Flattering percentiles?

Question is - how much the 100% is diluted with inactive players?

A rather mediocre 1700 cannot be so high in percentiles.

I'm not sure inactive players figure into it but it is very possible. In my national federation I'm at 83.9 Percentile with an OTB rating of 1539. One of my friends at 1751 is 90.1.

 I'm pretty certain that the percentile is based on active players.  Take a look at the global number at the top of the stats page.  That's too small to include inactive players.

Also, after a year of not playing blitz, my stats are now N/A.

When I look at USCF data, about 90% of the active players are rated under 1800, so it's not surprising that an 1800 or so Chess.com rating is 96+ considering that the pool of players is probably weaker overall.

Sorry, could you explain to me what a percentile is?  I've seen it on my stats page.  E.g. 86.3% percentile.  Does that mean that 86.3% of people are lower rated than you, or higher rated than you, or what?

The percentile is the % of players that are rated lower than you.

Thanks.

Ok, looks like these percentiles are not lying after all. Feels a bit like a paradox seeing how much further one can go in rating. Must be some kind of Cause curve with apparently endless sides and a rather fat and bulky middle.

Meant to say Gauss curve.

The rating formula was developed by statisticians.  They want to have a curve that resembles a standard distribution. 

Hmm .. I beg to disagree. Love and are not extinct and lie still did not replace truth everywhere.

Love and loyalty are not extinct.