What is it about? Presumably fighting?

_________________
be immersed

Yes. There are quiescent endgames, but endgame skill requires fighting strength. If you lose your fighting spirit in the endgame you will miss opportunities and lose games you should not.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Hi often, in case you missed it I would be interested to hear your thoughts on the below.

When you are behind, you should try a riskier strategy to get ahead. If you are ahead and your opponent is trying something risky, you will probably win.

_________________
Revisiting Go - Study Journal
My Programming Blog - About the evolution of my go bot.

I think the idea of "10 point lead in the opening" is misleading. After the opening, I get groups and shapes that are worth something, or at least I have to try to make them worth something. Maybe with perfect evaluation I could tell you how much that weak group or big corner is worth, but in a game I just have to guess who is winning and try to make what I can with what I have.

Is your position that the opening is so irrelevant that the answer will follow a binomial distribution with mean 50, in other words the same as if they just played all those 100 games from a blank board?
Yes. I think this is applicable for any rank (if you got 1000 Xk,Xd,or Xp). Look, of course there will be games that an opening might be so disastrous that it is impossible to come back from, or that there will be games that are completley lopsided in favor for one person until that "one move".

However,

If you're playing someone at your same rank there is no real guarantee that your opponent has any better idea than you as to whats happening or whats being played

There's two easily pointable phenomena in Go that i can use to illustrate this.

1. Handicap games
If the opening mattered as much as is being claimed, there it would mean that handicap stones are just "too big" of an advantage. But if that were so, it wouldn't explain why stronger players would have a chance to win. It is because of the middle game and fighting that ends up determining the game, not the handicap stones in the beginning.

2. Pro games with joseki/fuseki
It was once asked in a game review at go congresses if Pros bothered with the large avalanche joseki. The pro answered "no, it's too complicated, we just play some set openings and then go straight into the middle game". If the opening mattered even more than the middle game, then this would contradict a lot of set openings and pro behavior.


Again the claim that i'm providing arguments against is that the opening matters so much that it can determine the rest of the game. I feel that, especially at the amateur level, the concepts of the middle game are much more important than the opening.

Let's think of it in another way.

Take an opening mistake, maybe a mismatch in fuseki where you went for territory and gave your opponent all the thickness, or maybe a misplayed joseki where you get a poorer result. Yes it is important to know where you went wrong and to strive not to do it again.

However, what would be the conditions necessary for this to be seen again? Also, how often could knowing something like this determine the rest of the game?

Now take a middle game mistake, where you misjudged a groups life, or played too close to thickness, or did not properly attack something, or even just didn't know how to kill/make life. In my opinion, these are concepts that can be easily learned from and be repeatable from game to game.

Again, what would the conditions be for this to happen again? How translatable are these lessons, and how often could knowing something like this affect the rest of the game.


Yes the opening matters, but i don't think it is the reason you will "lose" a game.

This is so extremely absurd --- on the contrary, the fact that a well calibrated handicap can change the win/loss ratio from an arbitrary number to close to 50% shows that the handicap stones (which can be viewed as a very bad fuseki by white) have a very concrete role in determining the outcome.

I wouldn't say misleading, but it's not precise for sure. Perhaps the leading player (black) has more territory and the influence the opponent got doesn't compensate well enough according to perfect/excellent play from both sides which involves the black player invading and making sabaki. But these random 5ks we picked to play the game black is crap at sabaki, and white is particularly good at using influence so actually the game is pretty even for them. Or perhaps black has good influence that is worth a 10 points lead in the hands of an expert but he is bad at using influence and attacking. Or perhaps white has a weak group that black should get 5 points of benefit form hassling but fails to do so, but he also has a 5 point advantage from some joseki mistake in another corner that he retains. That's why I said pick 200 5ks and get them to play 100 games, the individual strengths and weaknesses of those 5ks will get averaged out.

Thanks for answering, but I must say I am incredulous. Let's break this down. We start with a board with one player 100 points behind (yeah this is hard to achieve, but lets's say we play 4 large avalanches where white screws up and both gives black the outside influence and then dies in the corner. And then maybe play a few self ataris just for good measure. We now pick 200 pros and they play 100 games. Do you still think white wins 50 on average? Assuming the answer is no (because c'mon you can't be that crazy!) then how many would black win. I will say 100 on average. Yeah there might be that 1 game that Lee Sedol won against Kong Jie after dying massively and being 50 points behind, but 100 is a lot bigger than 50 and there are a lot more examples of pros losing games they are 50 points behind rather than making some miraculous comeback. So your position would be a 10 point lead black wins 50, 100 point lead black wins 100. Where does this change happen? Is it a sudden jump or gradual? If black has a 50 point lead how many games would black win of 100? 100? 50? Somewhere in between? Perhaps 42 is the magic number of point lead at which it goes from 50 to 100: start with a 41 point lead and the opening has no effect and black still wins 50, but with 42, BOOM! suddenly black wins 100. Obviously I think this is absurd and there is a gradual increase in the number of black wins as the starting position black lead increases. I don't know what shape this graph will have, but feel confident to say it depends on the strength of the players and will be steeper for stronger players.


Not sure if this 'however' is arguing for or against your position. As I said above some 5ks or some pros or whatever may be better at using influence or invading or preserving a lead or whatever. There's no reason to believe your opponent in any of these 100 trial games is better or worse at some aspect so a small advantage to start will on average end up as a small advantage at the end.


What?????!!!! This makes no sense. In handicap games one player is STRONGER than the other. In my examples 200 5ks play each other and they are all the same strength (or within the relatively narrow strength range of one stone). If a 5k plays a 1k even the 1k will win most of the time, I presume you agree. How much? Maybe 80% (I took inspiration from http://senseis.xmp.net/?EGFWinningStatistics though those data come with caveats). The exact number doesn't matter, so long as it's >50%. Yet if the 5k and the 1k play with the 5k starting with 4 handicap stones (an opening advantage; perhaps we can say of about 40 points, or maybe 50, or maybe 60; I don't care exactly how much) then 5k will win about 50%, do you agree? (Actually for average 5k versus average 1k the 5k should win a bit less than 50% as white has a half stone advantage in handicap games but nevermind). So what just happened? An opening advantage (the handicap stones) changed the win rate! So now what happens if the 5k plays another 5k starting with 4 handicap stones. According to your previous argument (that opening advantages have no effect on the win rate between equal strength players) the 5k with the 4 handis will still only win 50% of the games. So 4 handicap stones have no effect if a 5k plays another 5k, but if they play a 1k it does? Bonkers.



I'm not saying the opening matters more than the middlegame. I agree with you the middle game is more important and has more effect on who wins. However I do argue that the opening has some effect on the outcome of the game, rather than none as you profess. Regrading the large avalanche, I've also heard that pros don't like to play it because it is too big and settles too much of the board with established patters, it's boring. Also others (Michael Redmond for example) have said that it was good for the player to start in the corner. But you do still see the large avalanche in pro games, for example Kim Jiseok has played it quite a bit recently. I've also heard quite a few pros say the opening is important.


I do not claim it determines as in a 10/20/60/whatever point lead in opening means you win 100% of games (until that lead gets huge), but that the opening does affect the rest of the game and have an effect on the win rate. Are you not able to appreciate that something can have an effect which is not black and white, all or nothing?


Actually I think opening mistakes and theory are often easier to teach and learn about than middlegame fighting, and that is one reason for their popularity. Although there are also lots of middlegame strategies and tactics and theories and shapes etc that can be taught, perhaps the most important part of middlegame is reading, and that is hard to teach other than saying "go away and do a thousand tsumego". But yes, as I said many times before, I agree the middlegame has more effect on the outcome of the game.


Finally! You acknowledge the opening matters. Does this mean you also agree having a good opening might increase your chances to win? Although I agree there are few games that are lost in the opening and there will usually be plenty of chances to turn around a game you are losing in the middlegame, as the level of play increases that becomes harder and if your opening mistakes are sufficiently large and your opponent is sufficiently skilled at controlling the game then there are some games that can indeed be lost in the opening*.

* Pretty rare in pro games, but this is perhaps an example, feature the large avalanche as luck would have it: viewtopic.php?p=180499#p180499

I never said opening doesn't matter. More that if you're trying to improve, you don't need to worry about/focus on the opening. More effort should be spent worrying about the middle game.

But also, a lead/deficit in the opening really doesn't matter for amateurs in the grand scheme of the game. So "snowballing your advantage" or having an advantage really doesn't matter for amateurs because they're not going to take advantage of whatever that means. Typically people don't even know what advantage they've been given or even know that they're behind, so it's not going to happen.

And finally, you're pushing some made up hypothetical situation of a "lead in the opening" and imaginary players. First you're going with a 10 point lead then you're talking about a 42 - 100 points lead. Yes, you can be ridiculously ahead and probably win the game, that is not typical so why even discuss it.

Again, what does a "lead in the opening" even mean, how does this go with any random player's play style or tendencies? You're looking at one variable when there are so many variables in a person's Go.

Does the opening game affect the rest of the game? of course, it's the opening.
Does the opening effect the win rate? still, no. i would be willing to put money down that if we took 10,000 amateur games, made a judgement about the result of the games after a perceived opening, that we would still only be right half the time. In fact, if we can arrange that some how within this forum that'd be an amusing experiment.

@often: I'm sorry, but you're really not making any sense.

An advantage in the opening is easy to create and quantify. Giving black 3.5 komi instead of white 6.5 komi creates a 10 point lead. It is as simple as that. How do you figure that reverse 3.5 komi will not affect the winning percentage of amateurs? That is frankly an insane position to hold.

Sure, snowballing is not a thing in go, and the middle game is more important than the opening, but that does not mean that the opening has no effect, or that such an advantage isn't real.

I discussed a pro vs pro game with an 100 point lead because I thought even you would concede the leading player would win more than 50% of the games, and in fact 100%. The talk of a 42 point lead was a reasoned argument to show that if a pro vs pro with no lead wins 50%, and a pro vs a pro with 10 point lead still wins 50% according to you (not me), but a pro vs a pro with a 100 point lead wins 100% (do we agree on this?) then at some point in between those two scenarios 50% has to transition to 100%. I posit it is a relatively smooth transition and not a jump at 42 or any other number (which is admittedly an imprecise way to characterise an opening advantage). It is a thought experiment to reveal the illogicality of your position.

I'll take that bet, but judging 10,000 games is rather time consuming. What could be scripted is taking a large number of games, say from KGS, using some monte carlo bot to assess the score difference after say 50 moves, and then counting up the win rate. If someone did this with KGS 3ds and looks at games where the bot thinks one player leads by 10-15 points, I will willingly bet you $100 that the leading player will win >60% of a sample size over 100. Deal?

For a more manual approach (which will take a lot of effort to be statistically significant), I invite you to take part in this thread.

This is a silly argument, of course tiny advantages matter and this can be demonstrated. http://senseis.xmp.net/?Komi%2FStatistics shows a switch from 51% black win at 6.5 komi to 49% black win at 7.5 komi at the 8D+ level. I don't have a link handy, but I've also heard that professionals decided to jump from 5.5 to 6.5 as a result of statistical analysis showing black was winning too much. 1 point advantages impact win percentages, never mind 10 point ones.

Or let's switch it around, do you also believe that the endgame can have no impact on the result of a game? That if I play a 3 point blunder in the endgame, it couldn't possibly cost me the game? If not, then why would it matter if I made that blunder significantly earlier: boundary setting moves can certainly occur in the opening.

??

Then why do you give different numbers of stones to players of different ranks? If a pro can beat an amateur 6d at 3 stones, it doesn't mean that the pro can beat the same guy at 9 stones. That's because the opening makes a difference.

often, I think most people agree with you that the middle game is important, and maybe more important in many cases than the opening. But I don't understand your arguments that the opening is irrelevant to winning or losing.

Think of Herman's argument. Surely, if one player starts out with a greater komi advantage, it affects the win rate, no matter how good the opponent is at middle game, no?

_________________
be immersed

too many responses, too tiring, goodnight.

Don't give up! This discussion is just entering the middle game!

_________________
be immersed

As someone who played a fair bit of league back in highschool I remember the game to much much more punishing due to the way how the game is designed. If any team member gets a couple of kills in the opening then they've got a very clear advantage in higher stats which can make a huge difference. The benefit to these extra stats is relatively simple and straight forward. That benefit being that they can play more aggressively in what they do. (i.e. A jungler can spend more time diving a tower, top and mid and the AD carry will have better time harassing and denying minion kills and the support will enable his AD carry to do all of that indirectly.)

On the other hand, in go if you get a result which is overwhelmingly better for one side then it still comes down to *how* it was better. Did one side get too many points or has he acquired overwhelming thickness? Was he able to seperate a group into 2+ weak groups? Has he forced his opponent into terrible shape? The thing with gaining the advantage in go is that you need to know how to keep it and how to capitalize on it. (Which is something being discussed above.)

Something which comes directly to mind on how go has snowballing aspects is when thinking about thickness. If one player manages to deal the other into the corner or has a massive wall that has super solid shape then he can often use that influence to haunt the other player for the rest of the game. Indeed, if you are playing an even game against a player a couple stones stronger than you then he will be able to make profit in such a way that it would be difficult to recover.

Obviously this kind of skill is required in both games but I feel that in go punishing your opponent with your advantage to gain yet more advantage is more subtle. Snowballing is something which can take place in go but I feel like it's more situational.