let's have a discussion about improvement and blitz games

[often]

alright, i have my answer, but i wanna hear everybody else's viewpoints

who here thinks that they can improve by playing a lot of blitz?

i used to think it could help, but i ended up losing a bunch of games on time and silly mistakes. now i play just a bunch of relatively slow games for the internet (15min 5x30 type) and i feel like i do a lot better and learn more.

i'm sure someone out there thinks it's helped their improvement, so let's hear it.

[Splatted]

I think what matters most is how you approach the game. If you view blitz games as an excuse to play the first thing that comes in to your head it's not going to be that helpful, but if you use it to force yourself to concentrate I think it could be very helpful. In games with longer time limits it's easy to procrastinate or waste time daydreaming etc, but in shorter games you need every second so if you're serious about it you're constantly pushing your reading to the limit and your opponent's time becomes a welcome opportunity to consider other areas of the board.

[jts]

Can I take the inverse position? I think I have improved less than I would have otherwise because I don't play blitz games. It's gotten harder for me to set aside 90+ minutes to play a game, but I don't think it would be hard to play 90+ minutes of Go if I did it all in small chunks. So no blitz --> uncomfortable with blitz --> play fewer games --> slower improvement.

(I don't dispute that blitz games might contribute less per minute of play. Most beginners who say "Why can't I understand this game" have only played two or three games ...but the ones who actually have a hundred games under their belts and are still 20k are playing super-blitz.)

[Phoenix]

I second Splatted on this.

It's also important to make a note about time management. You do know some patterns, and the more you know, the better you will fare on blitz. Typically, these moves take less than two seconds to make. Normally, you've even read these lines ahead of time.

But your opponent will not follow everything you want him/her to do. That's when they play an unexpected move and the board becomes a big tsumego. The best move and follow-ups can take ten minutes to figure out! Some pros are known to spend over an hour to think over a single move when a game hits a critical junction.

In blitz, you have to play one of the first three or four moves that come to mind, based on reading about 4-5 moves deep. Or whatever your reading proficiency is at that speed. This turns Go into more of a game of intuition and 'luck'.

Not that all of this is bad in itself, but your reading will definitely take a back seat to memorization and 'vital point spotting', and reviewing your games will be an exercise in futility, most of your thoughts being "What a simple mistake!" and "Why the hell did I play that move?!?".

My personal take on this is that Go is a thinking game, and the thinking itself should be the enjoyable part. If I want to win or lose at something quickly I'll work on my rock-paper-scissors skills. I hate to be left with ten seconds to reply to a good tesuji, life-and-death situation or a severe invasion. Or even choosing which site to atari a stone.

You can't just leave the stone for later. Not without reading out that it can't do any damage while you tenuki.

So that's just my two cents. In general, I've found that blitz games do nothing for improvement, and I get much more enjoyment, not to mention better feedback, from regular, slow games.

[cdybeijing]

My ideas have changed on this somewhat over the past few years. It also depends a lot on what you consider to be blitz.

The basic fact that I have realized is that time management is a real go skill and indicative of strength. I play mostly medium length automatch games on KGS (25 minutes + 5 x 30s) or 10 minutes + 3 x 30s on Tygem/Wbaduk and my teacher considers these games to be blitz. I use all of my time every game and strive to play the best moves, though it is definitely a fact that the quality of my play deteriorates once I hit byo-yomi. Nevertheless, the majority of my time is spent on the first 100 moves which are the most significant at my level. A quick review of each game seems to make these good learning opportunities.

[Shaddy]

I only played blitz in between KGS 6k and 2d. If blitz did not help my improvement, then it must have been faeries.

[Kirby]

I think blitz games are helpful if you are seriously trying to win. It reminds me of a game I have on Nintendo DS for "brain training".

In this game, you're basically presented with a bunch of easy math problems, one after another. They are all single digit math problems, like the ones you'd learn in elementary school.

Example:
5 * 7 = ?
56 / 8 = ?
8 + 4 = ?
6 * 4 = ?
...

There are like 100 of these, and you're being timed. You try to answer the problems as fast as possible, and it keeps track of your best time. I try to practice this every day. The idea here is not to learn the single digit math problems - I already can solve these easily. But by practicing every day, I can re-enforce these relationships, and maybe improve my thinking on more complex problems.

I guess this isn't the same as a "blitz" game, but I feel that, even if you are trying to think quickly, it can still help.

The key is to do this consistently, and do it mindfully, trying to actually think.

Practice thinking, and you'll get better at something. Besides, it's fun, isn't it?

[wineandgolover]

I only played blitz in between KGS 6k and 2d. If blitz did not help my improvement, then it must have been faeries.

This is the first pro-blitz argument that I actually buy. It proves that it can work for some (well at least one).

Did you review your games to find the mistakes? Perhaps this is where the greatest value/time is?

What else did you do to improve? Still not buying the faeries.

[Boidhre]

I only played blitz in between KGS 6k and 2d. If blitz did not help my improvement, then it must have been faeries.

This is the first pro-blitz argument that I actually buy. It proves that it can work for some (well at least one).

Did you review your games to find the mistakes? Perhaps this is where the greatest value/time is?

What else did you do to improve? Still not buying the faeries.

There's an Irish player who went from KGS 1d to KGS 4/5d almost solely through playing blitz games. He studied some joseki on the side, that was it, no other theory or tsumego. Before 1d however he didn't find it helped him much to play blitz.

[wineandgolover]

I only played blitz in between KGS 6k and 2d. If blitz did not help my improvement, then it must have been faeries.

This is the first pro-blitz argument that I actually buy. It proves that it can work for some (well at least one).

Did you review your games to find the mistakes? Perhaps this is where the greatest value/time is?

What else did you do to improve? Still not buying the faeries.

There's an Irish player who went from KGS 1d to KGS 4/5d almost solely through playing blitz games. He studied some joseki on the side, that was it, no other theory or tsumego. Before 1d however he didn't find it helped him much to play blitz.

Anonymous players mean nothing to me. Shaddy claims to have done it himself.

That said your point about playing blitz before you understand the game (1d, say) makes sense. Maybe blitz reinforces what you already "know" whether that's good or bad.

[EdLee]

I only played blitz in between KGS 6k and 2d. If blitz did not help my improvement, then it must have been faeries.

This is the first pro-blitz argument that I actually buy. It proves that it can work for some (well at least one).
Did you review your games to find the mistakes? Perhaps this is where the greatest value/time is?
What else did you do to improve? Still not buying the faeries.

I agree with wineandgolover and am also curious about the same questions, plus one more: at what age ?
Not sure about fairies, but we have definite data on the age -- at least Shaddy does,
but I understand if he doesn't want to share it -- it's private data.

(a) I believe most (or all) children start Go at blitz timings. (Children = around 10 or under.)
Later, as they improve, they may slow down.

(b) We know most (current) pros start very young (that is, under 10, or even close to 6 years old).

(c) Therefore, most (current) pros start as children with blitz.

Does anyone know of examples of people starting Go later in life (say, after 35 or 40),
who have a similar experience as Shaddy's ? That is, exclusively blitz (no reviews, no Go problems, no books,
no studying, no teachers, no lectures, no workshops, etc. -- absolutely nothing else but blitz),
and made it to KGS 2d or equivalent level ? Let's hear about it.

[gowan]

I have posted previously stating a negative opinion about the value of blitz play for learning and improvement. If you are studying go theory and life-and-death, you need time to think to use what you learned in a game. You can't improve at reading without doing it and fast games don't give you enough time. For example, it is well known that you learn better by figuring things out for yourself rather than memorizing. Suppose you encounter a life-and-death position in a game. Most likely it isn't identical to a position you studied but you think it resembles a known position. Maybe a few stones are placed differently. You need to remember what the key idea was behind the problem you studied before and then read out several variations to see if the basic idea works in the position in your game. You probably can't do that at 10 seconds per move. So what do you do? Probably you just try something and hope it works, hope it makes your group alive or hope it kills your opponent's group. In either case, without thorough reading you could easily be wasting a move. I mean you might be trying to make an already dead group live or kill a group that is already alive. As another example, you are involved in a joseki you don't know thoroughly and your opponent plays a move you don't recognize. The consequences of a mistake on your part could be serious. You need to read it out. Can you do it at 10 second per move? In my opinion the only thing blitz play can help with is rapid recognition of shape. Replaying and analysing blitz games is not so useful either. Since you weren't thinking how can you find and correct wrong thinking?

[Splatted]

I've put my thoughts in hide tags because I know it's ridiculous for a 5k to argue with a 5d about how to get good at go, but I just can't help but think I'm right.

You need to read it out. Can you do it at 10 second per move?

You can't read it out completely but you can still do some significant reading. I think getting the right answer is less important than time spent excercising your reading ability, and short time limits can force you to push it to the limit. You won't be able to work on depth of reading, but speed and accuracy are important aspects that come to the fore in blitz games.

Replaying and analysing blitz games is not so useful either. Since you weren't thinking how can you find and correct wrong thinking?

I don't know if this completely applies to go, but one thing that music practice has taught me is that there is no such thing as a meaningless mistake. It doesn't matter if it's easy and you'd never have made such a mistake if you'd been concentrating, because that just shows your instincts are wrong.

Edit:

Not that all of this is bad in itself, but your reading will definitely take a back seat to memorization and 'vital point spotting', and reviewing your games will be an exercise in futility, most of your thoughts being "What a simple mistake!" and "Why the hell did I play that move?!?".

I think "Why the hell did I play that move?!?" is a question that's well worth asking.

[xed_over]

I think it depends on your definition of blitz. Some blitz is too fast maybe.

Most beginners (not kids) play too slowly, and I think it hampers their improvement.

As someone else already mentioned, you can slow down later, to have time for deeper reading, after you've improved enough to have mostly internalized the fundamentals, which could be reinforced by faster or blitz play.

I just had a discussion with someone at the club this weekend about this. I had a long 2 hour 4-stone game with someone who was losing almost the entire game. I recommended that he might improve faster if he played a little faster, but quickly got dissenting opinions from at least one very vocal onlooker.

[EdLee]

I recommended that he might improve faster if he played a little faster, but quickly got dissenting opinions from at least one very vocal onlooker.

"At the end of the day we are likely to be punished for our kindnesses."