PDA

View Full Version : Is there a problem with the stats graphs?



TeAm Deception.Net
12-01-2006, 10:52 PM
Check out my person stats page (http://www.seventeenorbust.com/stats/users/user.mhtml?userID=10544)
Why are there those huge spikes in the graphs? I see others on Team AnandTech with similar spikes but don't see them on other teams that i've looked at.

Also, i logged into my account. There are 4 tests in my pending tests with an IP address i don't recognize. What's up with that?
Another person on our team also has unknown IP addresses in his pending test area.


746594 33661•2^12518064+1 67.151.45.xxx Wed Nov 29 23:06:43 2006 Thu Nov 30 04:03:31 2006 5320920 42 % 46821540

747081 55459•2^11979778+1 67.151.45.xxx Sat Dec 2 00:52:51 2006 Sat Dec 2 01:30:38 2006 10169181 84 % 643771504

747080 22699•2^12508534+1 67.151.45.xxx Sat Dec 2 00:52:51 2006 Sat Dec 2 01:30:52 2006 1742327 13 % 119513371

747082 55459•2^12287038+1 67.151.45.xxx Sat Dec 2 00:52:52 2006 Sat Dec 2 01:30:52 2006 8976720 73 % 594399122

allio
12-01-2006, 11:10 PM
these are the mysterious pending tests. what's going on?



746596 33661•2^6892344+1 67.151.45.xxx Wed Nov 29 23:09:00 2006 Thu Nov 30 04:02:36 2006 1604910 23 % 4327900
747083 24737•2^11971591+1 67.151.45.xxx Sat Dec 2 00:53:03 2006 Sat Dec 2 01:30:55 2006 8221216 68 % 518600352
747078 21181•2^12549212+1 67.151.45.xxx Sat Dec 2 00:52:39 2006 Sat Dec 2 01:31:02 2006 898242 7 % 61423360
747077 33661•2^12518064+1 67.151.45.xxx Sat Dec 2 00:52:31 2006 Sat Dec 2 01:31:06 2006 5866410 46 % 397096328

if someone's crunching for me with 1000M of power from 4 cpus, I'm not complaining, but I suspect there's more to it than that.

Alien88
12-02-2006, 12:35 AM
Hi,

It appears as if a member from team Anandtech has returned results for various users on team anandtech. What he has done is reported the results as different users from a test he's already partially crunched. The server sees this as "unexpected" but still scores it. It will eventually time out and the scores will reflect this, but in the mean time it gets an unexpected result for a test that's been partially crunched and it sees this as "hey they went from 0% to 45% in X time" and that results in a very high cEM/s.

I'm not sure if the person is doing this on purpose or by accident, but either way it's not appreciated and makes the stats look misleading.

I could ban the IP address for the time being, but I'd prefer if someone could bring this to team Anandtech's attention and hopefully that user will stop..

Mike Garrison

TeAm Deception.Net
12-02-2006, 08:42 AM
Thanks for the info. We'll try to get thinks figured out.