Of course, which is why you have to track the exact moment when you roll from 1% to 5%, as it gives you the smallest error for estimation of number of players.

Being in top 1% with certain rank indicates that the rank is enough to fit to top 1%, but says nothing about where the limit is. So player who is in top 1% with rank 1555 knows that his rank is enough for being in top 1%, ie. there is at least 155500 players, but that is lowest limit: it could be 155500 players but also 2M. Which is why you wait until you drop from 1%.

I will take @dawnsempires as an example:

He took last 1% screen at rank 12394, meaning there was **at least** 1239400 active players because there was enough players for 12394 of them being in top 1%.

His first 5% was at range 12620, which means that 12620 players cannot be 1%. He is in top 5%, suggesting there is **at least** 252400 players: which is correct, since it is the lowest limit. **But** we know he just fell of 1%, so he won’t be at the bottom of top 5%, more like it will be top 1.01%. So from his position we know there is at least 1239400 players but no more than 1262000 because 12620 is not enough for top 1% while 12394 was. So the number of players is actually the interval [1239400, 1262000[. This of course says nothing about whether the players are causal, occasional, or whether players that don’t participate in RT are actually not playing at all or just ignoring RT.

Basically it’s just a guesstimate.