Is it possible to get better at "chess" playing bullet/blitz. New Perspective.

It's a common topic of discussion whether or not someone can get better at chess playing bullet/low time control blitz. Now here's the fact of the matter. Most people that play on this website aren't interested in, and have never actually been to, a real chess tournament. Why would i care about getting better at something that I never plan on doing? 

Now here's a question that is very seldom asked. Can someone get substantially or even partially better, or perhaps just worse, at playing bullet/blitz by playing bullet/blitz? What are your opinions? Conditions for failure/success? Personal stories? 

I'll love to share my own after i get some insight on everyone elses'. 

Sure, I call it being a blitz or bullet specialist. You can far exceed (maybe 400-600 points) what's normal for your classical time control peers. That's how you can see IMs or even FMs pretty high on chess.com blitz or bullet ranks (higher than GMs).

The drawback is you'll hit a wall where your stop gap methods no longer work. I call them stop gap because speed chess only gives you enough time to play the strategy and tactics you know really well... so that means at some point to improve your speed game you have to improve your game overall. That means learning skills that are useful in classical.

As an example, IMO Carlsen is pretty bad with the mouse. Frankly I think I'm (at least a little) faster than him... meanwhile people like Naka and penguin are otherworldly in time scrambles... but Carlsen is better than them in speed games because he's better at chess overall.

Anyway there are other tricks like blitz-specific repertoires and blitz-specific evaluations (like correctly judging the value of initiative, or which endgame is a draw or loss). For example you should choose a lost endgame that is very hard for your opponent to win in 30 seconds over a drawn endgame that's very hard for you to draw in 30 seconds.

Lovely

What exactly does it mean to "improve your game overall" though. That's the real question for me. Aren't classical and blitz inherently different in "correct" methods given the time differences? 

For example in one friendly OTB blitz game I was losing, so I sacrificed my last two pieces for his last two pawns. Now he has 1 minute to mate me with bishop and knight and no pawns... and he did it. You don't learn proper technical endgames as part of being good at blitz or classical, it's knowledge you can use in any time control.

Most knowledge and skills are like that: you can use them in any time control.

Knowledge is a huge shortcut. That's why, for example, Carlsen could play 50 players of my skill at the same time and have good chances to win every game. I would out-calculate him roughly 50 to 1, but he's over 50x as efficient so it doesn't matter.

Being 50x (or 100x) more efficient is a huge help to speed games

For example, lets say I'm white in the game below (and instead of this it's some tactical line) and I offer any GM the chance to trade down into this bishop and king endgame.

I'm guessing a lot of lower rated players wouldn't see what the big deal is, but here white is completely lost, and a GM would be able to blitz out most of the moves for a win... not because they're good at blitz, but because they understand how these positions are won.

So we could do the same for an opening or middlegame. If I make a deadly mistake, they'll punish me for it and I wont even know what I did wrong... not because it's a fast or slow game but because the position itself is losing.

Obviously it is possible to get better at chess playing Blitz (or even bullet).  If people are discussing that and claiming otherwise then they're stupid since it only requires one person out of the whole world to give evidence of improvement in order to prove them wrong!

Well, no matter the time control, I think it's fair to say improving involves studying. In other words if all you do is play all day you're not going to improve much.

In an extreme case that is precisely what Alpha Zero did, to my understanding.  Of course, if you are going to assume that you forget everything you see and experience during each game it's a different matter.

Chess is an analytical exercise for the brain. The brain makes a better job when a number of pathways have been developed. It's known that such brain pathways are better developed in competitive (stressful) situations.

For the outsider, there's no difference whereas no real competition or just seconds instead of minutes to make a decision. The experienced player, instead, knows that certain mental processes are required to achieve precision and that these processes often require more than a glimpse of the situation. In turn, the processes become faster with practice due to the new brain pathways developed or strengthened. The overall result, for the tournament chess player, is that he sees more and becomes more precise, using the same amount of time, because his brain is in training to search deep into the positions.

As for getting better because of blitz or bullet intense and exclusive practice, there are two levels to consider. Better results in those time controls? Probably yes. Better at chess? Nah, because the player gets used to make practical decisions (rival, clock) rather than improving his understanding and precision on the board, leading to what's known as "superficial".

The way machine learning works is there is an adjustment after every, lets call it test.

So a better parallel would be if you made an adjustment after every game you played.

Machine learning basically brute forces the process by making random adjustments, then keeping the changes that were beneficial. Do this an astronomical number of times and you'll get really close to (but often never quite) perfect.

A human can make an adjustment after every game too, it's just likely to be less effective, but there's an assumption that people playing Blitz apparently just forget every game they play and learn nothing from them.

I probably play 20-30 (blitz and bullet combined) games a day. Almost 1000 a month sometimes. Not only do I typically learn nothing, but I'm also reinforcing bad habits.

Of course I'm just playing for leisure and to pass the time, and it's possible for someone to learn from blitz, but I doubt they could learn by playing as many as I do.

I've also played blitz for improvement, where I analyze immediately after every game.

That's nice, but it's also limited. You're stuck trying to interpret the engine. Plus it's not possible to learn all the basics the way you would, for example, with an endgame book where you're collecting all the fundamentals.

And frankly, sometimes the engine is wrong, or at the very least will give impractical advice.

It depends where your start point is and what you define 'better' too.  The original question was vague enough that, as I said, no-one in their right mind could claim the answer is 'No'.

I'd be very surprised if someone who starts off only knowing how the pieces move and nothing else, didn't improve through playing Blitz chess alone, if they have the will to keep playing and are aiming to improve, both of which are big 'ifs'.

It also depends who your partner is. If you played blitz chess with titled players from the beginning, then you'll learn a lot because after the game they'll make a comment or two. Do that x1000 and it's like years of private lessons.

I mean, that also has problems, but anyway, you get my point.

And yes, I agree that probably no one will say you can't improve at all.

There's the rare case of Arthur Dake (1910–2000) who began playing chess at the age of 20 and then mostly Rapid Transit (similar to modern blitz). About a year later he began playing in tournaments... with enough success as to obtain an individual silver medal in the Chess Olympiad 1933 and a gold medal in 1935, besides three team gold medals in 1931, 1933, and 1935. In recognition of his results in the 1930s (he basically retired afterward) he was awarded the IM title in 1954 and honorary GM in 1986.

So it's possible to increase the chess strength by just playing fast chess, but since there are few known cases it's not regarded as the better method.

These are the two main takeaways for me from this thread:

Llamonade2:  "As an example, IMO Carlsen is pretty bad with the mouse. Frankly I think I'm (at least a little) faster than him... meanwhile people like Naka and penguin are otherworldly in time scrambles... but Carlsen is better than them in speed games because he's better at chess overall."

Nicator65:  "The experienced player, instead, knows that certain mental processes are required to achieve precision and that these processes often require more than a glimpse of the situation. In turn, the processes become faster with practice due to the new brain pathways developed or strengthened. The overall result, for the tournament chess player, is that he sees more and becomes more precise, using the same amount of time, because his brain is in training to search deep into the positions.

As for getting better because of blitz or bullet intense and exclusive practice, there are two levels to consider. Better results in those time controls? Probably yes. Better at chess? Nah, because the player gets used to make practical decisions (rival, clock) rather than improving his understanding and precision on the board, leading to what's known as "superficial"."

Sure you can very slowly get better playing very fast chess but you can quickly get better playing slow chess!