Elo estimate is goofy ahh

Yesterday I (500 elo blitz player) played a game where I played 19 moves (would be more but my opponent resigned), and I game reviewed it today (i had already used yesterday's game review) and saw I had not even a single inaccuracy and my accuracy was 96.7. Game: https://www.chess.com/analysis/game/live/78352283477?tab=review. So how is my elo estimate 1300 when according to the insights feature, Hikaru has an average accuracy of only 88-89? There is one more question: in move 17, why is Rd1+ better than O-O-O+ when the difference is my king is safer and my rooks are connected when I castle queenside (which is what i did)?

It's based on the opponent's elo

huh. my friend @scroogemcduck12345 from school is like 1400 and he once got a game with an estimate of 2050. he is def better than me, the other day he played a game with 2 brilliants

I prefer one that is not based on the elo. After all, I may accidentally be brilliant, or blunder prone. An abnormal elo will be more apparent than simple accuracy meter.

yeah, i find it odd that i can get 96.7 accuracy and get an elo estimate of 1300 (https://www.chess.com/analysis/game/live/78352283477?tab=review) but @GothamChess can play a game with 63.2 accuracy and get an elo estimate of 2000 (https://www.chess.com/analysis/game/live/78678806093?tab=review). What gives?

Scoring 96% against a beginner is easier than scoring 60% against a master. A beginner will make obvious blunders that are easy to punish. A master will not blunder and will make your life difficult.

Other than that, this whole Elo-estimation thingy is nonsense. So of course the results will also be nonsense.

He played well too except he had one blunder