Overclock.net › Forums › Overclockers Care › Overclock.net Folding@Home Team › Chimp Challenge 2012??
New Posts  All Forums:Forum Nav:

Chimp Challenge 2012?? - Page 4

post #31 of 314

This ^ headscratch.gif

 
HDPC
(15 items)
 
 
CPUMotherboardGraphicsGraphics
AMD Ryzen 7 1700x Asus Prime X370-Pro MSI Hawk Ek GTX 1070 MSI Hawk Ek GTX 1070 
RAMHard DriveHard DriveHard Drive
EVGA DDR4 3200 Kingston SSDNow 120GB WD Green 2TB WD Black 2 TB 
Optical DriveCoolingCoolingCooling
Samsung DVDR Ek Supremacy Evo Blue Edition XSPC RX360 Radiator Swiftech MCP 655b 
CoolingCoolingOSMonitor
EK XSPC Dual Bay Reservoir Windows 10 e-Machine 23" 
KeyboardPowerCaseMouse
OCN Ducky DK9008 G2 Keyboard PC Power & Cooling Silencer 750W Antec Lanboy Air Logitech G400s 
Audio
Realtek HD Audio 
CPUMotherboardGraphicsRAM
Intel i7 2600k Asus Sabertooth P67 EVGA GTX 770 G.Skill Ripjaws X 
Hard DriveCoolingCoolingCooling
Kingston SSDNow 120GB XSPC Rasa Acetal XSPC Pump/Res  Ek GTX 770 
OSMonitorKeyboardPower
Windows 10 Pro Vizio 58" 4k Logitec Wireless Coolmax 600w 
CaseMouseAudio
Antec 300 Logitec Wireless Realtor HD Audio  
  hide details  
Reply
 
HDPC
(15 items)
 
 
CPUMotherboardGraphicsGraphics
AMD Ryzen 7 1700x Asus Prime X370-Pro MSI Hawk Ek GTX 1070 MSI Hawk Ek GTX 1070 
RAMHard DriveHard DriveHard Drive
EVGA DDR4 3200 Kingston SSDNow 120GB WD Green 2TB WD Black 2 TB 
Optical DriveCoolingCoolingCooling
Samsung DVDR Ek Supremacy Evo Blue Edition XSPC RX360 Radiator Swiftech MCP 655b 
CoolingCoolingOSMonitor
EK XSPC Dual Bay Reservoir Windows 10 e-Machine 23" 
KeyboardPowerCaseMouse
OCN Ducky DK9008 G2 Keyboard PC Power & Cooling Silencer 750W Antec Lanboy Air Logitech G400s 
Audio
Realtek HD Audio 
CPUMotherboardGraphicsRAM
Intel i7 2600k Asus Sabertooth P67 EVGA GTX 770 G.Skill Ripjaws X 
Hard DriveCoolingCoolingCooling
Kingston SSDNow 120GB XSPC Rasa Acetal XSPC Pump/Res  Ek GTX 770 
OSMonitorKeyboardPower
Windows 10 Pro Vizio 58" 4k Logitec Wireless Coolmax 600w 
CaseMouseAudio
Antec 300 Logitec Wireless Realtor HD Audio  
  hide details  
Reply
post #32 of 314
sorry if my last post was confusing smile.gif
post #33 of 314
Quote:
Originally Posted by zodac View Post

Adak, this was in your earlier post:


If you don't need to change your Folding name, that would imply no team CC name.

Again, while my initial formula had the conversion rate as having a large influence, that could always be lessened; none of the factors were at final weights - we could have tweaked them if necessary, but the entire thing was dimissed out of hand instead.


Now, on ot your basis system. While in theory it all seems fine, since you're basing it off a highly competitive period, you're missing one key factor -  PPD changes over the year, and the amount differs greatly for each team. In the case of OCN, we've got from ~10mil during last year's CC, to 20mil PPD... essentially we've doubled it. However, a team like CPC had ~4.3mil last year... and their total production right now is only 4.2mil. While they could increase their production, would it be comparable to the increase other teams have seen? Taking some of the basis from last year's CC, and some from a more recent period (showing current production) would balance that somewhat.

Secondly, regardless of the handicap/basis used, and when it was taken from, smaller teams have a much simpler job, and will see much faster benefits. A single 250k Folder could make up 10% for a small team, but only 1% for a team like Evga. So both teams are increasing the work done and research being generated, but one team gets x10 the benefit - that isn't a fair system. That's why last year's event didn't work, and why it won't work again this year. Your comment was this:


That was the opinion the captains had last year -  we made a mistake. It would be ridiculous for us to make that mistake again and expect different results.

Finally, I still don't see why we have two categories; we stopped the all-out PPD race because we knew it wasn't competitive anymore. Why are we bringing it back?

1) Each team WILL ABSOLUTELY have a team name. Don't know where you got that idea from, but it's completely off the mark.

2) I did not see your full proposal, or Xaviers. Just a sort description. If you want to discuss it, then let's discuss it. I would need to see some numbers like I posted, using your formula, before I could discuss it in depth.

3) If you make the basis based on folding "during the year", then you destroy the ability to handle teams that have numerous temporary folders, join into the race. That is unacceptable, since it lead to the problem of the CC last year.

4) If team A races with team B, and team B increases it's folding by 10%, then team A needs to increase their folding by 10% to keep up, in a handicap race.
Does that not seem fair?

5) The race last year had a mistake relating to the HWC. This fixes that problem, easily. But just for a data item, HWC has, WITHOUT NCIX HELP, one of the highest PPD average per folder, of any team in Folding @ Home, (minus the odd supercomputer teams like awachs). So in any handicap race, it would be very hard to keep them out of one of the top finishers, don't you think?

Yes, of course.

To put it into perspective, if you look back on other CC's, you would see that last years race was the CLOSEST race we've ever had - EVER! (by percent, among ALL the teams).

Yes, a mistake was made, but it was corrected (maybe not perfectly, but an honest effort), and the race was a great success. Maybe not from everyone's POV, but that's how it goes in a handicap race.

That's why I thought having both a Handicap race bracket AND an Unlimited race bracket, would bring a better balance into the CC. Because handicap races will ALWAYS favor either the little teams (because they're based on percentages), or the biggest teams (because they're based on points alone, even though it's a handicap).

You need both brackets to balance it out.

Would you rather change the unlimited bracket to a handicap bracket based on points, rather than percentages? Then the CC race would be a two bracket race with one based on percentages, and one based on points (but now with a handicap). I ran a projection of ONE type of a point based handicap last night on the spreadsheet. EVGA just ran away with it, and left us all in the dust.

But maybe.... I'll try and find a better format, but time is getting short, and it can't be too complex, or it won't sell to the CC captains, imo.
post #34 of 314
I ran a projection of ONE type of a point based handicap last night on the spreadsheet. EVGA just ran away with it, and left us all in the dust.

Was that based on last years CC, or using an average PPD since then? Maybe both?
    
CPUMotherboardGraphicsGraphics
Intel i7 2600k ASUS P8Z77WS EVGA GTX 980 SC ACX 2.0 EVGA GTX 980 ACX 2.0 
GraphicsRAMPower
EVGA GTX 970 SC ACX 2.0 Mushkin Enhanced Blackline 8GB (2 x 4GB) 240-Pi... FirePower ZX Series 1250W 80Plus Gold 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel i7 2600k ASUS P8Z77WS EVGA GTX 980 SC ACX 2.0 EVGA GTX 980 ACX 2.0 
GraphicsRAMPower
EVGA GTX 970 SC ACX 2.0 Mushkin Enhanced Blackline 8GB (2 x 4GB) 240-Pi... FirePower ZX Series 1250W 80Plus Gold 
  hide details  
Reply
post #35 of 314
As Chancellor Gorkon once said;

"Well... I see we have a long way to go."
post #36 of 314
Quote:
Originally Posted by hertz9753 View Post

I ran a projection of ONE type of a point based handicap last night on the spreadsheet. EVGA just ran away with it, and left us all in the dust.
Was that based on last years CC, or using an average PPD since then? Maybe both?

It used last years CC points as the basis, but my own projections. When I say "ran away", I maybe should have said "shot like a bolt of greased lightning". redface.gif

There ARE some other ways to handicap though, if you want to explore them.

1) TIME. The teams are assigned a time handicap. Each team then starts the CC at a different hour or even day! The goal would be to handicap it so each team would finish at the same time, same day.

This format diminishes the excitement of the race, imo. Until all the teams are in the race, it's rather boring.

2) BOOST: The teams are assigned a points handicap or boost. Each team then starts the race with a different points total, right from the start. The goal of the boost would be again, to have all the teams finish at the same time.

This is a fun one - better than #1. Another advantage is that the boost formula can be easily changed for next year -- it "tweaks" really easily.

3) SCALE: All teams start out at the same time, with the same zero points, BUT each team gets awarded points on a handicapped scale - big teams fold more points, but their actual points translate into fewer points for the race. This is the kind of race we had last year. Chimp points were the racing points. The basis or handicap is used to translate one type of points, into race points.

Also a lot of fun.

There are other ways, but they seem arbitrary, complex, or even silly, but I don't know them all by a *long* stretch. If you have a format you want discussed, jump on in and let's discuss. I can't analyze it fully without data in a spreadsheet, but we can discuss general formats.
post #37 of 314
This is an example of a Boost type of handicap race, using EVGA's CC total points * 2 as the goal that each team must reach. Note that each team starts with a different boosted value, but at the end of the race, (ideally), they will all have folded twice as many points as they did last time, and make it a close race. biggrin.gif

The actual number of days for the race will vary, depending on the handicapper's guess that the teams will be as strong as he/she estimates. With a poor estimate, the race could be over in three days, or run on for two weeks, for example. It all depends on how the handicapper can handicap it!
thumb.gif
Code:
Boost:       Equals EVGA * 2 240,558,512                                                                     

                Start >>            Day 1         Day 2           Day 3           Day 4           Day 5             Day 6           Day 7           Day 8          Day 9            Finish
Cbit_chimps     152,662,594     161,452,186     170,241,778     179,031,369     187,820,961     196,610,553     205,400,145     214,189,737     222,979,328     231,768,920     240,558,512
EVGApes                   0      24,055,851      48,111,702      72,167,554      96,223,405     120,279,256     144,335,107     168,390,958     192,446,810     216,502,661     240,558,512
ChimpPwrUp      186,025,842     191,479,109     196,932,376     202,385,643     207,838,910     213,292,177     218,745,444     224,198,711     229,651,978     235,105,245     240,558,512
MonkeyBollocks  210,260,306     213,290,127     216,319,947     219,349,768     222,379,588     225,409,409     228,439,230     231,469,050     234,498,871     237,528,691     240,558,512
MaxMonkeys      183,407,828     189,122,896     194,837,965     200,553,033     206,268,102     211,983,170     217,698,238     223,413,307     229,128,375     234,843,444     240,558,512
Beavers         159,237,784     167,369,857     175,501,930     183,634,002     191,766,075     199,898,148     208,030,221     216,162,294     224,294,366     232,426,439     240,558,512
OCNChimpin'      25,921,040      47,384,787      68,848,534      90,312,282     111,776,029     133,239,776     154,703,523     176,167,270     197,631,018     219,094,765     240,558,512
T32monkeys      155,217,880     163,751,943     172,286,006     180,820,070     189,354,133     197,888,196     206,422,259     214,956,322     223,490,386     232,024,449     240,558,512
TSChimp         149,232,538     158,365,135     167,497,733     176,630,330     185,762,928     194,895,525     204,028,122     213,160,720     222,293,317     231,425,915     240,558,512

For this example, I just used the highest point value from 2011 CC * 2.

As always, let me know what you think. I believe this is a very viable format. Both big and little handicapped teams tend to do well in this, in my experience with it. It may be the best format so far.
Edited by Adak - 4/12/12 at 10:52pm
post #38 of 314
I will agree with keeping the orignal team names. So the boost is based on Evga? I will ask again, Is this based on last years CC?
    
CPUMotherboardGraphicsGraphics
Intel i7 2600k ASUS P8Z77WS EVGA GTX 980 SC ACX 2.0 EVGA GTX 980 ACX 2.0 
GraphicsRAMPower
EVGA GTX 970 SC ACX 2.0 Mushkin Enhanced Blackline 8GB (2 x 4GB) 240-Pi... FirePower ZX Series 1250W 80Plus Gold 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel i7 2600k ASUS P8Z77WS EVGA GTX 980 SC ACX 2.0 EVGA GTX 980 ACX 2.0 
GraphicsRAMPower
EVGA GTX 970 SC ACX 2.0 Mushkin Enhanced Blackline 8GB (2 x 4GB) 240-Pi... FirePower ZX Series 1250W 80Plus Gold 
  hide details  
Reply
post #39 of 314
Quote:
Originally Posted by Adak View Post

This is an example of a Boost type of handicap race, using EVGA's CC total points * 2 as the goal that each team must reach. Note that each team starts with a different boosted value, but at the end of the race, (ideally), they will all have folded twice as many points as they did last time, and make it a close race. biggrin.gif
The actual number of days for the race will vary, depending on the handicapper's guess that the teams will be as strong as he/she estimates. With a poor estimate, the race could be over in three days, or run on for two weeks, for example. It all depends on how the handicapper can handicap it!
thumb.gif
Code:
Boost:       Equals EVGA * 2 240,558,512                                                                     
                Start >>            Day 1         Day 2           Day 3           Day 4           Day 5             Day 6           Day 7           Day 8          Day 9            Finish
Cbit_chimps     152,662,594     161,452,186     170,241,778     179,031,369     187,820,961     196,610,553     205,400,145     214,189,737     222,979,328     231,768,920     240,558,512
EVGApes                   0      24,055,851      48,111,702      72,167,554      96,223,405     120,279,256     144,335,107     168,390,958     192,446,810     216,502,661     240,558,512
ChimpPwrUp      186,025,842     191,479,109     196,932,376     202,385,643     207,838,910     213,292,177     218,745,444     224,198,711     229,651,978     235,105,245     240,558,512
MonkeyBollocks  210,260,306     213,290,127     216,319,947     219,349,768     222,379,588     225,409,409     228,439,230     231,469,050     234,498,871     237,528,691     240,558,512
MaxMonkeys      183,407,828     189,122,896     194,837,965     200,553,033     206,268,102     211,983,170     217,698,238     223,413,307     229,128,375     234,843,444     240,558,512
Beavers         159,237,784     167,369,857     175,501,930     183,634,002     191,766,075     199,898,148     208,030,221     216,162,294     224,294,366     232,426,439     240,558,512
OCNChimpin'      25,921,040      47,384,787      68,848,534      90,312,282     111,776,029     133,239,776     154,703,523     176,167,270     197,631,018     219,094,765     240,558,512
T32monkeys      155,217,880     163,751,943     172,286,006     180,820,070     189,354,133     197,888,196     206,422,259     214,956,322     223,490,386     232,024,449     240,558,512
TSChimp         149,232,538     158,365,135     167,497,733     176,630,330     185,762,928     194,895,525     204,028,122     213,160,720     222,293,317     231,425,915     240,558,512
For this example, I just used the highest point value from 2011 CC * 2.
As always, let me know what you think. I believe this is a very viable format. Both big and little handicapped teams tend to do well in this, in my experience with it. It may be the best format so far.

Edited
    
CPUMotherboardGraphicsGraphics
Intel i7 2600k ASUS P8Z77WS EVGA GTX 980 SC ACX 2.0 EVGA GTX 980 ACX 2.0 
GraphicsRAMPower
EVGA GTX 970 SC ACX 2.0 Mushkin Enhanced Blackline 8GB (2 x 4GB) 240-Pi... FirePower ZX Series 1250W 80Plus Gold 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel i7 2600k ASUS P8Z77WS EVGA GTX 980 SC ACX 2.0 EVGA GTX 980 ACX 2.0 
GraphicsRAMPower
EVGA GTX 970 SC ACX 2.0 Mushkin Enhanced Blackline 8GB (2 x 4GB) 240-Pi... FirePower ZX Series 1250W 80Plus Gold 
  hide details  
Reply
post #40 of 314
Like i said PAIN IN THE ASS! what?, are you two writing a book with your posts? learn to sum things up.
You two make it a pain in the ass in regards to to the point system it's ALL for fun ! and research rolleyes.gif

You guys are starting to miss the point rolleyes.gif
Edited by Bodycount - 4/12/12 at 11:20pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Overclock.net Folding@Home Team
Overclock.net › Forums › Overclockers Care › Overclock.net Folding@Home Team › Chimp Challenge 2012??