.m.

Comment | Drelza commented on AK 3rd barrel shove?

I think this was played well and is a good hand to balance our AA/KK value bets. You don't block diamonds, which is good. You block him from having slow played AA/KK himself. So I think this was well played.

May 30, 2016 | 11:11 p.m.

66 makes a lot of sense here but there aren't many combos of it so I think it really comes down to how weak you think villain is. Depending on preflop ranges a two pair (that riveted a boat) also makes sense too. If you have no info on them then it's probably safer to fold because you're most likely good outcome is a chop, but if you think they're weak enough to do this with A2 or even an overpair then of course you can find a call here due to the combos of those hands.

March 14, 2016 | 10:44 p.m.

Hand History | Drelza posted in NLHE: 2NL are we ever good here?
Blinds: $0.01/$0.02 (6 Players) BN: $2.03 (Hero)
SB: $1.63
BB: $3.59
UTG: $2.00
MP: $2.01
CO: $1.94
Preflop ($0.03) Hero is BN with 8 9
UTG raises to $0.06, 2 folds, Hero calls $0.06, 2 folds
Flop ($0.15) Q 9 8
UTG checks, Hero bets $0.09, UTG calls $0.09
Turn ($0.33) Q 9 8 A
UTG checks, Hero bets $0.28, UTG calls $0.28
River ($0.89) Q 9 8 A 6
UTG bets $0.57, Hero folds
Final Pot UTG wins $0.86
Rake is $0.03

March 3, 2016 | 1:33 p.m.

Hand History | Drelza posted in NLHE: 2NL AK preflop?
Blinds: $0.01/$0.02 (6 Players) BN: $0.78
SB: $4.23
BB: $3.33 (Hero)
UTG: $3.00
MP: $2.18
CO: $3.06
Preflop ($0.03) Hero is BB with A K
2 folds, CO raises to $0.06, BN folds, SB calls $0.05, Hero raises to $0.24, CO raises to $0.78, SB folds, Hero folds
Final Pot CO wins $0.54

March 3, 2016 | 1:30 p.m.

Hand History | Drelza posted in NLHE: 2NL Top set on four flush board
Blinds: $0.01/$0.02 (6 Players) BN: $2.04
SB: $5.36
BB: $0.91
UTG: $2.53 (Hero)
MP: $2.10
CO: $1.95
Preflop ($0.03) Hero is UTG with K K
Hero raises to $0.06, 3 folds, SB calls $0.05, BB calls $0.04
Flop ($0.18) 3 K 9
SB checks, BB checks, Hero bets $0.14, SB calls $0.14, BB folds
Turn ($0.46) 3 K 9 8
SB checks, Hero checks
River ($0.46) 3 K 9 8 Q
SB bets $0.22, Hero folds
Final Pot SB wins $0.44
Rake is $0.02

March 3, 2016 | 1:23 p.m.

Load more
Runitonce.com uses cookies to give you the best experience. Learn more about our Cookie Policy