Chess.com VS USCF Rating

Blitz ratings are better predictors. One reason is that there is less of the notorious "extenuating circumstances" in blitz. Even so, it is a poor predictor.

If your rating is increasing here, you are likely improving some aspect of your play. Ratings are really only "accurate" within a pool of players.

You can get inaccuracies in regional players OTB, that play relatively the same pool of people all the time. They then go to  a different area and have worse/better results than their rating would indicate, without any real change in chess knowledge/strength.

Also, look to see if you continue to make the same types of mistakes; if you are, you probably aren't improving much

Very similar to me.  1938 USCF and 1701 Blitz here (all 10min games).  I scored 1.5/3 in my last OTB RR so I'm hanging with the 1900-crowd in the USCF but find it hard to do so here against 1700s in 10min blitz

my USCF rating is somewhere in the 500s while my chess.com rapid rating is 994. kinda weird but yeah.....

 10 0 and any blitz with increments adds the likelihood of a certain leveling of skill due to extenuating circumstances.

"extenuating circumstances"

The rapid ratings overestimate strength at the lower levels but then underestimate probably somewhere around 1900. Unlike the blitz ratings, which go up into the 2700s for the super gms that use the site, the standard ratings are all generally under 2000 or just over 2000. They attempted to make the ratings more accurate by giving players extra rating points sometime back but they are still too low.

one way to test it is finding a player who plays just as many blitz,daily,rapid and OTB games and see what there rating says which is very tedious although I may do it just to test all of this.Ill start off by not going off rating but how much rating I gained over time in all formats. I cant do OTB tho since im pretty inactive but perhaps I can simply play some Classical timed games against even leveled opponents. The one big flaw with all this is allmy opponents would have to be the same strengh but that means knowing already how accurate chess.com ratings are to begin with. A pretty annoying paradox.

10 minute blitz is full of engine users, so the ratings are deflated. Stick with 3 minute blitz - I have played the great majority of my blitz against 3 minute players, and there are 0 account closures for cheating amongst them.

My blitz is 3 0 and 5 0 almost exclusively. When I go back a month in my history, I see 2-3% account closures.

About half of my 10+0 games, including unrated, have been engine users, and blatant ones at that. 

I take my online rating seriously enough so that I can motivate myself to play seriously.

It has no value other than that, compared to official ratings.

I used to play a lot of 60 min games on ICC, and my rating was solidly around 1600. My USCF is 1907. If the ICC ratings and chess.com ratings are similar, then Id say that online chess and otb rating correlations just vary from person to person. For me, the 2d vs 3d and the home environment against a tournament setting is just too different.

I think your guess of around a hundred points difference is pretty accurate. My USCF is higher than my online rating here and on other chess websites, but my bullet and blitz ratings are a lot lower of course because I make so many mistakes on the shorter time controls.

I believe there is a serious rating disparity. For example I have a 1225 Rapid/Classical rating on chess.com, and my OTB is a weak 505, this is 700+ difference, minus some considering the fact I find it more difficult visualizing OTB. If anyone has any idea how I can improve the gap between my chess.com rating and my OTB rating please comment.

Remembering online games have many disturbances such as internet issues,lack of concentration,etc

More like 1000-1300 according to the post. USCF ratings themselves are overstated to FIDE. Maybe to stay on the safe side we say its at least 500. I had no idea I could be 1400 away from 2000 according to Silman.

Forget Silman's comment.  Isn't the default here for new players 1200?  I would think the further you are from 1200, the more accurate your rating is (and perhaps more comparable to USCF).