Folding Forum

Since my previous "crazy idea" has had more holes shot in it than a slice of swiss cheese, I thought I would try another crazy idea. This one does have something in common with my previous crazy ideas - it requires a nice chunk of money to be donated by somebody who can afford the donation ...

This idea is actually a combination of two of my previous ideas - namely the FAH "Business Card/Promotional Postcard" idea combined with an annual "FAH Donor Recruiting Contest" award. My brother joined a national sales organization a few years ago. Each year this organization has a "Rookie of the Year" contest where the top rookie FSR (Field Service Representative) wins a fairly substantial cash award along with a number of prizes and other recognitions. I'm proud to say that my little brother won the ROY award a few years ago. So it occurs to me, why can't we have an annual "FAH New Donor Recruiting Award" (along with say a $100,000.00 prize) which would go to the member of the FAH community who recruits (and brings in) the most new donors to the FAH project during the calendar year?

To qualify for this prize, there must be certain conditions (and certain qualifications) that a new donor must meet before that "new donor" can be included in a contestants total. An obvious condition would be that in order to be counted, a new donor must have completed (and returned) some minimum number of Work Units during the year in which they are recruited. I'm not sure what such a number should be, but I'm thinking a minimum of 50 completed WUs would be required before a new donor could be added to a contestants total. (Fifty is an arbitrary number - others might feel that the new donor should complete a minimum of 100 WUs before they are credited to a recruiters total.) The goal of establishing a minimum number of completed WUs is to try and encourage new donors to continue folding as some may fold for a week, maybe do 25 or 30 WUs, and quit. The goal here is twofold: Not just to bring in new donors, but to (hopefully) keep new donors folding. There are other conditions (and "rules") for this contest that I haven't thought of, but I'm sure I can depend on all of you folks to point out the things I've missed. (The most obvious "problem" with this idea is: "OK Alan, who's going to put up the annual $100,000.00 prize for the winner of this contest? You!?" That's a good question ... I don't know the answer.)

Assuming we can solve the "big problem" of where to get that 100K prize money that makes this "contest" possible, one way a contestant might go about getting credit for recruiting a new donor would be having a special "code number" printed on your "FAH Business Card/Post Card" and asking your new donor prospect to enter that number when they are downloading the software and registering. (This would require a bit of coordination with Bruce and the Pande Group I suppose.) If we don't use some kind of "code number" scheme to tag new donors to their recruiter, I suppose each donor will have to get the name and email address of the individuals they recruit. The key here is that if you have a pre-printed FAH card, (such as what I have suggested here: viewtopic.php?f=86&t=23382), that will make your recruiting efforts a bit easier. (Heck, I've recruited several new donors via casual conversations at my local Barnes & Noble - without the benefit of a FAH Business Card - so it can't be that hard!)

Coming up with the money to fund the prize is the major problem with this idea, but a legitimate shot at a $100,000.00 prize is a lot of motivation. If we only have 100 dedicated (and creative) folks vying for such a prize, somebody will get a pretty good return on the investment of their time - and Dr. Pande and his team will be a lot closer to those 100,000 new folders that he would like to see. (And hopefully the team will be able to take simulation of the Alpha synuclein protein out of the "experimental project" catagory and get it on the front burner!) I would like to see Folding@Home go from the largest distributed computing project in the world to the largest distributed computing project in the universe. If we currently have a sustained performance rating in the neighborhood of 3-5 petaFLOPs, I want to see that rate increase by a factor of ten - at a minimum. Bring in enough new donors and we should be able to blow these numbers out of the water!

My brother won a nationwide "Rookie of the Year" contest and he was only up against 40-45 other FSRs. It appears to me that there are approximately ten (maybe twenty) folders who read and post to these forums. I wish I were wealthy as I would put up the $100K (and "suggest" some of the contest rules) myself. However, even if we can't secure a generous benefactor who would provide the seed money for this prize fund, maybe we can institute such a "contest" anyway - just for the hell of it. This is a challenge where you don't have to devise a "fair" way of comparing uniprocessor x86 FLOPs to quad-core processor FLOPs to GPU FLOPs - all you have to do to win this contest is get out there and hustle. Use your gift of gab and talk to people!

Alan C. Lawhon wrote: It appears to me that there are approximately ten (maybe twenty) folders who read and post to these forums.

I think there are a few more people around than that... At one time there are the following stats:

The board software wrote:In total there are 26 users online :: 6 registered, 0 hidden and 20 guests (based on users active over the past 5 minutes)
Most users ever online was 175 on Sat Oct 27, 2012 7:59 am

Alan C. Lawhon wrote:Since my previous "crazy idea" has had more holes shot in it than a slice of swiss cheese, I thought I would try another crazy idea. This one does have something in common with my previous crazy ideas - it requires a nice chunk of money to be donated by somebody who can afford the donation ...

Hi Alan, I don't think anyone disliked or "shot holes" in your other ideas, everyone was just thinking up different ways of how it could work.

For this new "crazy idea", how about if each recruiter had a different team number? That way, it would be very easy to compare the number of new folders that each recruiter had got to join, as well as how many new folders had "stayed with it" for the duration.

BTW, I think everyone here really appreciates your energy and new ideas, so don't stop coming up with them!

Art

Shooting holes in a "crazy idea" doesn't mean it was a bad idea. At the very least it gets people thinking and the ensuing discussion is always valuable, whether the idea is adopted or not.

I don't think there's a reliable way to ascertain who is a Rookie and who is not. In your brother's case, the employer could see the SSN and the person and know who was new and who was not. In fah's case, I could move all my machines to a new name, a new passkey, and a new team (or maybe some subset of that list) and the only change is that 10 WUs don't get QRB and I have to mentally combine my stats from all of my (former) accounts.

Alan C. Lawhon wrote:I'm not sure what such a number should be, but I'm thinking a minimum of 50 completed WUs would be required before a new donor could be added to a contestants total.

Using number of wu's isn't a good idea, for one thing AFAIK Nvida-GPU will easily finish many wu's per day while with SMP it's maybe only 1 wu/day. For another, if SMP and single-core both uses roughly 1 day/wu, to meet the requirement it's just to run N single-core instead than SMP to finish 4 - 8 wu/day (depending on #cores) and wouldn't expect this would be an advanage for FAH. Also, where's large variations between wu's for the same type and ocassionally there'll be a string of same-project wu's, meaning one donor can be lucky and get a string of 4-hour/wu so will hit 50 wu's in 9 days and stop folding while another donor running FAH a month later gets nothing but 24-hour/wu so needs 50 days.

Rattledagger wrote:

Alan C. Lawhon wrote:I'm not sure what such a number should be, but I'm thinking a minimum of 50 completed WUs would be required before a new donor could be added to a contestants total.

Using number of wu's isn't a good idea, for one thing AFAIK Nvida-GPU will easily finish many wu's per day while with SMP it's maybe only 1 wu/day. For another, if SMP and single-core both uses roughly 1 day/wu, to meet the requirement it's just to run N single-core instead than SMP to finish 4 - 8 wu/day (depending on #cores) and wouldn't expect this would be an advanage for FAH. Also, where's large variations between wu's for the same type and ocassionally there'll be a string of same-project wu's, meaning one donor can be lucky and get a string of 4-hour/wu so will hit 50 wu's in 9 days and stop folding while another donor running FAH a month later gets nothing but 24-hour/wu so needs 50 days.

Mr. Rattledagger (what a cool moniker - is there any chance you are a herpetologist?)

If setting an arbitrary number of Work Units returned (or points earned) is problematic, I suppose we could consider some other minimum performance metric. How about a minimum time threshold that the new donor must meet and/or exceed before they are officially added to the recruiters count? For instance, there are 24 hours in a day and 30 (or 31) days in a month, so the product of 24 and 30 is 720 hours. (Maybe we could round that up to a nice even 1,000 hours.) If we adopted this minimum performance metric, a recruiter would not get credit for bringing in a new donor unless (and until) that new donor has crunched (and returned) completed WUs for a cumulative total of 720 (or 1,000) hours. Adding this rule might (or might not) encourage long term donor participation, but it would get over the technical problems you have noted.

I don't know of a practical way to measure that 720 (or 1000) hours. While that number is reported when the WU is uploaded, I don't think we can obtain a total from the stats database.