I don't understand....
3 Mar 2005 10:53:17 UTC
Topic 188219
(moderation:
I don't understand the stats - though perhaps this is because I am a former classic seti user.
How do I find out my "rank" compared to other users?
Why does the most important statistic appear to be the recent credit, surely this will, on average be the same over time unless you add more and more machines to your userid?
Am I just too interested in leaderboards and rankings over the science?
I don't understand....
)
> Am I just too interested in leaderboards and rankings over the science?
Put it this way... I currently get around 200,000 hits a day on my BOINC WORLD STATS site... so I figure your not TOO interested compared with the rest of us *smile*
The BOINC Projects intentionally have minimal stats on their sites for a reason. They produce XML files which 3rd party stats sites then can be made around (like mine... link below). That way a LARGE proportion of hits and bandwidth can effectively be transferred off site... allowing smoother running of their own web servers (smart hey)
Anyway... here is a link you may be interested in
http://www.boincsynergy.com/stats/boinc-stats.php?id=3717&project=eah
That is of course on My stats site. Follow the links around. The entry point to the site is here http://www.boincsynergy.com/stats/
There are of course other stats sites run by Willy, Toby, Janus and a few others... which can be found here http://einstein.phys.uwm.edu/stats.php
Enjoy
Zain
---
Teamless? Drop by BOINC Synergy and look around
Also checkout our famous BOINC World Stats
Well thanks very much for
)
Well thanks very much for that - it's clarified it alot for me. Logical for the project to do it this way - after all they are interested in the science rather than providing a stats service :-)
Still not sure why they seem so interested in recent average credit though. Perhaps it's so they can see users who stop sending stuff in.
===
Read my book blog
http://resolutereader.blogspot.com
> Why does the most important
)
> Why does the most important statistic appear to be the recent credit, surely
> this will, on average be the same over time unless you add more and more
> machines to your userid?
I've often wondered about that too ;) but don't worry, the stats sites have rankings for total credit - and both for individuals and teams, just like seti classic ;)
> Well thanks very much for
)
> Well thanks very much for that - it's clarified it alot for me. Logical for
> the project to do it this way - after all they are interested in the science
> rather than providing a stats service :-)
>
> Still not sure why they seem so interested in recent average credit though.
> Perhaps it's so they can see users who stop sending stuff in.
It can be used to determain how many TerraFLOPs are produced - from the BOINC website:
"BOINC-based projects projects now have 80,721 participants in 188 countries. Based on current credit statistics, together they supply 106 TeraFLOPS of computing power - far more than any conventional supercomputer"
The relationship between FLOPs and RAC is explained here: http://boinc.berkeley.edu/flops.php
You can also pass by on
)
You can also pass by on Mundayweb.com for signatures and stats.
If we're listing them anyway. ;)
Since I don't want to create
)
Since I don't want to create new thread, and I don't understand...
I tried to read abour Recent average credit (RAC) - and don't understand.
I have two machines, which was joined same time. First of machines computed 390 credits so far, the second 379. However first have 218 RAC, and the second 118.
Why?
Moreover, I get this suspicion that's only becouse first machine was the first which gained any credits.
And moreover, I have machine which was added to project later than those machines, has 419 credits so far and only 78 RAC. Why?
> And moreover, I have
)
> And moreover, I have machine which was added to project later than those
> machines, has 419 credits so far and only 78 RAC. Why?
Did you read the definition for RAC in the glossary?
In fact I did it before
)
In fact I did it before posting.
And I still don't understand how it answers my Q (maybe simply language problem?). My machines runs all the time. Were added same time, were granted same credits, have different RAC.
Maybe I simply don't understand what is RAC. I expeted that machine which gained more credits within shorter time, will have larger RAC. Then it will be higher in hall of fame ;)
But it's not.
> That is of course on My
)
> That is of course on My stats site. Follow the links around. The entry point
> to the site is here http://www.boincsynergy.com/stats/
Nicely done, I like the layout quite a bit. My compliments. =)
http://www.wolverinesden.org
> In fact I did it before
)
> In fact I did it before posting.
> And I still don't understand how it answers my Q (maybe simply language
> problem?). My machines runs all the time. Were added same time, were granted
> same credits, have different RAC.
Well, with your computers hidden we can't help much ...
If you want detailed answers you have to open up your machines for examination ...
However, without that, lets try this ...
RAC is a running average. It is done per computer. So, both of your machines would have to run identical WU with identical run times, with identical credit claims with identical credit grants created from identical quorums.
Too many identicals.
The bottom line is that even if you ONLY run Einstein@Home on both machines, never use them for anything else, the fact that the credit grants cannot happen at the same time will cause the RAC to be different. It is only calculated at the time of the grants, so, because you have to wait for the quorum to form, and they will form at different times for the two machines ...
Simply put, they are never going to be the same, close maybe, but never the same. Think of them as a rough speed indicator of the rate at which each computer is accumulating credit ...