Chimp Challenge 2012 rules discussion thread

[chrisk]

This is what I posted in the captain's forum, from what I have gathered in this thread. I have not been monitoring other threads in other forums, for various reasons:

The main thing I can bring from our team is that we just want a Chimp Challenge. We will take part no matter what the system is.

Straight-up PPD, total points, total amount of users with the vowel 'e' etc. we are in.

If we can make people happy with a mix of last year's PPD with the best 2 weeks (including HPCS) then we will accept it. Might be a good compromise. Nobody is going to get everything they want out of this I think, so there must be give and take.

Personally, I am for any system that encourages a huge push in F@H recruitment. The best way I see teams doing a 'frantic' recruitment drive is if each team has a realistic chance of winning from the beginning.

That means that EVGA needs to have a good chance of winning as OCN, HWC, TPU, etc.

[Xavier Zepherious]

from what I gather OCN won't be in if we include the HPCS months in the handicap

the longer this goes on (delays - inactivity on the CC threads)... the more inclined to go with OCAU and have multiple categories multiple badges (for 1 year)
because there is no way we can agree on stuff with little time remaining (the way it's currently going)


then starting immediately after the challenge work for a full year hashing out the new rules/ format/formula... and no whining about it lets get er done... even cross posting the info on each teams forums...public - so they all know what is going on

Don't be surprise if [H] members are folding along with us - some indicated they will (not necessrily for EVGA either)

(although some joined us for the TZC last year)

PS: EVGA is not against any handicap - just against one that solely designed and biased by excluding certain months so that a certain team stands a 100% chance to win

It would also be a good idea to read the CC thread in the EVGA forums - plenty of ideas there

[SugarJ]

Our captains speak for us. While there's some ideas tossed around here in the hopes of making the CC competitive for all, most of us here remember that the whole point is that it's for the science. Whether EVGA or [H] is in or not, we'll be torturing the shit out of our machines before, during and after the CC. "FOR THE SCIENCE!"

[Adak]

Howdy HWC!

I'm with OC.com, and wanted to make some comments on the CC.

1) You guys did a GREAT race last year. Please pay no attention to those who only love sour grapes.

2) Your website for the CC is the best we've ever had, and greatly appreciated.

3) The race needs to be short if we're going to require teammates to fold under another name. A long race would need to be a different race, and might be set up for whole teams, with no change of folding name required. Basically matching up every point from team N with team M, on a handicap basis.

The CC could be set up to avoid folding name changes, quite easily.

4) For the handicap "basis", the easy way is to simply take the points earned during last years CC, as this year's basis. That has a real leveling effect, encourages long term gains in folding power, and stops all the whining about which weeks or months from the year to include in the handicap basis figures. (well, most of it ).

We WANT to encourage racing teams to avail themselves of EVERY opportunity to fold more, during the CC. That includes cloud servers, battery powered abacuses, etc. If one team has an advantage one year with a particular cloud server - good for them! That's smart racing, not cheating.

5) A starting set time is good. The more timing, the more we get to laugh at teams that can't figure out the offset for the Pacific Daylight Saving Time (remember TSC! and Bit-Tech last year - ).

You time yours, I'll time mine - there's no way to stop timing without prohibitively invasive "rules", that we know can't be enforced equitably. So let the timing be a part of the race. We WILL melt the FAH servers, but that's Stanford's problem! <LOL>

Fewer rules, simpler rules, are the best rules. Everyone understands them, and it allows and encourages maximum folding power to be brought to the CC. It ensures maximum fun for the racers, and maximum buzz for Folding@Home, on the social networks, blogs, imo.

That is the three goals of the race: 1) Fun for the folders 2) Increased buzz/exposure for FAH, and 3) Greater folding power for FAH, even if it's only short lived, FAH definitely gets a boost during the CC.

Thanks for all you've done for the CC, and good luck in the upcoming race.

[stoanee]

Originally Posted by Adak

Howdy HWC!

I'm with OC.com, and wanted to make some comments on the CC.

1) You guys did a GREAT race last year. Please pay no attention to those who only love sour grapes.

2) Your website for the CC is the best we've ever had, and greatly appreciated.

3) The race needs to be short if we're going to require teammates to fold under another name. A long race would need to be a different race, and might be set up for whole teams, with no change of folding name required. Basically matching up every point from team N with team M, on a handicap basis.

The CC could be set up to avoid folding name changes, quite easily.

4) For the handicap "basis", the easy way is to simply take the points earned during last years CC, as this year's basis. That has a real leveling effect, encourages long term gains in folding power, and stops all the whining about which weeks or months from the year to include in the handicap basis figures. (well, most of it ).

We WANT to encourage racing teams to avail themselves of EVERY opportunity to fold more, during the CC. That includes cloud servers, battery powered abacuses, etc. If one team has an advantage one year with a particular cloud server - good for them! That's smart racing, not cheating.

5) A starting set time is good. The more timing, the more we get to laugh at teams that can't figure out the offset for the Pacific Daylight Saving Time (remember TSC! and Bit-Tech last year - ).

You time yours, I'll time mine - there's no way to stop timing without prohibitively invasive "rules", that we know can't be enforced equitably. So let the timing be a part of the race. We WILL melt the FAH servers, but that's Stanford's problem! <LOL>

Fewer rules, simpler rules, are the best rules. Everyone understands them, and it allows and encourages maximum folding power to be brought to the CC. It ensures maximum fun for the racers, and maximum buzz for Folding@Home, on the social networks, blogs, imo.

That is the three goals of the race: 1) Fun for the folders 2) Increased buzz/exposure for FAH, and 3) Greater folding power for FAH, even if it's only short lived, FAH definitely gets a boost during the CC.

Thanks for all you've done for the CC, and good luck in the upcoming race.

Great post and comments, thanks alot for that, I agree with all your comments. As you may have noticed we take our folding VERY seriously but still like to have FUN!

[Tim_H]

I've been almost thinking that the chimp challenge wasn't going to happen this year. I'm game for whatever is decided. I just want to know when to make the switch, I think I have some spare parts that I might try and find a home for yet too. Nothing major but work is still work!!

[BrainEater]

Oh it will happen forsure.Even if I fold under "chimp challenge" , by myself for a week....haha.....I know everyone here feels that way.

The last couple months has been really good for the chimp challenge.There's been a lot of really good open discussion by a lot of teams , I've seen a lot of good idea's.

I look forward to the future of the chimp challenge , and folding in general.

[Xavier Zepherious]

EVGA set a deadline this week

either select an all out PPD race or, Split category or figure out a formula
at least get something otherwise EVGA won't be there

[Adak]

If the CC isn't a handicap race, it loses interest with a lot of teams. Having brackets is interesting, also. We could have two brackets - unlimited for the huge teams like EVGA, OCNET, and any other team that wanted to be in that bracket, and a handicap bracket for those not in the unlimited bracket.

This is what I'd suggest for the handicap formula:

Use last years CC as your guide. Take all the total points earned (not handicap points, but raw FAH points), and divide it by 10 (the number of days). That is the "basis" for your team's handicap. Each team gets one point each time they earn that number of points.

The result would be if two teams folded exactly the same as they did last year, they would both wind up with exactly the same number of handicap race points. The problem is, this favors small teams, since they have far fewer points to get to improve their performance, by percentage. (a 10% improvement is far easier for a smaller team to achieve, than a larger team).

So we add in a percentage improvement penalty. If your team has improved by 10% over last year, you will get a 5% penalty assessed to your points. If you improve only 5%, you have a 2.5% penalty only. 15% improvement, means a 7.5% penalty assessment to your points. "Points" here, means raw points, not race points.

You can call the penalty several names to make it more politically correct, (like a multiplier), but it's really a penalty - and it penalizes smaller teams that are favored by the first part of the handicap, more than it penalizes the larger teams.

Somebody with a spreadsheet and last years data could confirm the above assertion would make for a much closer race.

[Xavier Zepherious]

actually if you just made the handicap the best two weeks of production by any team

then (CC PPD/ Handicap PPD) would give you a point scheme that was fair (all by itself)

but multiplying that by conversion is not (cc ppd/handicap PPD) *(CC PPD/ TEAMS PPD) is just increasing the disparity (Zodac formula)
then zodac wanting to take best months away from handicap to favor themselves didn't suit EVGA too nicely either

I even suggested some conversion additive like (cc ppd/H PPD)+ (cc ppd/h ppd)x(cc ppd/team ppd )/ range (where range is 10 or 20 to give a 2.5%-5% bonus on getting people out)

Ive suggested different formula and formats - we even ask zodac to talk with our contest Guru on a formula(work together) - and all we get is posturing and them forcing their formula only

Honestly we been on the CC forums - suggesting things waiting for teams and ideas and agreements to take place
It's almost getting too late for EVGA to get onboard

had the teams been really interested they would be on the CC forums religiously suggesting ideas, discussing things rather than the mess we are in
if the CC is to succeed the teams will have to plan and start earlier and plug in(the CC forums on a daily basis) - or the CC is over


we given this week as a deadline...if we can't get it together we plan starting NOW for next year

judging by the mess it's in this year.... I don't hold all hope any more