Phil Ivey vs. Magnus Carlsen..poker and chess

I would rather think really really hard about how to get better at bluffing, than have to learn hundreds of entirely distinct concepts and tactical ideas and then have to execute them with absolute precision (in chess one blunder can cost you the game regardless of how well you played beforehand).

Do I have to assume that this single action of bluffing isn't compiled of hundreds of things like that such as in chess? Yeah, just as I have to assume that when a dog is making gestures, there's not some secret, deep poetic meaning behind it. I am making claims as reasonably as I can. My opinion could change if I studied poker for months, or it might not. I'm not just going to assume that it will change. I have taken a cursory look at arguments people make with poker, not looking into equations they might use of course, but yes I'm fine with the judgment I'm making for my purposes and I suspect studying it more deeply would result in diminishing returns.

I brought this up before but what if chess had a 5000x5000 board. Am I still supposed to give a benefit of the doubt? (although, 10^50+ positions is pretty complicated on its own) When I am relentlessly swamped with tactical ideas just to play a constructive move without blundering material, I'm not going to need as much convincing that chess is fucking hard and that it hates you; I could even be biased towards looking for reasons to think chess is easy and still totally fail to do so. And it's not like I haven't played poker before. Never at all got the feeling like the game was ever unplayably difficult, not even as a little child. Simply the fact that some people get really really good at it is not in itself a very convincing argument to the contrary.

But why then do you think anyone who has actually played and studied both would care for or give any credence to your estimations?

I don't necessarily think they would. I'm saying from my perspective, it doesn't make sense for me to blindly trust people who claim they have studied both and can interpret their results accurately (both of these virtues by the way don't necessarily go hand in hand). Maybe they can, but it's not something I will trust just because they are acting like an authority.

So, sure, you can be unconvinced, but my perception of you being unconvinced is not a good reason for me to change my view. I am contributing my view, and it might as well be the one that, all things considered, including what others have said here, I trust the most -- no need to copy someone else's view unless I hold it.

I've been annoyed by your posts and probably resorted to being personal so for that I apologise. You haven't lowered yourself to personal attacks so fair play.

Chess is about 95~99% skill I'd imagine, the luck percentage happens in time pressure and other stuff like that.

That's why people like blitz and bullet, its more confusing, hectic and luck based in fast time controls.

Phil Ivey will lose every game vs Magnus. 

Magnus may win a few games off Ivey in poker. Poker has a signficant luck factor invovled. 

That's cool. To me though it's hard for me to understand why someone would get annoyed by someone else's view -- for example if I think someone's view is totally wrong, I can simply be unconvinced by it.

I guess you perceive me as attacking poker or something. But it's not true; I don't have anything against poker and think it is a cool, challenging game. I also have my opinion, but I do think before I express it, and I also qualify it -- I have conceded that I am speaking from ignorance to some extent. So take it for what it is -- you can be totally unconvinced by my argument and there's nothing wrong with that.

"Win a few games" is nonsensical and meaningless in and of itself. 

The luck factor of poker converges on zero as the number of trials increases. A solution of a poker game, like the three card toy game, does not refer to luck. Any time someone even uses the word when comparing chess to poker all that person is doing is confusing the issue. That there is a short term luck factor in poker makes it in no way less or more complex (other than the psychological impact). It is completely and totally irrelevant as to theory. The only way a lesser player comes ahead of a better player is if there has not been a sufficient number of trials. Depending on the circumstances, that number can be absurdly large. That is it. Magnus, making the same assumption he is not the superior player, never wins. Ever. This is pretty much by definition. 

Fair enough. So essentially we are asking, is it harder to make the best decisions in chess, or the best ones in poker, regardless of the outcome.

(Well maybe that's not totally precise. For example it might be easy to get 95% good at something but for whatever reason it's harder to get the last 5 percentage points, whereas in some other game it's easier to get those last percentage points, but harder to get to 95%, maybe because of the learning curve or something. But ok, this is mostly quibbling)

I will. If we are talking about computational complexity, that a computer has more trouble with a task means it is more complex. That perhaps you find this incredible is besides the point. That the human brain is better at handling some complexities better than a present day computer program is also besides the point. The brain just prunes complexity more efficiently than programmers can dream of today. There is perhaps the independant variable that programmers have worked more on one problem than another, so performance isn't completely tied to complexity, but that is about it. 

A program can be written to play poker just as it can to play chess. The data is out there. There are many people that play millions of hands each year that have saved every hand via a databasing tool. Plenty of people are trying to construct poker bots with some success. While a different kind of problem than chess, it isn't something a computer can not do. 

The solution for the basic three card toy game is somewhat simple. A computer isn't necessary. Various simplifed versions of poker have been solved by computers, and a computer can play the more complex forms of poker reasonably well. Just not as well relative to humans as chess. 

That is evidence that it is more complex on a computational level. Not proof, but it seems a bit stronger than incredulity otherwise.

Apples and oranges, really. There are theoretically perfect moves in both, and neither have been solved other than minature versions like limited piece tablebases in chess and toy games in poker.

The main difference is that there are probably fewer reasonable choices to select from in typical poker decision than in chess. In fixed limit this is obvious as there are at most three choices. In other forms, bet sizing comes into play, which makes it a bit more complex.

It is probably way easier to accidently do the right thing in poker than chess. This, along with the luck factor, is what makes poker seem way, way simpler than it is. 

"It is probably way easier to accidently do the right thing in poker than chess."

Is it so weird for me to think that this should count for something? I know not everything, but something? The fact that in chess you really have to understand something to play the right move (in all likelihood) seems to give more of an impression of unfathomably sophisticated strategy, to say the least. Squeezing out tiny bits of extra strength can be a bit beside the point -- it's like saying we should care about not just mating with queen vs king, but mating in the absolute fastest possible way. Most of the way I will consider your skill will simply be based on, can you win the position at all, not how fast can you win it.

I will. If we are talking about computational complexity, that a computer has more trouble with a task means it is more complex.

That's a nonsensical statement.  By qualifying the word complexity, you're just trying to dance around by omitting the implied "...for a computer" at the end of your statement.

"If we are talking about people whose favorite color is blue, then if someone has a favorite color, it's blue."  Yeah, okay.

You're "pruning complexity" is saying the same thing I pointed out with the analog/digital comparison but re-labeling it complexity.  

Complexity is defined (literally) by having more connected and varied parts...whether those "parts" are fuzzy human factors does not affect how complex the result is.  It just changes the ability for one or another "perceiver" of said complexity to process it.  

Complexity is primarily a quantitative function, not a qualitative one.  Humans process fuzzy/random complexity better than computers.  Computers wipe the floor with humans when it comes to structured complexity.

The main difference is that there are probably fewer reasonable choices to select from in typical poker decision than in chess. In fixed limit this is obvious as there are at most three choices.

Choice is the wrong word.  Choices represent the end result of the decision-making process.  Framing choices for a selection first involves evaluating decision criteria which number more than 3 ;)...and if you only give the computer 3 anyway, you're never going to beat anyone.

It is just different. If you consider bet sizing and frequency it might actually be harder in a lot of crucial spots to accidently make the right move in poker because game theory will dictate a certain bet size made at a certain frequency.

For example, there is an optimal frequency to bluff with a missed draw, and an optimal bet size to do it with (depends on the game, pot size, etc.). The beginner might bluff in the same spot an expert would, but without understanding the purpose behind it a beginner is likely to think a bluff is always correct in that situation. An expert would know that it is only right to do it, say, 33% of the time. That alone changes things. Add in bet sizing and it is even less likely to stumble upon optimal play. 

I think a computer might be programmed to play optimally in the near future as to frequency of actions and bet sizing. It seems plausable. However, this is where a major difference between chess and poker comes in.

The added complexity is that, while optimal (meaning non-exploitable) strategy would always be the best way to play over time, against most fallable opponents playing in a non-optimal but exploitative manner has better results. Like if your opponent has a history of folding way too often, maybe where bluffing 33% is optimal and can not be exploited, stepping out of line and doing it 80% wins more even though an opponent playing optimally would punish you for doing so...  

This is mostly irrelevant to chess. It rarely makes sense to commit an exploitable error to gain a greater advantage when the non-exploitable move still wins (it might make sense to play an unsound opening or variation if you think it will help you win against an opponent unschooled in exploiting that opening, but that is not what I mean, I mean in a situation where optimal play already wins). A chess game is a discrete entity to win lose or draw on its own, and is only against one opponent. In poker, there are a lot of reasons you would want to increase your edge. In a tournament you would want to win a poor player's chips before someone else does, etc. 

This where judging relative complexity becomes a nightmare. Consider playing against eight opponents making various mistakes and/or considering your exploiting those mistakes as a mistake and adapting to exploit you.... and head asplode. I don't know if we can conclude the possible combinations are finite, much less solvable. Chess is at least finite. 

I've played Gin Rummy a bit as well. It had a learning curve, but over the course of a few weeks I could play it decently, and I was maybe 11 or 12. The scale here is just in a different universe from chess. And of course there are people who play on websites and have gin "ratings," and there are probably professional players out there. Well, yes, of course you can almost always get more precise with your strategy. If I wanted to measure something super precisely, I could keep remeasuring it, trying to be correct to the nearest tenth of a millimeter or something (look in between the little millimeter marks on a meter stick). Even if I did so successfully, the measurement would, for humanly purposes, not be much different compared to if I just took the measurement once.

Do I think I could use my intuition to become a better than average poker player in a relatively short period of time? Yeah, I kinda do. A lot of the logic seems pretty intuitive and something I could pick up along the way without super intensive study. I guess I would compare getting good at poker to raising your blitz rating a few hundred points. It won't happen overnight, but it's no yearly commitment either. Again in my estimation. I'm not going to prove it here, I just want to say what I think, because that's what we do here.

Why not go and get rich? Talk is cheap.

Gin rummy 

Maybe (maybe) I could do that, but it wouldn't be very satisfying anyway. It seems much more mechanical -- thinking about formulas, like a math quiz. And all the psychological stuff feels like it would get old eventually -- even if (repeatedly) wondering about what your opponent has is the hardest thing ever, it seems so tedious, like memorizing digit after digit of pi endlessly.

Again the comparison here is hilarious: I've gotten up to 2000 OTB (the 97th percentile of tournament players give or take) and the question of me becoming a pro chess player is way sillier than the question of me becoming a professional poker player. Yes people care much less about chess, but it's the 97th frickin percentile and I'm not even in the same galaxy as pro, lol.

You must be trolling now...

No not really. All I can really do here is try to come up with words to describe what is going on in my head to believe these things. But it's as hopeless as hoping to become a 2700 by hearing a 2700 teach you chess. Sure poker has strategy. Lots of it. And it seems pretty learnable. Even when seeing it on TV I'm thinking about a lot of the same things the players are -- do you want to look weak/strong, and how might that affect your opponent's reaction. If you think they're trying to bluff but not confident enough to call a large amount then you can bluff back even with nothing -- even if they have something they may not think it's good enough to beat you. Maybe (even if you have nothing) call a bluff instead of raise because that may reduce the bluffer's confidence (why wouldn't he raise if he wanted me out of the pot? He must have something if he's letting me stay in the pot). Just off the top of my head. Cool stuff, but not rocket science.

And also like trying to explain chess as 'one guy attacks, the other guy defends then when he is safe counterattacks'. Cool stuff, but not rocket science.