Since Dr Anderson's statement on cross project credit parity, I think that we are beginning to enter the realms of politics, hence this thread here.

Is Dr Anderson correct?

---

Since Dr Anderson's statement on cross project credit parity, I think that we are beginning to enter the realms of politics, hence this thread here.

Is Dr Anderson correct?

Umm, I missed the original statement, do you have a link?

---

I think you'll find it's a bit more complicated than that ...

Posted by DA on boinc_projects mail list:

There is significant variation in average credit
per CPU second between projects: see
http://boinc.netsoft-online.com/e107_plugins/boinc/get_cpcs.php


The following projects are giving more than 1.5X S@h's credit/sec:
QMC (1.59)
RieselSieve (4.3)
Cosmology@home (1.68)


... and a number of projects are giving somewhat less (in the .7 range).


It is critical to the success of volunteer computing that participation
be based on factors like scientific merit - not points per CPU sec.
So can all projects please take steps to make their ratio
as close to 1 as possible (and no more than 1.2).
If not, I will eventually ask the statistics sites to scale down
the credit of projects with high ratios.


To adjust your ratio:
1) if you're giving fixed credit per WU,
or using the FLOPs-reporting API, change your numbers.


2) if you're giving credit based on benchmarks*CPU time (the default)
set the <fp_benchmark_weight> configuration parameter
to a value that's appropriate for your application: see
http://boinc.berkeley.edu/trac/wiki/ProjectOptions
If you're not sure how to do this, or it doesn't fix things,
please reply to this group.


-- David

and currently being discussed in Number Crunching in this thread.

---

It is critical to the success of volunteer computing that participation
be based on factors like scientific merit - not points per CPU sec.

Who gets to decide the scientific merit? The National Science Foundation? The BOINC admins? What about people participating in whatever project they darn well please, regardless of credit? I submit to you, fellow BOINCers, that number three will have the most influence.

---

Capitalize on this good fortune, one word can bring you round ... changes.

It is critical to the success of volunteer computing that participation
be based on factors like scientific merit - not points per CPU sec.

Who gets to decide the scientific merit? The National Science Foundation? The BOINC admins? What about people participating in whatever project they darn well please, regardless of credit? I submit to you, fellow BOINCers, that number three will have the most influence.

I quite agree, but if the Director states otherwise, I can see changes ahead.

My personal intentions are the 1st 25 customer agreements I get, will be seti followed by the next 25 to Einstein, etc etc.

My personal opinion is that allied forces went into Iraq to depose a dictator, so why should volunteers to scientific projects accept it?

Edit - also, as seti has nothing but problems, why should that project be allowed to set the standards for others - they cannot maintain their own!!!

Edit 2 - As Dr A states there are another 41 projects besides S@H!!!

Edit 3 - Just had a very instinctive feeling that I'm heading for a ban!

---

I don't see any problem with the way different sites allot credits.

If everyone in a given site receives credits at the same rate, everything is relative whether the rate is set at .5 per unit or 2.5 per unit.

I don't think people join these projects to gather useless online points in the first place.

While the accumulation of credits does allow one to check their progress against others, as long as they all receive these credits at the same rate, what's the problem?

While the accumulation of credits does allow one to check their progress against others, as long as they all receive these credits at the same rate, what's the problem?

Since you appear to be a SETI-only person, you may not realize that payouts are lower or higher at other projects. You should look at this chart.

Since I'm "all up in it" already, here's my deal:

The idea to normalize is a perfectly legitimate idea. The general idea behind it is that the selection criteria for projects could be influenced by the credit payout, with those that offer very high payouts attracting more participants and those with lower payouts attracting fewer participants. The general idea is to have all projects offering the same general amount of credit, so then the selection of a project by a potential participant would be more solidly based upon the participant's enthusiasm over the project's goals, aka "the science".

This is a laudable goal, but it has some issues:

• Assumes many participants already do not select a project based on the science.
  
• Assumes that competition exists between project administrators to attract large numbers of volunteers by offering very high credit.
  
• In its' current form, disregards the fact that some of the higher credit projects are in alpha or beta phases. Alpha or beta projects should, at least in my opinion, offer more credit to users because there could be problems with the applications crashing or validation problems resulting in the user not getting any credit at all for their time.
  
• Also in regards to alpha/beta projects, some users may experience problems, while others do not. The statistics for the project could then be weighted more in favor of those who did not experience problems, thus due to the higher total credit granted to those participants, the average could be higher for them than for someone who did have problems. This begs the question, is it fair to then penalize the person having problems in an alpha/beta project by recalibrating the entire project's credit so as to bring the people that are not experiencing problems "back in line" with production-level projects?
  

Anyway, the "politics" angle of this is whether or not projects should be coerced to comply. The initial way of handling has been a request. That is proper. A statement has been made that appears to indicate that if compliance is not done willingly, action will be requested to be taken to make it appear that a non-compliant project was actually compliant so as to create an illusion of more of a "level playing field" so that the selection of a project by a potential participant would in theory be more about "the science" and less about "the credit".

I am seeking clarification about the whole illusion bit, and whether or not that is really what is meant, but I can tell you that if it is, I have some serious issues with that concept.

---

Well, I don't really see a big scandal there, as it's just statistics. And as BOINC is the platform, the other projects are using, it's fair, that they should adapt to reasonable rules given by the project management.

Of course you can debate if it really is the wisest idea to force them to comply. On the other side you can also ask, if it's fair of some projects to give out higher credits, to manipulate the people on their project's choice. (Assuming this is the reason for the differences in the credit-policy)

Edit 3 - Just had a very instinctive feeling that I'm heading for a ban!

Now THAT would really be a scandal. (But certainly won't be happening)

---

Best regards
Jens

Credits should be based on the flops in a work unit. Then a super fast computer will not have an advantage per work unit over a slow computer. The speed will give the fast computer the possiblity to complete more WUs.

Credit should not be given based on time spend on a work unit then the fast computer will have a enormous advantage over an older slow computer.

You volunteered to work here and if you don't like it simply leave rather than whine over credits. geesh

---

my website... Our Famlies Guitar strap company

You volunteered to work here and if you don't like it simply leave rather than whine over credits. geesh

It is correct that we have volunteered. It is not appropriate to characterize this as "work." We ourselves do not do work. We have set our machines to do work for the various projects that interest us. The poster's characterization of the discussion that has gone on in this thread so far as "whining" is odd. For someone that has only crunched for the project for 3-4 months to tell people to go away is even stranger. Where (from whom) have we heard such things before?

---

Capitalize on this good fortune, one word can bring you round ... changes.

In answer to one of Brian's observation's, or is a question.
He noted that several site are not yet mature, that they are still in a testing phase. Therefore maybe the question should test sites be in the main stream stats analysis at all?
The question then is would the main sites be willing to make steps to cross project equalistaion, if they don't already comply to the +/-20% ruling.
With the follow on question, would enough volunteers join test projects if the credits don't contribute to the main BOINC statistics.
Or more truthfully if test sites operated 'correctly' in that they only granted credit on completed, validated results, which in many cases would be below the expected rate because something unforeseen caused the test application to run slower than expected (see Einstein S5R3 Windows 4.25 app), and gave zero credits for all canceled, incomplete work.

We ourselves do not do work. We have set our machines to do work

Oh sure... Now ya tell me, three calculators later... ;)

---

It may not be 1984 but George Orwell sure did see the future . . .

The question then is would the main sites be willing to make steps to cross project equalistaion, if they don't already comply to the +/-20% ruling.
With the follow on question, would enough volunteers join test projects if the credits don't contribute to the main BOINC statistics.

There are other distributed computing projects that are not using BOINC. Some of my teammates run them. I understand Folding@Home is one such project? So, apparently, there are volunteers joining projects not at all concerned about BOINC credits.

---

Capitalize on this good fortune, one word can bring you round ... changes.

We ourselves do not do work. We have set our machines to do work

Oh sure... Now ya tell me, three calculators later... ;)

LMAO

---

Air Cold, the blade stops;
from silent stone,
Death is preordained

Calm Chaos Forums : Everyone Welcome

There are other distributed computing projects that are not using BOINC. Some of my teammates run them. I understand Folding@Home is one such project? So, apparently, there are volunteers joining projects not at all concerned about BOINC credits.

Very well said. People only have to look at my sig to see that I don't really give a monkey's regarding credits. Had I done so, I would have left all rigs crunching away for little green men & my rac would probably be over 2000 by now (it was over 1500 before selecting other projects).

Personally, I can work as an individual or team player. If I am given the choice, I would rather be a team member as you can have more fun. I apply this fact to the projects & where rac is concerned, I'm only interested in helping my team. A little competition never hurts anyone.

I'm just saddened that the foremost project is underfunded & undermanned. However, should it's leadership become totally isolated in their ivory towers, then it deserves to die a weary death.

---

I'm just saddened that the foremost project is underfunded & undermanned. However, should it's leadership become totally isolated in their ivory towers, then it deserves to die a weary death.

Well, careful there.
There may be truth to the underfunded/undermanned concern. So, there's always the donation route to go for the first. For the second, I do not know what to tell you. People have talked about whether the WUs we've returned get analyzed. How many people in the world could get together to both determine how to analyze and how that should get coded?
Not me. My intelligences/skills lie elsewhere, even though growing up 30 miles from Cornell (where Carl Sagan taught) made my first thoughts of going into academia be about astronomy.
As for the second, please recall that there is both the leadership of this flagship project and then there's leadership of the overall BOINC efforts. They're neither the same nor mutually exclusive and the people in the intersection do not have to be in agreement on all things.

---

Capitalize on this good fortune, one word can bring you round ... changes.

That's what I meant about blind faith & pandora's box. Dr Anderson created Boinc & now that crunchers are leaving the foremost project & joining projects where progress is made, it does not sit well with them.

So much so, that the other projects have to set their standards by seti's!

That's like telling GM, Chryslar, & the rest to comply with Ford's standards because Ford's were the 1st! Or telling the US Navy along with other navies that they have to have the same standards as the Royal Navy just because they were the 1st!

I don't think any will comply nicely over a cup of tea, do you?

---

It's an interesting analogy. We'd have to think about whether the two situations are really comparable.

---

Capitalize on this good fortune, one word can bring you round ... changes.

It's an interesting analogy. We'd have to think about whether the two situations are really comparable.

Okay, how about this one - I'm the MD of a well known international corporation & make a similar public statement. It drives away customers. The shareholders contact the board of directors - I'm fired.

True case in point - Gerald Ratner & Crap Jewelery - Bye Bye Gerald Ratner!

---