Forums

Why’s Blitz so much harder than Rapid?

Sort:
YellowVenom

Blitz and Bullet are easy if, and only if, you have the right mindset. Most of my games are won on time by strangling my opponent with repetitive checks. Limit their opportunities, force them to use up valuable time. With less than 30 seconds left, your opponent will panic and a win should be a formality.

PawnTsunami
AlexiZalman wrote:

There is nothing unusual in this. You can see the same effect in intermediate games, 1500+ - the only the difference is that the 'needle' doesn't swing as much, but in relative terms the consequences can be the exact same. 

That situation becomes more and more rare as the rating of the players increases.  You can also see good examples of it watching Levy's "Noob Arena" series.  The lower rated players tend to just make moves with very little thought, no plan, and hang pieces left and right.  Players in the 1500 range start to come up with semi-logical plans and do not hang pieces nearly as much.

There was a discussion years ago about the differences between players at certain levels.  A GM replied and it went something like this: Below 1200, almost every other move or more is a blunder; at 1500 they only blunder 1-2 times per game; at 2000, maybe once per game; at 2200, once every 5 games, at 2500 once every 10 games.

The numbers may be a bit off, but the idea is the same.  You can see that by looking at the statistics on sites like AimChess.

And with that, when players are blundering every other move, it doesn't matter much if they are 100, or 900.  The match is "even" in that they both will have many winning chances left on the table.

tjt85

It's the speed that I can't deal with. Even in the blitz formats where you get a few seconds back, I'll do too much thinking and run out of time long before my opponent does. And if I compensate by moving too quickly, I make mistakes I'd almost never make in a rapid game. Either way I still lose. At the lower levels of blitz chess, this game is just a whole bunch of weird openings and pure chaos. It's fun (sometimes) and addictive, but I don't think it's improved my play one bit.

I also think the pool of players in the blitz formats may be larger and for that reason full of stronger players at all levels (or at least there are more players out there better suited to blitz than I am).

AlexiZalman
PawnTsunami wrote:
AlexiZalman wrote:

There is nothing unusual in this. You can see the same effect in intermediate games, 1500+ - the only the difference is that the 'needle' doesn't swing as much, but in relative terms the consequences can be the exact same. 

That situation becomes more and more rare as the rating of the players increases.  You can also see good examples of it watching Levy's "Noob Arena" series.  The lower rated players tend to just make moves with very little thought, no plan, and hang pieces left and right.  Players in the 1500 range start to come up with semi-logical plans and do not hang pieces nearly as much.

There was a discussion years ago about the differences between players at certain levels.  A GM replied and it went something like this: Below 1200, almost every other move or more is a blunder; at 1500 they only blunder 1-2 times per game; at 2000, maybe once per game; at 2200, once every 5 games, at 2500 once every 10 games.

The numbers may be a bit off, but the idea is the same.  You can see that by looking at the statistics on sites like AimChess.

And with that, when players are blundering every other move, it doesn't matter much if they are 100, or 900.  The match is "even" in that they both will have many winning chances left on the table.

Not my experience, it's actually very rare for a sub-1000 opponent to just outright hang a piece, maybe a combination that loses a piece, those do happen - pawns being the most likely victims. Sub-1000 players are actually far better than to randomly hang pieces without some sort of end -of-game time handicap. 

Just because a computer evaluates a move as poor after a few seconds of calculation doesn't make a move poor. A poor move is only a poor if and only if the opponent capitalises on it! And if the position is complex enough for one player to make a poor move a player of similar rating is likely to respond with just as poor a move for failing to capitalise.  If a player capitalises on a poor move then generally the complexity decreases, if not you often get runs of 'poor' moves before the complexity decreases. At lower levels these runs of poor moves are longer but once resolved the moves become solid - albeit one player will be at a significant disadvantage. This is a very familiar pattern.

I analyse my games off-line running Arena/Stockfish through the games, often leaving the engine to run for many minutes for certain positions. I also look at the five-best moves all the time, and most of the time there is less than half-pawn evaluation spread across the top five moves. If say, d4 is the strongest move as determined by the engine, then if I don't play d4 chess.com analysis will flag each subsequent failure to play d4 as a blunder, yet my opponent during those subsequent moves never countered the possibility of me playing d4. How are these subsequent moves blunders? All this is aside from the fact that more often than not playing the strongest computer move leads to more a complex position and therefore I would never make the move even if I had known it was the strongest. 

Also note that high-level youtubers and bloggers etc are more likely to select lower-level games that correspond to an expectation of the gameplay, yet these high-levellers never actually play low-levellers, indeed if rating was involved, they would avoid low-levellers like the plague!

PawnTsunami
AlexiZalman wrote:

Not my experience, it's actually very rare for a sub-1000 opponent to just outright hang a piece, maybe a combination that loses a piece, those do happen - pawns being the most likely victims. Sub-1000 players are actually far better than to randomly hang pieces without some sort of end -of-game time handicap. 

That is just factually incorrect.  The reason they are sub-1000 players is because they hang pieces!  Here is an example from another forum post earlier today:

 
AlexiZalman wrote:

Just because a computer evaluates a move as poor after a few seconds of calculation doesn't make a move poor. A poor move is only a poor if and only if the opponent capitalises on it! And if the position is complex enough for one player to make a poor move a player of similar rating is likely to respond with just as poor a move for failing to capitalise.  If a player capitalises on a poor move then generally the complexity decreases, if not you often get runs of 'poor' moves before the complexity decreases. At lower levels these runs of poor moves are longer but once resolved the moves become solid - albeit one player will be at a significant disadvantage. This is a very familiar pattern.

As you can see from the game above, there were blunders (from both sides) everywhere.  And I'm not talking "Black missed this 5-move combination that wins the game".  I'm talking "my queen is hanging and I do nothing about it for 3 moves".

And the idea that a poor move is only poor if the opponent capitalizes on it is flawed.  If a move is bad, it is bad.  It doesn't matter if the opponent sees it or not.  If you continue playing it, eventually a stronger player will punish you for it.  That is why they are a stronger player.

AlexiZalman wrote:

I analyse my games off-line running Arena/Stockfish through the games, often leaving the engine to run for many minutes for certain positions. I also look at the five-best moves all the time, and most of the time there is less than half-pawn evaluation spread across the top five moves. If say, d4 is the strongest move as determined by the engine, then if I don't play d4 chess.com analysis will flag each subsequent failure to play d4 as a blunder, yet my opponent during those subsequent moves never countered the possibility of me playing d4. How are these subsequent moves blunders? All this is aside from the fact that more often than not playing the strongest computer move leads to more a complex position and therefore I would never make the move even if I had known it was the strongest. 

Again, I'm not talking about moves that shift the position from +0.25 to +0.5.  I'm talking about moves that shift the position from roughly 0.00 to +3 or better and then back to -5 or worse.  As you can see from the game above, they both made those errors many times during that game, and that is a pretty typical game for that level.  The way they get out of that level is by not making those wild swings (i.e. not leaving pieces hanging!)

AlexiZalman wrote:

Also note that high-level youtubers and bloggers etc are more likely to select lower-level games that correspond to an expectation of the gameplay, yet these high-levellers never actually play low-levellers, indeed if rating was involved, they would avoid low-levellers like the plague!

Not true at all.  Hikaru, Danya, Levy, Magnus, Svidler, Bartholomew ... they have all done various streams (speed runs, banter blitz, climbing the rating ladder, etc) streams where they play people as low as 300.  I cannot think of a single speed run by Hikaru and Danya (for example) where they lost to anyone (legitimately) before reaching 2500+.

Granted, this is online rating and none of them care about their online rating (especially for speed run accounts).  But even if they were playing an OTB event, they would not fear playing 1000 1000-rated players at once (other than the exhaustion from running from board to board and the boredom from seeing pieces fall off the board left and right).  If you think any master-level player is worried about 1000-level players, you have not been exposed to any masters.

PawnTsunami
ALinkToChess wrote:
Blitz is just a harder pool, without a doubt. Blitz at 900 is roughly equivalent to anywhere from 1200 to 1400 rapid. I bounce around 85th to 90th percentile in rapid (1.5 mil to 1.1mil rank out of ~10mil) and 65th to 70th% in blitz (1.8 mil to 1.5mil out of 5.3mil). This trend is the same for a large amount of my opponents who have played in both pools recently.

The average for the pools is 792 (rapid) and 814 (blitz) right now - with 23m and 9m players in each pool, respectively.  In short, the numbers do not support your assertion (in fact, just the opposite).

ALinkToChess wrote:
Comparing masters to 1000s is silly but there are levels to chess and dismissing someone’s skill increase just because they have only played for maybe a year or two compared to someone who has played their whole life is not a good idea.

That was not the point.  I was addressing @AlexiZalman's assertion that strong players would not play much weaker ones if rating was on the line.

magipi

It is amazing that everyone just misses the point.

The advice to beginners goes "if you want to improve, play slower games and think about your moves".

It has absolutely nothing to do with how easy or hard it is or where you can get rating points faster of slower.

 

PawnTsunami
ALinkToChess wrote:

Yeah I know the player pools, but if you go to your stat page, it shows your percentile and rank based on active players, 25 games in the last 3 months I think or something along those lines. If you use math, you can figure out the numbers based on rank and percentile of active players in each pool, which works out to the numbers I posted. Average is different than median, and if you check in to that you'll see the numbers work out. I'm sure you're very intelligent with your Gordon Wood and all that, and obviously you are are a very strong player compared to sub 1000 "scrubs", I'm just making a point that sub 1000s are not blundering every other move, and the skill level for two 1000s playing each other is actually less of a clusterf**k than the ivory tower elite like to make it out to be.

Oh?  It isn't?

Every so often you will see a game that doesn't look terrible (like the one you posted in #26), but most of the time games between 2 sub-1000 players is riddled with mistakes - and I'm not talking about small mistakes (like playing d5 instead of dxe5 against the Philidor when Black plays Bg4, but mistakes that drop material instantly).

That is not to disparage players in those rating bands (we were all there at some point), but to refute the notion that somehow players that are rated 1000 are somehow stronger than they actually are.  The simple fact is they are rated <=1000 because they are not yet strong.  In the same way someone who bench presses half their body weight is not strong - they simply haven't gotten there, yet.  Trying to say that rapid players with a higher rating are somehow weaker than blitz players with a lower rating is 1) comparing apples to tomatoes, and 2) depending on the gap, not even close to accurate.  You can get "good" at faster time controls by making bad moves fast.  That does not make you a stronger player, but a faster clicker.

Praveen_bhat97

My understanding about ratings is this! Once you master some basic strategies, openings and certain patterns , your game automatically improves! This automatically impacts your rating positively! But , the catch here is , to improve your understanding of chess , you should play longer time control games where you can clearly execute your ideas ! So, once you go to certain level in such format, it affects your play strength in shorter time control games also. Now, even in blitz you can come up with ideas very quickly! So, it finally boils down to your understanding about the game! 

That's why, you can see top GMs play precisely even in bullet games! It's because, they have taken their own time before to master their understanding about chess! So, it doesn't matter if you give them 1 min, 3 min. They just play their game!

AlexiZalman

Don't consider the Knight on e4 to be hanging, as there is a combination required to win it.
Also note that 9 g4 is playable, albeit with a -0.7 position evaluation to follow but with the Queens plus four minor pieces off the board, still a winnable game at these levels.
Even after Bxe4 Black's King is going to be kicked around - Qd4 followed by Qa4+.
Indeed, White may even end up with an endgame consisting of three passed pawns on the Queen's side against an un-castle-ed Black king on h6 for the minor piece gain. Not a clear-cut win imo. As far as I can see taking the Knight left the Black King vulnerable hence the priority given to castling, which could have lost the bishop - f5! being the only move to keep it, which the lowbie 650 found! 12 .. Rd5?? there is only ONE move to keep the bishop, the lowbie 650 didn't find it, and I am sure a few higher-level players wouldn't find it either having blown half their time by this stage!

Our definition of 'hanging piece' clearly differ. For example, on move 17 I wouldn't consider the Queen to be hanging as it can be moved to avoid capture.
A hanging peice to me is one that can be immediately captured with little consequence, chess would certainly be a different game if you had to avoid moving piece only to protected squares.
And this is aside from working out what piece protects a Queen.

Also note the game timings, Rapid 10mins:-

Move 13 4.36 (Black) against 8.39mins (White).
Move 20 1.58 against 7.41mins.
Move 30 0.52 against 6.51mins.
Move 40 0.33 against 5.48mins.
Move 50 0.23 against 5.23mins.

It's not hard to guess the actual gameplay here. Black burned time on resolving how best to win the Knight in a complex position. White realised Black was short of time so went for a flagging, thereby compromising their winning position out of the opening.
The gaming tactics may have been dubious, but I have no doubt these players would have seen most of their 'computer blunders' within seconds if viewed outside the context of the game.
What made this game poor wasn't low ability but poor gaming strategy. This is one reason I would encourage players to WATCH higher level games as it's not all about 'computer says'.

The game had a fairly complex opening position followed by manic flagging gameplay, I would expect a computer to record a few blunders in such circumstances. Might be an interesting exercise to repeat the same process in a 1500 v 1500 game where one player blew half their time by move 13 and the other didn't. You would see fewer 'computer blunders' but I suspect quite a few strange moves. In fact, feel free to pick one of my own bot games if you want to see how well (or not) lowbies can play Chess without time constraints happy.png

 

That aside: some of your computer comments were wrong, analyse the game off-line with an engine capable of running for more than a few seconds per move at a depth of 18-20ply- as I did! 

Eck I only saw the first part of your post!

Kraig

I believe blitz just has a stronger player pool overall.

Virtually all of the titled players exclusively play bullet or blitz and probably less than 5% of them play rapid online. This likely skews the ratings of a particular pool which will make it difficult to directly correlate a particular rating with a particular level of skill when trying to compare rapid and blitz.

Online chess is for fun. My advice would be to play what you enjoy.

Contrary to popular advice, I improved as an adult learner by almost exclusively playing blitz (Only 10% of my total games have been rapid, and probably 90% of those have been 10 minute games with 0 increment). I was 600 in 2019 and am 2100+ now.
On one hand, rapid does give you extra time to 'think' and come up with ideas OTB, but you can substitute this by learning ideas from study material, videos, exercises, etc and then implementing that in blitz too.

AlexiZalman
PawnTsunami wrote:

...

And the idea that a poor move is only poor if the opponent capitalizes on it is flawed.  If a move is bad, it is bad.  It doesn't matter if the opponent sees it or not.  If you continue playing it, eventually a stronger player will punish you for it.  That is why they are a stronger player.

...

I would disagree, if a poor move results in no immediate negative consequences how is it a poor move in the context of a single game? Sure, a higher-level or for that matter a lower-level player might capitalise on a move but it's not poor till they do so. Afterall there would have been some purpose to the move, however mistaken. It's only poor when the opponent recognises it's poor, if they don't it can be an excellent move! I've won many a game with a poor move because my opponent only saw the threat - I have lost a quite a few in the similar manner too.

Just because a computer or GM would immediately pounce on a poor move isn't of any consequence to the person I am playing or at least it shouldn't be, 

Heck, I have sometimes played a move just because I have spent so long trying to figure the consequences, - and failed to do so with any degree of certainty - that I just play the move to see what happens. It's a game, why not? My thinking is that if I can't work it out, then there is every likelihood my opponent of similar level, can't either.

AlexiZalman
PawnTsunami wrote:

...

Not true at all.  Hikaru, Danya, Levy, Magnus, Svidler, Bartholomew ... they have all done various streams (speed runs, banter blitz, climbing the rating ladder, etc) streams where they play people as low as 300.  I cannot think of a single speed run by Hikaru and Danya (for example) where they lost to anyone (legitimately) before reaching 2500+.

...

Doubt they would do this if there was any risk of losing FIDE rating.

I also suspect if they did legitimately lose to a much lower rated player, the player would be instantly banned by chess.com. When I play these guys I always let them win happy.png

I am pretty sure there are plenty of 1500+ players on chess.com that would not welcome being drawn against a sub-1100 player in a tournament. 

PawnTsunami
AlexiZalman wrote:

Don't consider the Knight on e4 to be hanging, as there is a combination required to win it.

After the move played (Be3) there is no combination.  The knight is simply hanging to Qxe4 and Black is up a whole piece.

PawnTsunami
AlexiZalman wrote:

Doubt they would do this if there was any risk of losing FIDE rating.

When a 2200+ player plays a sub-1000 player, there is no risk of losing unless he happens to suffer a heart attack mid-game.

AlexiZalman
PawnTsunami wrote:
AlexiZalman wrote:

Don't consider the Knight on e4 to be hanging, as there is a combination required to win it.

After the move played (Be3) there is no combination.  The knight is simply hanging to Qxe4 and Black is up a whole piece.

Fair enough, my mistake, Qxe4 if played would have been hanging piece. - only looked at White's possible responses to Bxe4 which were complex.

However, would still argue that the time element had a lot of to do with the quality of the game, one player was very short of time for most of the game whereas the other was going playing for a flagging.  The latter is a situation that can occur in mid-level games and albeit more frequent in lower games, it's not really reflective of chess ability but more of poor gaming tactics.