Psychology in online bullet chess
The evidence of this is even more overwhelming than OJ's guilt as a murderer.
The proof lies in the fact that the very best players in speed chess are also the very best at time controls in which the game may last several hours, such as Magnus Carlsen's triple crown (world chess champion in slow, rapid, and blitz) and then double crown (winning both the classical and rapid and taking second place to Grischuk in the finals of the blitz), the very next year. The top one-minute online players, such as Nakamura, are all elite super grandmasters in classical chess. The only thing required is to arrange a series of games between the same two players, to even out the chance factor of a probable blunder in any single game.
However, there are some fascinating 'tricks' of the trade which tend to be more effective, or in some cases are only possible, in an online bullet game. One such example of the latter is the judicious use of the "premove" feature, which is afforded at all decent online chess servers. This feature allows you to set up your next move to execute immediately, with no drain on your clock from the time that your opponent makes his (her) move to the time his (her) clock starts again for his (her) next move. Some servers, such as chess.com, try to mitigate the tremendous power of this wonderful feature by inducing a small drain of time, such as a quarter of a second or so, every time premove is used. But this is certainly much less than the amount of time that would have been expended if even the fastest hands were to make the move manually. There are many situations where premove is obvious and with no potential consequence, such as when your premove will result in the forced recapture of a piece, or when your premove involves recapturing a piece that you expect your opponent to capture. That way, if your opponent doesn't capture the piece in question, your premove is benignly cancelled and you need not fear a blunder being made. The real skill in using premove occurs when the situation is more ambivalent, such as when the move being premoved is likely to make sense, but could end up being a terrible blunder if your opponent does something you didn't expect. I have dubbed this phenomena the premove error. An obvious example of this would be when you want to shoot out a bishop say to 'b4' or 'g5' to pin the enemy knight, but all of the sudden you find out that your opponent has played the prophylactic 'a6' or 'h6', and now you are staring at a board in which you are about to drop a wholesale piece!
But there are many situations in which a premove error is much less obvious, such as when you have captured material and have premoved something in full confidence after your opponent's recapture, but there exists a subtle intermezzo that you have overlooked which now turns your premove into a blunder!. There is also the practice of trying to induce a premove error, such as when your opponent senses that, in order to obtain a time advantage, you have been premoving all of your standard opening moves with confidence that these "generic" opening moves are good against any standard reply, such as creating a "London System" type setup. But at some point there might exist a vulnerability, such as an unprotected bishop when you have just played 'g3' with the idea of fianchettoing (would you believe the blogger spellcheck wanted to change this word to ghettoizing?) your bishop to 'g2'. So now your opponent, who suspects you will probably premove the natural-looking Bf1-g2, bangs out Bh3, which is hanging but you didn't take it because your move was premoved. Now he (she) will win at least a bishop and maybe a whole rook (or both), depending on whether or not your knight is still on 'g1'! This is why I liken using premove in situations where there is some risk involved to stealing bases in baseball.
In baseball, you can only stray away from the base so far before the pitcher gets the idea of whipping the ball back to the current baseman to nail you out. Similarly, in chess, an astute opponent will start to get the idea that you're getting a bit too greedy with all the premoving you're doing and will throw in the occasional hanging-zinger (another technical term I coined to describe one of the many premove countermeasures) to keep you honest. You premoved with confidence because you knew that your opponent had no "good" move to make your premove bad, but you didn't take into account the hanging-zinger, which is a blunder if you didn't premove, but a double-exclam if you did.
These are the kinds of mistakes normally made only by amateurs. A strong player will very rarely make a premove which could transform into a premove error, unless they (or their opponent, or both) are in draconian time-trouble and now don't mind throwing away pieces to give check, or even better, placing a piece beside the enemy king to "zug" it in the directions that the piece controls, so as to invalidate the opponent's probable premoves as he was likely expecting you to give check. It is difficult to quickly (meaning fraction of a second) assess that an enemy piece, especially a heavy piece, next to your king is hanging, as psychologically you are expecting it to be protected by another piece as would usually be the case in a normal game. So even if you didn't premove, you are forced to reckon with the fact that there is an enemy piece threatening to molest your monarch and this will likely distract you from your other plans and cost you time on the clock. This is one of the "higher-level" premove countermeasure techniques that I use myself, although to be honest it comes up quite rarely. Perhaps less than five percent of the time. In the vast majority of the games it's straightforward and safe use of premove that is all that is required. One or two servers even offered a feature dubbed "aggressive-premove" in which you could set up a series of premoves to all execute (assuming still legal in the resulting positions) immediately after each of your opponent's moves. This is useful, but not easy to calculate, in situations where a series of moves are forced.