Can someone please investigate this credit result for this task? The calculation looks off to me.

https://boinc.bakerlab.org/rosetta/result.php?resultid=347543422

---

Bad luck, things like this happen. But it happens the other way as well. I was 'overgranted' here i.e.

It eventually will even out.

cu

Jochen

---

Jochen wisely said:

Bad luck, things like this happen.

I agree, the Fickle Finger of Fate and all that good stuff. However, that said, I think that the original poster may have a point. Looking back over the past day or so I too see a handful of tasks that got unexpectedly low granted credit.

I am willing to accept a 50% or so deviation as a result of the the expected "Fickle Finger of Fate Factor" but not when the difference amounts to 90% or more. At that point I just tend to jump to the conclusion that something is just not right with that job.

Note that I said "not right" - I don't think that it is an evil plot of a Windows user to skim credits from the enlightened few, nor is it a big deal.

I offer up as a couple of examples from the past day (not an exhaustive list - my bladder calls)

Task 347680997 - claimed 109.2, granted 1.6
Task 347344569 - claimed 146.1, granted 1.1
Task 347618079 - claimed 123.7, granted 9.8

These all "ran long" but ended cleanly (they were not bit by the Watchdog) and generated very few decoys. In my case they were run on AMD Phenom II CPU's running Linux.

Yes, I had another one yesterday. This time I was granted a whopping 1/100th of the claimed credit.

No offense, but if my machine is going to be doing 3.5 hours of work on a WU that is granted that little... I might as well stop running Rosetta.

https://boinc.bakerlab.org/rosetta/result.php?resultid=347749871

---

No offense, but if my machine is going to be doing 3.5 hours of work on a WU that is granted that little... I might as well stop running Rosetta.

I know that unexpectedly low credit can be both disappointing and frustrating and that everyone has to make decisions based on what they fell is right but I guess in light of the fact that Rosetta credits have no "street value"I want to ask one question:

Do you run Rosetta for the credit or do you participate in it because you believe in the cause and the work that is being done?

I agree, something appears to be wrong with the task, and I have had a few of them too. But I have been writing software for more years than I care to admit and bug-free code is a fantasy. And while that particular work unit may not have provided much direct benefit to the science at hand, exposing bugs when found is also important.

Just my two cents worth. Now I will return to the back of the room where I can be seen but not heard...

Cnick6 - I'll guess I will ask one more thing - you appeared to have generated 12 valid decoys with that work unit - whatever a decoy is. Did you consider that it is very possible that while the scoring appears to be a little screwy, the results generated may very well be useful?

Now I'll shut up. I promise.

Yes, I had another one yesterday. This time I was granted a whopping 1/100th of the claimed credit.

No offense, but if my machine is going to be doing 3.5 hours of work on a WU that is granted that little... I might as well stop running Rosetta.

https://boinc.bakerlab.org/rosetta/result.php?resultid=347749871

I guess these ProteinInterfaceDesign-tasks are related to CASP and are different to the usual tasks provided by Rosetta (AFAIR). I had a couple of those as well, usually running 10 in stead of 6 hours and with rather low credits granted. As I said, bad luck. As well CASP won't last forever...

If you take a look at my other threads, you will find out, I am not happy with the crediting system either. But even I have to admit, that it is a fair approach. And as long, as I can't come up with a better idea, I won't complain anymore.

@Chris: Decoys (models) are these neat little 'subtasks' where the credits are granted for. Some run longer, some run shorter and for these types of decoys there will be a different amount of credits granted (there is a thread explaining it in detail: The new credit system explained). The calculation is based on the average claimed credits...

cu

Jochen

---

Best way to deal with these low credit WUs is to ignore them. It happens to all of us every once in a while and thus does not affect your credit value vs another cruncher's credit value under rosetta. (Just like inflation and deflation).

---