Perfect avoiding of worst possible actions
It's quite obvious that the worst possible strategy would be to constantly fold everything, or check if fold isn't possible. It would 100% guarantee a loss in the long run.
But nobody will do this, so what about the different situations that your current strategy puts you in:
1. Is there an exactly opposite action to the optimal action or simply "the worst possible action" for each situation ?
2. If you're 100% flawless with avoiding these "worst possible actions" in the long run, will this guarantee you a profit ?
15 Replies
It’s far from obvious that fold everything is the spray possible strategy, especially since folding about 3/4 of your hands preflop (give it take a bit) is actually part of an optimal strategy. A strategy of folding everything except 72o, for example, would indeed be a worse strategy than folding everything.
As for your questions, the answers seem obvious - yes, there is a worst possible strategy and no, you cannot guarantee a profit by avoiding that worst possible strategy. It is generally accepted that there is a best possible strategy, that is a strategy that maximizes your EV. It makes sense then that a worst possible strategy exists, one that minimizes your WV. However, just as there are plenty of +EV strategies that are suboptimal, there likewise would be plenty of -EV strategies (such as “fold everything”😉 that would yield -EVs, but aren’t as bad as the absolute worst possible one.
Good decisions lead to certain situations. There can be for example 6 decisions for you to make in one hand. And what I meant was something more like: you can make the worst error 6 times in all 6 spots. So this "worst possible decision" only sees a certain moment. It sees but it doesn't care if you played previous actions well or not. By this I mean that you don't need to follow the worst possible long run strategy to face a decision that will give you the chance for the worst mistake.
I just want to point out that both my questions refer to situations which see only a particular moment. And of course this moment multiplies then as the time progresses and you make your next decisions. Before time progresses and those decisions multiply, you must first make the current decision (and avoid the worst mistake).
I think the way I started post #1 could have been a little bit misleading and redirect these 2 question towards the long run, which wasn't my intention.
Good decisions lead to certain situations. There can be for example 6 decisions for you to make in one hand. And what I meant was something more like: you can make the worst error 6 times in all 6 spots. So this "worst possible decision" only sees a certain moment. It sees but it doesn't care if you played previous actions well or not. By this I mean that you don't need to foll
I've reread this three times. I have no idea what it means. It's like, the words are all English, but arranged so that it seems like a foreign language.
I've reread this three times. I have no idea what it means. It's like, the words are all English, but arranged so that it seems like a foreign language.
OK 😀
I meant:
Following the worst possible long run strategy would lead you to different situations than avoiding "worst possible decision" in each spot you play. You'll find yourself in different spots if you play great and in each of these spots you can still make the worst mistake.
His answers to my 2 questions were more based on the existence of the worst long run strategy. It's something different.
"Perfect avoiding" is being flawless on all decisions and not making the worst mistake there.
"Perfect avoiding" isn't just not following the worst possible long run strategy.
First one would be very difficult to execute, while the second one would be quite easy.
I think that worst possible actions just see the long run in a special way. And this terminology phenomenon could have been misleading here:
A) When there is a certain hand scenario and you make the worst mistake in your 1st move, then in your 2nd move you will be on the "worst long run strategy path".
B) If you play your 1st move according to GTO, then in your 2nd move you won't be on the "worst long run strategy path" anymore.
I don't know, I might be tilting a bit, but I still desperately want to see precise answers to both of my questions 😀
♠ By the way - I've come up with this brave side-conclusion:
You can only apply GTO to a spot where your previous actions in this particular hand were played with correct GTO.
I think it makes sense.
1. Is there an exactly opposite action to the optimal action or simply "the worst possible action" for each situation ?
Yes, just select the lowest EV move.
2. If you're 100% flawless with avoiding these "worst possible actions" in the long run, will this guarantee you a profit ?
Nope. Avoiding blunders doesn't guarantee profitable strategy. Choosing the 2nd worse move instead of the worst move is still torching money often.
B) If you play your 1st move according to GTO, then in your 2nd move you won't be on the "worst long run strategy path" anymore
Also not true. For example, you can correctly open AA and fold to a raise. That would be a bigger mistake than open-folding AA.
You can only apply GTO to a spot where your previous actions in this particular hand were played with correct GTO.
This part is true.
Awesome. Thanks for a precise reply tombos21 !!
You can only apply GTO to a spot where your previous actions in this particular hand were played with correct GTO.
Is it? Does GTO look backwards at your prior actions? I thought it only looked at the current situation. Of course as an overall strategy, it'd wouldn't work to only apply GTO if you got to the river and just randomly bet/fold/call/raise the rest of the hand. But I thought GTO didn't care how you got there, just what was happening now.
Is it? Does GTO look backwards at your prior actions? I thought it only looked at the current situation. Of course as an overall strategy, it'd wouldn't work to only apply GTO if you got to the river and just randomly bet/fold/call/raise the rest of the hand. But I thought GTO didn't care how you got there, just what was happening now.
It doesn't look at how you got there but it's still important that your starting assumptions are clean.
A solver doesn't care what inputs you plug in. It will optimize the strategy for whatever ranges you give it. But if your starting assumptions are not GTO, then your output is also not GTO.
As an extreme example, imagine plugging in completely random ranges. Obviously the resulting strategy would be meaningless noise. Garbage in garbage out.
The class of game theory that deals with this is called safe / unsafe subgame solving.
So maybe I originally misunderstood, or the thinking has evolved, but for pure GTO isn't the range of your opponent "any two cards"?
Solvers, of course, are different and provide a better outcome, provided you are reasonably accurate with your ranges.
So maybe I originally misunderstood, or the thinking has evolved, but for pure GTO isn't the range of your opponent "any two cards"?
Solvers, of course, are different and provide a better outcome, provided you are reasonably accurate with your ranges.
Solvers can start at different points in the hand. If you're solving preflop, then yes, the starting range is any two cards. But you could also run a solve that starts on the flop, in which case you need to assign the starting ranges.
ITryDeuces said earlier:
You can only apply GTO to a spot where your previous actions in this particular hand were played with correct GTO.
Which is true in the sense that if you're solving a spot later in the hand (say you're solving flop->river), then you need to plug in good starting ranges, else the strategy will be exploitable in reality.
