Did this guy take me seriously?

Indeed there was some people here that are rated 3200.
It seems to be a fake!.

Even Magnus Calsen does not have this rate.
3200 is the rate of chess.com max engine.

Chess.com ratings are easier than FIDE ratings. The maximum engine's rating is supposed to imitate a FIDE rating, I think. The Chess.com bot ratings are notoriously inaccurate, though. 

maybe he plays only unrated games

plot twist: @alphaous is a GM in disguise

Ur a secret prodigy.

Edit:  Well... it's not a secret anymore. Now everyone knows. 

Lol, true, but the games were just there so you would have a little background. One of the reasons why I don't play much bullet is because you cannot analyze the games with a computer without getting depressed.

But I'm not five! I'm a little older. 

Edit: There's no reason for him to assume my age, and there's no reason for me to disclose it.

How did you find out?

Nooooooo.

In the 1997 I made a page at Geocities. A page about HP48 calculator.

There was a boy claiming to have over 1 million visitors to his site. This almost made me give up on making my website. Then I saw that it was fake. He tampered with the access counter and put in there the number he wanted. Maybe there are people able to change their data here on Chess.com, putting there the rating they want. 

Who knows?

Almost certainly a case of them playing blindfolded, hand and brain with a lower-rated partner (this one seems particularly likely given they were so slow in every single game), or someone else entirely being behind the computer. Alternatively, they could very well have cheated to get to 2000 in bullet ─ this seems especially plausible given their sudden jump from 1300 to 2000 in just over a month in September 2017.

I'll tell you why I'm so confident in my prediction:

1. That same day, the account had played a total of 51 games, with most of the games being against opponents in the 500-1200 rating range, and many players below 900. If you look at the account's winrate in sub-900 games, it in fact works out to be 72%, which would put its real strength at around 1000. This is consistent with your experience, as that's exactly what you'd expect the elo to be given an 80% lose rate against a 1000-1500 player that is yourself. Furthermore, this strength estimate lends credence to the hypothesis that they cheated to get to 2000, as it is close to the rating they had before that rocket-fast jump all the way to 2000.

2. Even though most of the account's games since creation seem to have been rated, all the games that he played that day were unrated, which suggests there might be a good reason for their decision to play unrated that day. Usually, people play unrated either because they don't put their rating at stake or because they want to do something which would be against the rules in rated chess. Both of these reasons are consistent with all of my explanations ─ if they were playing blindfolded or had a lower-rated acquaintance play on their account, then they would simply lose rating points unfairly had they decided to play rated instead, and if they were playing hand-and-brain, that could be construed as using outside assistance and may or may not be allowed in rated chess. My alternative explanation also fits the bill here ─ perhaps they wanted to find out what their true strength was and see if it had improved since they started cheated; in this case, they'd know that, had they played rated, they'd simply lose every game and lose all the rating that they cheated so much for, hence the decision to switch to unrated all of a sudden.

I think the 4 options that I have proposed are the only ones which can adequately explain these two points.

Thank you! That does make sense, especially considering I only lightly surveyed their account. But what doesn't make sense is the fact that he plays almost exclusively bullet, when I've heard that it's almost impossible to cheat in bullet. Why not play blitz and rapid more often? And it seems very difficult to play hand and brain in bullet as well, in my estimation it would take twice as long, and it would be impossible to premove. And now that I've looked closer, he's played over 40,000 unrated games to 1500 rated games. He has over a 50% win rate in both, but still loses regularly. I think you made a lot of good points, someone else said he might have had a friend playing, and I thought it was very plausible, but I didn't consider hand and brain.

ha I can probably do blindfold bullet and still beat a 1000 a couple times... just move fast, premove, and wait for it to end...

maybe it is blindfold, ever thought of asking them?

I happen to know a 1000 bullet. Mind if I ask him if he wants to participate in this experiment?

His rating looks fake. Search his rated games and you'll see he has long series of win vs the same opponent... all wins in 3 moves.

This doesn't happen just once or twice.

His rating looks fake. Search his rated games and you'll see he has long series of wins vs the same opponent... all wins in 3 moves.

This doesn't happen just once or twice

In any case, players in unrated games like this are very often always grossly overrated. Either because they're involved in some form of rating manipulation, or because they're a very good player who doesn't want to play rated games due to e.g. intoxication.

I see. But what makes things really confusing is that he has a good, but not perfect, score against titled players with mixed accuracies, not suspiciously perfect games.

Wow, it's even worse the more I look.

He's won 100s of games that have lasted 1-2 moves.

Chess.com is extremely pathetic for not having automated detection in place that flags this sort of thing.

Chess.com search function is screwing up, so I could only search the last 2 years.

In the last 2 years, I filtered out unrated games and opponents rated below 2000. He's played a grand total of 9 games, and won zero.

In unrated, he's played 2300 games against players over 2000. He's won 13%... many of them in fewer than 10 moves...