I also observed some strange effects some days ago (suddenly I had more than 1600 total reads and about 80 last week). But in the Diagram there were the 'normal' (and much lower) numbers. Now they seem to have fixed it and I have only 10 total reads again :)
Aaaah, I solved the puzzle. The reason for the difference in read numbers is the simple fact that the bigger number aggregates publication reads and project updates as well as questions. while, the smaller number only contains publication reads. It is as always a good idea to read carefully :)
there are many strange things-- people having better citations but no questions and answers have lesser RG score whereas those having 3000 answers have better score
on other hand those having 2000 answers have lesser score in comparison to those having 3000 answers.
the criteria for scores against any quastion or answer and against any follower or citation better to be known to members.
some members complained that they added a paper and actively participated in answers but at end of week their score was reduced (0,4-0,8) instead of increasing.