Today I did my first drill of going back to a marked hand and allocating myself a range at each decision point. I took this hand:
PokerStars Zoom, [i][color=#1515FC]Hold'em No Limit[/color][/i] - $0.50/$1.00 - 6 players
[url=https://pokeit.co/public/124MlpplX/r]Replay this hand on Pokeit[/url]
UTG: $107.88 (108 bb)
MP: $288.11 (288 bb)
CO: $127.40 (127 bb)
[b][color=#C4302B]BU (Hero): $116.69 (117 bb)[/color][/b]
[b][color=#C4302B]SB: $189.27 (189 bb)[/color][/b]
BB: $118.40 (118 bb)
[b]Pre-Flop:[/b] ($1.50) Hero is BTN with A:spade: J:heart:
[color=#666666][i]3 players fold[/i][/color], [b][color=#C51F1F]Hero raises to $2.25[/color][/b], [color=#C51F1F]SB 3-bets to $11[/color], [color=#666666][i]1 fold[/i][/color], [b][color=#C51F1F]Hero 4-bets to $24.50[/color][/b], SB calls $13.50
[b]Flop:[/b] ($50) 8:spade: K:club: A:club: color=#009B00[/color]
SB checks, [b][color=#C51F1F]Hero bets $13[/color][/b], SB calls $13
[b]Turn:[/b] ($76) K:diamond: color=#009B00[/color]
SB checks, [b][color=#C51F1F]Hero bets $14[/color][/b], SB calls $14
[b]River:[/b] ($104) 9:spade: color=#009B00[/color]
SB checks, [b]Hero checks[/b]
So I 4bet pre which is ok at a frequency, though the mix I used was not the correct application of my prefigured range. Thats fine, I only made it yesterday and intend to do a lot more work drilling preflop spots over the coming weeks.
OTF, I used a tiny sizing since it was a 4bet pot and planned to just range bet. I actually could have gone a little larger and still range bet, though my play is totally fine.
TURN. This is where I was totally wrong. I thought I had still a decent range advantage since 1/4 of my range was full houses or quads. In fact my opponent should lead almost their entire range and I should have checked back 100%.
I wildly underestimated how bad the king was for my range. Betting tiny would probably be ok with the overall equity advantage alongside a lot of hands which had my opponent drawing dead. Checking range makes sense when none of my value hands needed protection and wanted to allow for river bluffs from my opponent.
Jan. 15, 2021 | 8:05 p.m.
Holdem manager 3 doesn't calculate standard deviation which it was odd since HM2 did. Anyway I have recently moved back to holdem from PLO and am not sure what a "Typical" standard deviation might be in bb/100.
I'm playing a TAG style though leaning a touch to the aggressive side at 6max NL.
Jan. 15, 2021 | 1:04 p.m.
At 33.30 he is putting in $25 to win $20, he is in position and you will go to the flop with a capped range. I don't think you can call any two profitably here vs a min 3bet any more than you can defend any 2 in the big blind vs a min open. My guess would be that you can fold around half of your opening range without being exploited here.
I will run it in Monker.
Jan. 15, 2021 | 8:27 a.m.
Mon 11 Jan
I have been playing poker a long time and have never quite managed to break through from breakeven/slightly winning to being a solid winner at midstakes. I tend to exhibit a pattern where I am a winning player when I haven't played much lately. But when I play several days in a row my results have tended to fall off. I don't know exactly what it is that is changing about my game when my results take this negative turn, I just see patterns in my graph. Now the obvious explanation is that I am tired, less sharp or more tilt prone when I have played a lot lately. I certainly have noted that, if I play a lot of poker at stakes where the money wins and losses are significant to me, my sleep tends to suffer.
The first time I made a serious run at studying the game and playing regularly to make money, I was playing PLO. I really improved my preflop game and found I was beating $25PL zoom for a fairly decent winrate of 4 bb/100 over a 1/4 million hand sample. The only problem was that this was an EV winrate only. I ran 100 buyins under EV over that sample and ended up just breaking even. Sure, other people have had worse runs, sure I still had plenty of room to improve but as 6 months of my life melted away with nothing to show for it I kind of just gave up. I pursued other ambitions and only played poker occasionally.
This year, like so many, I find myself with additional time on my hands. I have been almost exclusively playing PLO the last few years but I stumbled on Ginge Poker on youtube and it made me want to take another look at holdem. I think his content is really good because he grounds his basic strategy in solid fundamentals but is willing to go with his reads too. Holdem has gone well and I have really enjoyed playing it again after so long so I have decided to focus on it again this year.
Now I have a lot to learn. I have not put in an serious volume since before solvers were a thing. I can see that my preflop fundamentals aren't there. My ingrained habit would be to open 87s ahead of K8s. So I need to work on that a little. Really though, the biggest part of my preflop game that needs work is responding to 3bets.
Really though, I need to develop my understanding in every spot if I am to have any hope of becoming a consistent winner. I plan to post here Monday through Friday going forward, marking my plans and progress for study.
Tuesday 12 Jan
Its interesting to be returning to holdem with a newbie mindset. I don't know how many hands I have played lifetime at holdem but I would guess that its something like 1 - 1.5 million. So in some ways I am a pretty experienced player. I have those developed instincts that only come with experience. So I am reasonably well able to look at a player's timing and bet sizing and have an idea that something about the story they are telling doesn't add up. I also know pretty well what sort of hands should be in a value betting range for a given size in most spots and have an understanding of blockers. I can identify the best value bets and the best bluffs.
What I don't have at all really is a good strategic understanding of how bet sizing relates to board textures. I tend to just bet a bit bigger with a tighter range on draw heavy boards and range bet small on all but the most unfavourable boards as the 3 bettor and basically all boards as the 4 bettor. I also don't have a strong approach to balancing my barreling frequencies. Working out a good understanding here will take some time. I favour a two pronged approach to study
1: Learn simplified opening strategy
2: Learn response to 3bet strategy vs most common sizings. What are my calls, 4bet stack offs, 4bet folds?
3: Learn 3betting strategy + response to common 4bet sizings
1: Figure out a prioritised list of lines leading to the flop. I will look at holdem manager and do some quick estimations of which spots are most common and which lead to the largest pots. So Btn open SB 3bet will be probably the 1st thing I study because its the most common way to see a 3bet pot.
2: Work out which textures favour which sizings and approximate frequencies using solvers
3: Drill coming up with simplified betting strategies which are not giving up a ton of EV when I node lock them. This technique was covered in a Sauce123 video back a few years ago. I always thought it was a great way to learn in theory but, since I played PLO, I never got the chance to try it.
1: Improve understanding of favourable and unfavourable turn cards for each player.
2: Improve understanding of bet sizing theory
3: Use 1 and 2 to generate an approach to approximating a balanced strategy and hence have a reasonably clear idea of what my river range is
1: Be able to approximate entire range
2: Use MDF principles combined with sense of whether line is over or underbluffed to decide when to call down.
3: Be able to use hand distributions to decide on which 1-2 sizings to use to bet and estimate which hands would go in each at equilibrium. Then use that as a factor in the decision of what hands to bet for what sizings.
Based on spots where I was unsure
With this approach I will either mark hands as I play that gave me trouble or do session reviews. I will manually pick my ranges for each decision point and node lock them in a solver noting the EV and the frequencies the solver takes against me. Then I will compare it to a solve where I just input preflop ranges and let it run. I will note the EV loss of the hand as I played it. If the loss is substantial, I will have another go at manually creating a nice simplified strategy which doesn't get owned too badly.
Although I will mainly be using approach 1 and working from preflop to river over the coming weeks and months, I want to spend 2-3 hours a week on approach 2. I hope that I will get to the point where I can do the exercise in about 1/2 hour and be able to do 5+ hands a week with this approach and time commitment.
Database analysis to inform population reads
This one I will not be doing for a while. But when I have played more hands I will use some empirical data to tweak my approach. I don't have enough hands yet for anything but the most common preflop spots yet though.
Wednesday 13 Jan
Last night I had a really poor night's sleep, being more than 2 hours short of my typical 7ish hours. So I decided to take it relatively easy today and to avoid playing. Taking it relatively easy means that I only did what I had specifically noted in my journal for today:
7 minute ab exercise routine - I am also new to working out. I'd done this workout once before and failed to complete all of the minute long segments with 10 seconds rest in between. So I repeated a few days ago giving myself 40 seconds on, 30 seconds off. Today I did it 45 seconds on, 25 seconds off and found it pretty easy. I'll keep going until I can do it properly...
I also did a 10k stationary cycle which is not a big challenge but I decided to combine it with meditation rather than trying to push myself with pace. I really found that a fulfilling experience. I went about 3/4 of what a pushing pace would be for me and counted my breaths which I kept as slow and deep as I could comfortably do. The time really flew by and I felt I got the best of both slightly with relaxation and stimulation.
As for poker, I had planned to do some simplification of preflop ranges. For the past couple of weeks as I played I had been using ranges from the site preflop guru. My approach to implementing them was to make every decision I was 100% sure of without checking and then pause my two zoom tables every time I was on the fence about whether a hand should be opened, 3bet or whatever. This approach was fine in terms of the fact that I wasn't often making any big errors. However those strategies are two heavily mixed to simply learn by rote an I think its good to have rote baseline preflop strategies. I can still exploitatively adjust away from them when it makes sense. But primarily using rote strategies allows me to:
Free up mental energy preflop.
Understand my flop ranges better
So today I used the preflop guru opening strategies as a base but simplified them so that I was opening everything either 100%, 50% or 25% of the time. I'll do preflop drills based on my charts 20 minutes per day until I have them all down, then check every month or so to make sure I still retain the information going forward.
Thursday 14 Jan
Why I want to post every day here
So I have been writing lengthy posts about what my plans are and how I am planning to improve my poker game over time. This is time consuming stuff and I am not sure that anyone will ever really want to tackle the wall of text I intend to generate here. So why am I doing this?
I was listening to the Joe Rogan podcast a few years back and on came Dorian Yates the body builder. If you don't know who he is (I didn't at the time) he is a 6 time winner of the sport's most prestigious title: Mr Olympia. In the world of body building, he is someone who is known to have succeeded less from genetic gifts than other body builders with similar success. Yates was known for his extreme dedication to training and diet. As he talked with Joe Rogan, he explained that he had recorded every single workout he had ever done in his journals. He did this so that he could experiment with and track the effectiveness of various training regimens. As a passing comment he also said seeing the progress over months and years in the weights he could lift or number of reps he could do was a great source of motivation.
Once I started trying to study poker, I did not find it easy. I was surprised by how difficult I found it as I generally would say I have a decently sharp mind and a really good memory for factual information. The limiting factor was not my capacity to absorb information though, it was in unsystematic in my approach. So I decided I needed to systematise and started planning my study approach but I started getting buried in the details. I would spend so long on things with such a narrow field of application that I would end up getting nowhere and losing motivation.
Even since I have started writing I can see that creating simplified balanced ranges for the most common preflop spots is actually a much larger undertaking than I thought it was. Instead of either getting discouraged or skipping on to flop bet sizing study before I am done with preflop ranges, I can see that I will be able to get finished with constructing ranges before the end of next week. Then I can learn those ranges by just spending a few minutes every day drilling them until I am satisfied (and maybe revising once a month or so after that to make sure I am not picking up bad habits). All of that will take maybe 30-50 hours, a time investment that I am pretty comfortable with.
So mainly I am posting here to try to maintain motivation and to keep track of my own plans. I honestly can't see a world in which I manage to post 200+ times this year with considered reflection on my plans and progress and do not achieve a steady winrate at $100z nl. We will see though. And if anyone does end up reading this and taking something from it, then all the better.
Friday 15 Jan
So here I am at the end of a week of journal keeping (mon-friday 40 weeks this year is my goal). I am really excited about my progress so far. Even though it has been slower than I expected in some ways, it has been concrete, organised and foundational. I will remain primarily focused on creating preflop ranges next week but then I will get to flop study with a solid platform already created.
Every weekday I will
Study/create new simplified strategies
Drill/test my ability recall/recreate strategies I am creating in a quickfire manner
Play (taking 1-2 days off from play based on energy levels)
On weekends I'll just play.
At the moment I am not too focused on putting in volume. I am more interested in working on focus and implementation of new ideas. I have only put in 2.5K hands this week so far but I intend to at least double that tonight and this weekend. Over the next couple of months I will include volume goals for play, though I think it is easy to push too hard with volume to the detriment of both life quality and bottom line.
Monday 18 Jan
The weekend did not go as well as I hoped. I lost around 6-7bi which by itself is a fairly normal swing. However I noticed that I was loosing patience in some spots too readily. I made a couple of bluffs which I still think in retrospect were sound but which got snapped off. Then I paid off a couple of hands where I really knew even before I clicked that I was making a mistake. I ended up backing away from the amount of volume I had intended to put in.
Today I got back into my study of preflop. I am fairly confident that I am not making any massive errors with my frequencies at this stage. Mainly the reason I am putting so much time right now into studying and drilling preflop is to free up mental energy. Also, though I do think about factors beyond blindly following a chart when I actually play, like everyone I can find that I autopilot a little bit when putting in a bit of volume. So I really want to get my ranges embedded into my subconscious mind. Still though I am looking forward to getting my attention on flop bet sizings next week
Jan. 11, 2021 | 12:34 p.m.
One thing about blocking board pairs is that it doesn't just block their value betting range. It also blocks the portion of their checking range which has showdown value. This skews their range more towards bluffs when you have a pair from the board and they bet. So, although 88 beats all the bluffs in villian's range I am never including it in my river calling range without a diamond
Jan. 10, 2021 | 6:15 p.m.
3bet pre. Pairs t+ an A blocker and nut suit are standard 3bets against ranges much tighter than his.
Otf, you should make up your mind before raising how you would respond to a 3b. Neither flatting nor calling is bad vs the cbet. When he pots its not a great spot but we know he makes loose aggressive plays pre and we have the 2nd nuts. Just because he didn't bluff with 5s in a previous hand doesn't mean he can't overvalue QJxx here. 37 hands is nowhere close to enough to consider folding the 2nd nuts here. Calling makes no sense either. Stick it in.
Sept. 13, 2019 | 12:49 p.m.
Thanks very much for responding so fully. Most of the points you make about solve duration and iterations align with my own tentative theories. Another tentative theory I have is:
The solver has run a flop sim for 2 hours and gotten to 10 i/n and a cbet frequency of 22%. This is still drifting downward significantly with iterations and would reach 15% if it was let run to 100 n/i.
I think that, while the frequencies are quite different, the EV leakage from the non equilibrium 10 i/n version of the sim will be small enough to be dwarfed by errors in implementation from even expert players .I think this because Monker fixes its biggest leaks first, both the 22% and the 15% ranges with be rather balanced and because the piece of hardware known as "human" is remarkably shoddily constructed and requires an upgrade urgently.
That you value studying river spots so highly caught me by surprise though. I have shied away from solver work on rivers so far because its so hard to model what ranges I show up on various rivers with, let alone my opponents. I'll play around with some river sims this weekend for sure though, see if I can find a few patterns. Thanks
Sept. 12, 2019 | 7:55 a.m.
First off I would like to say thank you for your content. I've watched videos from many coaches over the years including Phil G, sauce and Jnandez and I like yours the best by a comfortable margin.
With that said, I hope you consider yourself suitably buttered to consider the following:
Inspired by your vids, I sat down with a sims pack and a plan. I was going to study the following boards. 772,776, TT9, TT2, T22, AA6, A66. I was going to take each from the perspective of Btn vs BB in a single raised pot and look at suited and unsuited variants for each. The plan was to finish this weekend with a number of formal hypotheses about the effects of connectivity and pair rank on strategies for these types of boards to be recorded and tested later.
In the end I noticed that the rake was set for 500z in the monker sims I was working with. So I reconfigured the tree and reran it for my stakes where the rake cap is way higher. After 19 hours of running and 100 iterations/node the strategy still hasn't stabilised fully, though I suspect the equity leakage is now minimal.
This weekend was not a complete waste of time by any means, I learned some stuff about the one board I got around to really attacking. Its also useful to better understand the challenges I face with regard to doing useful work in a useful timeframe. This leads (at last!) to questions.
Do you have any useful heuristics for deciding if a particular solve has run long enough?
If there is a clear trend of the frequency with which a line is taking moving in a specific direction iteration on iteration, is there any way to estimate where it will end up?
If I was to say to you that I have drilled a reasonably optimal preflop strategy already, do you have any intuition as to the likely most useful spot(s) to spend my next 20 hours of study? In other words, what types of spots do you think you gained the most value from solver study?
Sorry if the setup was overly long or the questions were too many,
Sept. 8, 2019 | 7:33 a.m.
Thanks for the reply.
Yeah, its definitely true that post flop tendencies will affect optimal exploitative pre flop play. However what I am really aiming for with the current study I am doing is to get a nice functional set of preflop ranges together to give me a platform to start analysing post flop play. When I am done studying preflop, I will start into looking for population tendencies post and I might well tweak these ranges based on what I find. One thing at a time though :)
Aug. 15, 2019 | 7 p.m.
I understand the points you are making but I think you might underestimate just how much value one can get from doing this type of study. First off, having a more strongly defined sense of how wide your range is preflop helps you to make better decisions postflop. Also, having a simple but systematic approach allows you to stop second guessing whether you should defend K7o or not, freeing up mental bandwidth.
Most importantly though its not just the size of the mistakes you make in a given spot that impacts your winrate, its also the frequency that you are in that spot. Having spent all this week looking into various preflop spots I have increased my opening frequency in the small blind from 30% to 65% and on the button from 38% to 55%.
I intend to study preflop for maybe 40-60 hours more before moving on to systematically approach flop play, building on the ranges I am identifying.
Aug. 15, 2019 | 6:49 p.m.
I am trying to build a preflop strategy based on data from Holdem manager and solver work. I decided to start with defending the big blind vs various open sizes from the small blind. I am using data from $10 zoom.
I observe that the small blind opens 41% when folded to (1700) instances. The most common open size is 3x, followed by 2.5x and min.So I ran a simple strategy over a (representative) sample of 33 flops for a 3x open.
So here, any hand that has a number greater than 2 beneath it is more profitable to call than to fold. I'm pretty happy to now assume that these results are valid. What I don't know though and can't think of a way to figure out is "Other than premium holdings, what hands should I be 3betting?" If I 3bet and get 4bet, how should I proceed?
I don't have a big enough sample to figure out how people are actually responding to my 3bets in these spots and it would take years to build one. Also the solver runtime needed to start just toying with strategies is a lot. So I wonder if anyone knows how to figure out what approach to take with 3betting?
Aug. 12, 2019 | 9:53 a.m.
In practice a lot of our opponents bluffcatchers will be pairs in a lot of spots. How would you block those without having a bluffcatcher/value hand ourselves?
In your K7238r example would I be right in thinking that, if we got to the river with them this way, hands like QJ or T9 would be the best bluffing hands? I'm thinking that QJ blocks KQ and KJ which might play this way and T9 blocks pocket pairs which are potential bluffcatchers and could play this way on all streets. Does this make sense?
May 30, 2015 | 10:27 a.m.
Hi Jack. I enjoy your videos and your style of explanation. I think that the fact you don't use a HUD and are playing against players who you don't know well is actually great for video making. I am grinding 10 and 25 PL and find that I really want to know about readless situations a whole lot because of the fluid nature player pool. I do use a HUD myself but when you have only 50 hands on an opponent or whatever it may give you a vague idea of their preflop ranges but almost nothing about whether they barrel aggressively or not on the turn. So I find these "vanilla" approaches to certain spots really useful.
At 28:12 you fold QJ97 suited to the queen in an unopened pot from the button. Seems you were just distracted by your hand history review right? For me that hand is maybe a somewhat marginal cutoff open but a slam dunk to raise when it folds to your button.