Why would someone's correspondence rate have a big difference to his classical rate?

Maybe in correspondence blunders are rare?

Since you have loads of time to think, yes blunders are rare
OR
They cheat using an engine

@deadlyinstincts said in #2:

Since you have loads of time to think, yes blunders are rare
OR
They cheat using an engine
Or maybe they could use the analysis explorer?

Different pools.

If u see the profiles of players who play correspondence games then u can see most of them don’t play time controls other than correspondence. Most correspondence players are cheaters. They r only a few players who play fair correspondence games.

In correspondence chess you will sometimes win games from equal or inferior position, just because your opponent is away from the computer for a day or two. When real life gets in the way of them moving in time, you win. So if you make sure not to lose any of your games on time yourself, this alone makes a bit of a difference.
I find timeouts in classical timecontrol relatively rare compared to correspondence, and in classical it is often just rage quitters who time out when they choose to "punish you" by making you wait till their time runs out.

I have the opposite problem. I have a lot of time in correspondence, so I usually don't think enough on my moves.

Ratings are relative i.e same number on different pool does not mean same thing. Lichess glicko starting rating pull median rating to 1500 hence that is roughly the median for all pools. But it is not the same people who play in these pools
Blitz draws more strong players than rapid which draws more strong players than classic which draws way more than correspondence.

hence normally you would that with same level of skill Correspondence rating > Classic > rapid > blitz
Think of this way; Bunch of granmaster would start the lichess clone with no seeded rating since they would win each other about 50% of time the rating average would be 1500. rating does not measure skill but ability to win.
Sam thing for bunch of pre-schoold kids having their own clone - average would still end up on 1500.

@IsaiahCui

There's one big difference. In Correspondence you can legally use the analysis tool. That with no time pressure it means players don't have to visualise any calculations. This makes the accuracy level higher and lowers the number of blunders.

I don't want to open the can of worms of cheating, but it's harder to tell if average players are cheating because players are genuinely more accurate when they play Correspondence.

(Assuming you mean why are some players Correspondence higher than their Classical ratings).

And then for some players they get off to a good or bad start and they artificially start around 1400 or 2200 because of the extreme provisional Glicko system. In Bullet you can level off your true rating in a few days. It can take many months for players to get to their real rating if they're playing less than one game a day.

Your ratings are a good example of that - 22 games isn't enough games for the ratings to level out.

no more clock distraction.