Open Side Menu Go to the Top
Register
GTO Heuristics & Learning Framework GTO Heuristics & Learning Framework

01-26-2024 , 05:12 PM
I'm seeking an efficient pathway to begin my studies in GTO. To me, most of the training and videos I see online jump into solvers using random hands / flops. It seems the selected hands are randomly chosen from billions of possibilities, lacking clear context or explanation of their significance within the broader poker taxonomy. Is the expectation really to study all 1,755 unique flops, not to mention the millions or billions of potential turns and rivers, or is there a more methodical approach?

I acknowledge my understanding of GTO is incomplete, and I welcome all feedback on starting points and reliable information sources.. I've subscribed to resources like GTO Wizard, PokerCoaching, and have also tried RedChipPoker. For context, I have a background in math and deep understand of quantitative analytics.

Overall, here is where I am struggling.

- GTO: The term 'unexploitable' is frequently used in GTO discussions, yet it seems paradoxical given that no human can play perfectly 'GTO', rendering everyone exploitable to varying degrees.

- GTO feels like an algorithm wrapped in human behavior (NIT/TAG, etc).

- GTO works for computers and solvers with perfect memory, perfect math skills and ability to perform nuanced frequency of bets (eg 22% of the time, etc.).

- Has GTO been simplified into a learning framework with easy heuristics / rules to follow based on classification of important hands. If not, the numbers are simply massive. 1,326 starting hands, 19,600 flops (1,755 distinct flops), 2.6m turn hands and +3b river hands and I am left wondering how do I start learning? Jumping into complex UI of a solver and hunting and pecking my way through an endless array of hand combinations.

- Given the impracticality of memorizing billions of solutions and bets, there must surely be a way to distill this data into easily memorizable heuristics. I have only begun my search but I haven't found any tool, service or coach (yet) that has set up a new user learning agenda with this as the goal. If one exists, please let me know! Without simplifying, how do you go about understanding poker taxonomy and what's most important to learn first?

- Common Flops: Here are the most common flops based on the subset published by PioSolver. (https://piosolver.com/blog/2015-11-05-flop-subsets/)

- Common Board Textures: I have put together the most common board textures based on the work by Cepheus who solved Limit Hold'Em http://poker.srv.ualberta.ca/about.

- I can then map them together and put the above into a solver and hand roll heuristics to memorize or practice.
- PioSolver 25 ((3s3dKs, 7s7d6s, QsQd7s, 2d3sAs, 2s4d8c, 2s5dQc, 2s6dQc, 2d9sKs, 2sQsKd, 3s5d8c, 3sTdJc, 3sJsAd, 4s6sJd, 4s9dTc, 4sTsJd, 4dTsKs, 5s6dTc, 5s6dAc, 5d7s9s, 5s9sKd, 7d8sTs, 7d8sJs, 7sQsAs, 7sKsAd, 8s9dAc)
- High Card Dominant Boards:
- High Card with Middle/Low Connected Cards:
- 2d3sAs (1 High, 2 Low Connected)
- 3sJsAd (High Card with Middle/Low Connected Cards)
- Boards with Two High Cards:
- QsQd7s (High Card Textured Boards)
- 2sQsKd (Boards with Two High Cards)
- High Card Textured Boards:2d9sKs, 4dTsKs, 7sKsAd
- Mid Card Dominant Boards: 7s7d6s, 5d7s9s
- Low Card Dominant Boards: 2s4d8c, 3s3dKs
- Connected Board Textures:3sTdJc, 4s6sJd, 4s9dTc, 4sTsJd, 5s6dTc, 5s6dAc, 7d8sTs, 7d8sJs, 5s9sKd, 8s9dAc
- Suited and Monotone Boards: 7sQsAs
- Special Draws and Textures: 2s5dQc, 2s6dQc, 3s5d8c

Then use PioSolver Common Subsets To Study
- IP players in 3-bet pot, BTNvsBB; 100bb
- IP player in single raised pot, BTNvsBB 6max, 100bb
- OOP player in 3-bet pot, BTNvsBB 100bb
- OOP player in single raised pot, BTNvsBB 6max, 100bb
- IP player in single raised pot, SBvsBB 6max, 100bb
- OOP player in single raised pot, SBvsBB 6max, 100bb

-I presume the most critical positions/flop-board textures have been identified, prioritized, cataloged and solved. Are there videos or online training modules that delve into these specifics?

Your insights would be greatly appreciated.

Brian
GTO Heuristics & Learning Framework Quote
01-26-2024 , 06:27 PM
Think of GTO as simply 'solid fundamentals' or 'rough guidelines' rather than a replicable strategy you aspire towards. Sure, over time we may get closer to it, but that's not the sole reason we study it. We are trying to get a jist of how our range play our range vs the opponent's range on various boards, or in other words, how to play various situations efficiently and effectively.

I think where you might be overcomplicating is zooming in too much. Rather than worrying about which combos do what at the exact frequencies, think in broad strokes:

- High/ Low frequency bet?
- Big/ Small bet size?
- If heavily mixed strategy, which types of hands like to do which action?
- How is my value playing? What types of hands are bluffing?
- If facing a bet/raise, what are my defence thresholds? What hands need to be defended? This one solvers are particularly handy for because without them it's easy to misjudge which hands are worth defending.
- Why? What is this strategy trying to achieve? This one is trickier but can be explored in part by looking at the opponent's response. What are we folding out/ getting called by? Be curious and ask questions. Watching those coaching videos should help you get started with this.


I think you're on the right track using a small subset of flops and studying those (you can set the 25 flop subset as a filter in GTO Wizard reports if you were unaware). You can simplify the same way for turn and river (overcard, flush close, straight-completing, brick). Of course, as we get more confident we can get more and more specific and explore the subtleties between slightly different boards.
GTO Heuristics & Learning Framework Quote
01-27-2024 , 05:33 PM
Yes, you must learn every single flop, that is absolutely much more important than learning skills like "hmm, that guy there has raised every other hand for the past three orbits, I wonder if he is a maniac"
GTO Heuristics & Learning Framework Quote
01-31-2024 , 11:08 AM
The new Sklansky/Malmuth book Small Stakes No-Limit Holdem: Help Them Give You Their Money gives a pretty detailed explanation of why playing GTO at small stakes costs you money.
GTO Heuristics & Learning Framework Quote

      
m