Ananas:
I was graphing the new credits as credit/second like the pic above and then reporting the wildly unusual new credit grants. When graphing granted credit along with the claimed credit, it was also possible to see the effect of benchmarking/rebooting and unloading unneccessary software. The value I quoted is my subjective call on how much higher (or lower in one case) than the eyeballed average for that area of the chart.
In the chart above, only 2 of the WUs were reported.

---

I just processed two t354__CASP7_MINHELIX WU. One was reported back 2 Sep 2006 19:49:46 UTC and got 2.88 credits/decoy, the other was reported back 3 Sep 2006 4:55:42 UTC and got 2.45 credits/decoy. I find this a surprisingly high deviation for a WU, which is older than 24 hours.

It appears to me that there is a trend for all WU, that they get less and less credit the "older" they are. This could be explained by the fact, that 5.5.0 reports results back immediately, after they are finished while the stock client reports them back later. So the first reported results are probably most from 5.5.0 clients and are corrected the more standard clients report back. Can anyone confirm my hypothesis with his results?

I just processed two t354__CASP7_MINHELIX WU. [snip] I find this a surprisingly high deviation for a WU, which is older than 24 hours.

It appears to me that there is a trend for all WU, that they get less and less credit the "older" they are. This could be explained by the fact, that 5.5.0 reports results back immediately, after they are finished while the stock client reports them back later. So the first reported results are probably most from 5.5.0 clients and are corrected the more standard clients report back. Can anyone confirm my hypothesis with his results?

Perhaps you were part of the "bonus group"?? The last, 3rd from last, and 5th from last wus on this chart are from the same type wu.

156 points for a day's work is low on my 2.8 GHz Intel P4: https://boinc.bakerlab.org/rosetta/workunit.php?wuid=30835037

The Result: https://boinc.bakerlab.org/rosetta/result.php?resultid=35426560