Open Side Menu Go to the Top
Register
Regarding solver study Regarding solver study

01-11-2019 , 08:40 PM
I've been studying a lot and kind of focusing on what specific combos like to do in specific spots, but I had a realization that I wanted to run by people. Perhaps it's obvious to players better than I.

Any given strategy can be broken down into frequencies, bet sizings, and EV, yah? So really I should just study those things when I'm looking at boards and learn to intuit what combos would fit into given frequencies. I

Often times it's somewhat obvious what combos fit into a frequency. Like lets say you need to have a decent value range but the board clearly does not have many natural bluffs, the solver will just often take a low pocket pair and use it for the bluff.

I feel like I just want to move away from rote memorization, because the game tree is so vast, I could just be writing things in my notebook forever. So if I'm not on the right track with this thought process, what other ways are there to simplify the solver study?
Regarding solver study Quote
01-11-2019 , 10:01 PM
(I Think) the key is to understand what the solver is doing and "why" it is doing it. Start to do the solver work on your own and construct your ranges in given spots. in time you should learn to become better and better "solving" your hand while playing for that given board.
Regarding solver study Quote
01-11-2019 , 11:15 PM
Following on from what Zuko said. I try to pick a random flop and then write down how I would play my range on it. What's in my check raise portion, what's in my cbet portion, etc. Then use the solver to see where the differences are between what it thinks is right and what I think. This can sometimes highlight pretty eye opening stuff like JJ being a xf half the time on Ad9s8s as SB vs CO 3b pot. If I don't understand a decision by the solver, like say betting more combos of AQ with heart and club instead of club and diamond then I'll write it down and try and work it out on my own. Really try and struggle at the problem before posting it on a forum so you can learn better and exhaust all of your poker logic first.

I think that would be a good starting point.
Regarding solver study Quote
01-12-2019 , 07:12 PM
i think the best kind of insight that solvers can give us is that they reveal the broad structure of optimal play. there is some value on focusing on specific boards and doing 'deep dives' into tough spots, but IMO the best way to use a solver is to look for the overall shape of our strategy and how larger hand classes perform/mix. this involves the 'why' of our strat, as mentioned above

another thing is that when people use solvers the complexity is daunting, but we can use solvers to simplify our strategy, for example:

if we see that we are soposed to be mixing, but one option is very low frequency we can use the node locking function, removing that low-freq. option and see how much EV is sacrificed

oftentimes we sacrifice very little EV, while making our strat much simpler to implement 'in game.'
Regarding solver study Quote
01-22-2019 , 05:42 AM
Quote:
Originally Posted by Zuko
(I Think) the key is to understand what the solver is doing and "why" it is doing it. Start to do the solver work on your own and construct your ranges in given spots. in time you should learn to become better and better "solving" your hand while playing for that given board.
This.

Trying to find the logic for example why certain combos are bluffed over more natural bluffs etc is very valuable. And when you start getting better at the logic of why stuff happens, making good strategies becomes more natural instead of just trying to memorize what solver does.
Regarding solver study Quote

      
m