How long should a 19x19 game take?

[Koroviev]

A game should last exactly 40 mins. One second longer or shorter than that and you are doing it wrong.

[skydyr]

At the same time this is another problem that will not go away as you move to SDK and SDD. Even pros suffer from this, which is why competitions are so exciting.

I long for the day when I can reach the DDD ranks.

[SoDesuNe]

If an unknown opponent plays me with just 20 seconds thinking time each move, I would most likely don't play him again because he obviously does not treat the game seriously. A live game is not your everyday disposable KGS game.

At an average of 20 seconds per move, a game takes about 90 minutes to completion (45 minutes each). You would consider that "not taking the game serious"? :scratch:

Granted, I skipped the word "average" in your post : /
I just like the idea to take my time while playing Go, it's for me an important aspect of the game. You might be right, longer thinking times do not neccessarily make my moves better but they make the game more enjoyable for me.

But actually it's good to hear that several stronger players seem to be offended by taking too much time. If I ever play in different clubs, I will rowel myself.

[HermanHiddema]

Granted, I skipped the word "average" in your post : /
I just like the idea to take my time while playing Go, it's for me an important aspect of the game. You might be right, longer thinking times do not neccessarily make my moves better but they make the game more enjoyable for me.

But actually it's good to hear that several stronger players seem to be offended by taking too much time. If I ever play in different clubs, I will rowel myself.

If you enjoy longer thinking times, then by all means use longer thinking times. As I said, the purpose of playing is to enjoy ourselves. Just make sure that you and your opponent are on the same page. Don't make your enjoyment come at the cost of his. Personally, I don't mind if my opponents want to use a lot of time, but I would consider it polite if they asked whether it is OK beforehand.

If I visit a new club, I like to know what people expect. Every club is different, after all.

[Polama]

Personally, I just try to match my opponent's speed. I'd like to play good go at any speed. If my opponent learned as a child and instantly plays moves, I want to be able to beat them on those terms. If they're contemplative, I want to be able to squeeze just as much understanding from the board.

You can always have better instincts playing fast, and on a 361 point board after the very first moves and before the very last there's always more to read and count. Even if you get muddled reading past a certain distance, there are positions we leave intuitively as aji to be developed later based on how everything else turns out: you can always go back to those positions and read more of what could be accomplished, what stones would let you accomplish more, and how you might get those stones down. You can always go to long shot invasions and see if maybe there is a tricky path to life you can find.

[lemmata]

If both players spend >= 30sec on each move, and a game has 240 moves, it will take at least two hours. That's quite long for a beginner, IMHO. (1 min / move average would mean four hours, which is horribly long)

A beginner doesn't know joseki, proverbs, basic shapes, etc... These are all things that help shave thinking time dramatically. A beginner has more to think about. A more experienced player can quickly play joseki moves, hane at the head of two stones, extend before sacrificing, play vital points and so on in many situations. I also don't find anything wrong with a 2 hour game although 4 hours may be too much. I've had games of Catan go longer than that. Also, we're just talking about one side's moves here. The other side might be a typical beginner that slaps down stones in 5 seconds or less.

That said, the 30s/1m recommendation is somewhat arbitrary. The main point I wanted to make was that thinking a bit about every move is a good habit and is preferable to thinking insanely long about a few moves (and not thinking about most moves!) at a point when deep reading is not a skill one possesses. Adjust the actual numbers to your own tastes.

[Uberdude]

Games I play at the club are a relaxed affair, we are there to have fun after all, and people play at their own pace. Some are fast, some are slow. Games usually last somewhere between 30 minutes and 2 hours. We don't use clocks, and they are rarely if at all used in club play in the UK, but it seems using clocks in clubs is normal in some other countries.

[lemmata]

If you enjoy longer thinking times, then by all means use longer thinking times. As I said, the purpose of playing is to enjoy ourselves. Just make sure that you and your opponent are on the same page. Don't make your enjoyment come at the cost of his. Personally, I don't mind if my opponents want to use a lot of time, but I would consider it polite if they asked whether it is OK beforehand.

If I visit a new club, I like to know what people expect. Every club is different, after all.

Well said, Herman. However, I think that you would agree that both sides should be considerate. I consider a player who signals his opponent to move quickly during a game to be far ruder than a player who plays at a pace that his opponent does not like. The first form of rudeness seems to be far more prevalent than the latter. That said, it is best to avoid conflict altogether if possible. In that regard, your method of inquiring about preferences beforehand seems like excellent practice. Also, I find that proper handicaps help alleviate asymmetric time usage.

FWIW: There have been so many times when players on KGS have said immature things to me for taking "too long" to play in automatch games. These are games in which my opponent specifically decided to seek games with the time settings that were ultimately chosen! This makes me suspect (only suspect, but still...) that people who complain about their opponents' time usage are probably throwing temper tantrums based on self-centered and self-serving faulty logic. One of the benefits playing on Asian servers is that Asian temper tantrums show up as random glyphs.

[Discipline]

Interesting and useful post for me as well, to get an understanding on sensible time controls for a DDK.


I have been having some trouble playing too quickly , and alas sometimes automatically, and making some very stupid mistakes like missing ataris, potential ladders and cuts.

I'm always a bit apprehensive that my (usually) stronger opponent might get irritated or laughing behind my back (while I'm trying to figure out if an obviously dead/alive (L,J..) group to him/her is alive )

[Bill Spight]

Well, I see that this discussion has heated up.

I had planned to write a long post drawing largely upon my own experience and observations. But let me start with my main conclusion. I am not sure what I am going to say after that.

I think that a better question than how long should a beginner think is how should a beginner think?

I can't say that I really know the answer, except that I am pretty sure that a beginner should not do deep calculation of variations.

That said, that is not what I did during my first year. Well, I did start out not reading at all. But before too long I was trying to read out local situations as best I could. Even so, my games lasted about 1 hour (15 sec. per move or so, on average). Still, what catapulted me into the SDK range (to about 7 kyu) was stopping after I had done my reading and taking a second or two to look at the whole board -- just look. Often I would see a better move somewhere else. My local reading would usually pay off later. This is the kind of thing that Kotov makes fun of, but all I can say is that it worked. I jumped about four stones in a week.

I learned go when I was already an expert at bridge. Bridge is a faster game than go. Freddie Sheinwold said that he preferred a quick mistake to a slow one. It was much less embarrassing. Terence Reese advised, Don't dither. In chess Kotov makes the same point without stating it directly. No disrespect to my SDK friends, but when they take two hours to play a game, my guess is that there is a lot of dithering. One idea that I picked up from Botvinnik about studying your own games --which every serious player should do, even beginners --, is to note which plays you think about the longest. Such positions are usually difficult for you, and you can benefit from studying them. I found that noting which plays took me more than 1 min. worked for me.

When I comment on a game here I will often just suggest one or more moves. It isn't just that I am saving myself time. When reviewing kyu games I have found that a common response when I suggest a play is, "I never thought of that." What good is all your thinking if you never even consider the best play?

Which brings me back to my question about how to think when playing go? It seems to me that most SDKs should broaden their scope, so that they are more likely to consider the best play. But beginners? How should a beginner think?

P. S. There is tension between playing to win and playing to learn. IMO beginners should play to learn.

[Bill Spight]

One of the benefits playing on Asian servers is that Asian temper tantrums show up as random glyphs.

[oren]

P. S. There is tension between playing to win and playing to learn. IMO beginners should play to learn.

When is someone not a beginner?

Ok, ok... all new tangent.

[billywoods]

In chess, which is also based on pattern recognition, blitz games are generally accepted to only be mindless fun, and to improve you must play slow games and really think things through.

I suppose it depends what you want. I reached 8k by blitz alone (and I've met a few people who say they got much stronger by blitz alone).

Consider the case of a complete beginner who misses ataris. While it's true that they would eventually mostly spot the ataris after looking at each intersection individually if they took a minute per move, that is somehow the wrong thing to train. The correct thing to train is that instant *ping!* feeling you get in your head when a group comes into atari. In this way, the beginner has streamlined the process of reading, and can now read one move further ahead in significantly less time than before. So you would of course tell the beginner to try to spot ataris while still playing at 5 seconds per move, and very quickly various atari shapes would fix themselves in the beginner's head simply by repeated failure. That seems to be how most learning works - you learn most efficiently when trying to do something that's only slightly harder than you know how to do.

Well, your case and mine are the same: people 6 ranks above me probably spot some shapes instantly that I struggle for a while on, and it's simply not efficient for me to spend 15 minutes trying to read 30 moves ahead along 5 branches which contain double hanes and squeezes and snapbacks in the corner. If someone pointed a gun at my head and told me to read it out, I probably could, but it's too hard to be useful. Most of my time will be wasted.

In the same way, if you can't work out whether something's a snapback or not, don't spend 10 minutes trying to work it out - spend 15 seconds on it, go with your intuition, and review the game afterwards. After 10 games like this, you'll be able to spot snapbacks easily. If you can't work out whether some cutting points are dangerous, do the same.

(By the way, "blitz" is very much a relative term for this kind of reason, and the kind of reason the Korean 5 dan mentioned - the stronger you get, the more you can usefully think about.)

[billywoods]

I notice my post was entirely about local reading, which may well serve to back up Bill Spight's point that kyu players don't look around them very much. I think the same sorts of things can be said, though - my intuition for when a situation can be profitably ignored and tenukied has improved mostly through getting it wrong over time, rather than through spending 15 minutes each game counting the precise point swing if I tenuki. Or something.

[PeterPeter]

my intuition for when a situation can be profitably ignored and tenukied has improved mostly through getting it wrong over time, rather than through spending 15 minutes each game counting the precise point swing if I tenuki.

That has hit the nail on the head for me. Thank you.