Re:Chimp Challenge 2012
2012/04/16 07:04:40
(permalink)
devdog51
Exactly so, if I could get one of the admins to run a couple polls? The first one is whether or not people would like the idea of the CC being pushed to colder months? This would really help us hammer out a new system, plus like some have mentioned it will allow us to fold at higher rates during colder weather.
Second poll is: Does everyone wanna keep the Chimp names or do the want to retain their own usernames? These are two questions that definitely need to get answered now.
Yes I can... Offered it a few days ago and a sample in in this forum... I just need to know what questions to ask and for how long you want it. One thing to note. I will not allow anyone to vote, they have to have a log in. And I will only allow two votes per user. This means if two of the questions are something a person could go with, they can vote for both. I feel forcing a person for one idea is not showing the true flexibility in our team and we may end up with frustrations as a result. Or worse folks deciding to not vote at all. http://forums.evga.com/tm.aspx?m=1549213 
Attached Image(s)
|
devdog51
ACX Member
- Total Posts : 389
- Reward points : 0
- Joined: 2011/02/21 23:58:17
- Location: Lafayette, IN
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 07:25:08
(permalink)
Thanks that works. So I guess the options for time would be:
Late spring (no change)
Early spring- late winter (Jan-Apr)
Summer (June-Sept)
Fall- early winter (Oct- Dec)
|
texinga
CLASSIFIED ULTRA Member
- Total Posts : 5121
- Reward points : 0
- Joined: 2010/07/03 14:30:32
- Status: offline
- Ribbons : 20
Re:Chimp Challenge 2012
2012/04/16 07:26:13
(permalink)
|
Re:Chimp Challenge 2012
2012/04/16 07:34:30
(permalink)
|
texinga
CLASSIFIED ULTRA Member
- Total Posts : 5121
- Reward points : 0
- Joined: 2010/07/03 14:30:32
- Status: offline
- Ribbons : 20
Re:Chimp Challenge 2012
2012/04/16 07:37:15
(permalink)
AB, for the other question that Chris requested, it could go like this: Annual Chimp Challenge Contest Timing: There have been some suggestions that it would be better for our Folding Rigs and for all-out contribution to Chimp Challenge each year if it was held at a time when outdoor temperatures are cooler. Do you feel that it would be better to hold the CC during cooler months rather than on May 5th each year? This question assumes that CC could be held in a month that benefits Folders World-wide, so the best alternate month would be one that addresses the issue for all. - Yes, I would like to see future CC's moved to the Fall or Winter months.
- No, I like it on May 5th just where it is today.
- Don't care when it is scheduled.
How's that?
|
Re:Chimp Challenge 2012
2012/04/16 07:42:47
(permalink)
adjusting **Edit** Check the link... I will add to it but are we OK with the poll or do I need to them?
post edited by Afterburner - 2012/04/16 07:44:19
|
Re:Chimp Challenge 2012
2012/04/16 07:43:11
(permalink)
|
Re:Chimp Challenge 2012
2012/04/16 07:48:25
(permalink)
I wonder... Where should we have these polls? Here or the general folding category for traffic?
|
devdog51
ACX Member
- Total Posts : 389
- Reward points : 0
- Joined: 2011/02/21 23:58:17
- Location: Lafayette, IN
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 07:50:04
(permalink)
Afterburner
I wonder... Where should we have these polls? Here or the general folding category for traffic?
I say general folding just for the traffic. But youre the mod so you probably have a better idea of the amount of people who visit this thread.
|
texinga
CLASSIFIED ULTRA Member
- Total Posts : 5121
- Reward points : 0
- Joined: 2010/07/03 14:30:32
- Status: offline
- Ribbons : 20
Re:Chimp Challenge 2012
2012/04/16 07:56:04
(permalink)
I thing General Folding too to get the best visibility. It may also start raising some awareness to CC for Folders that have not been following the Contest thread.
|
texinga
CLASSIFIED ULTRA Member
- Total Posts : 5121
- Reward points : 0
- Joined: 2010/07/03 14:30:32
- Status: offline
- Ribbons : 20
Re:Chimp Challenge 2012
2012/04/16 07:57:33
(permalink)
Is there a way to also put some sort of Sticky Pointer in the Contest thread to the Poll(s)? That may help cover more bases for awareness. Ooops, I already see that the Polls are in the Contest thread which is OK. Maybe we need something over in the Main Folding thread to call attention to the Polls?
post edited by texinga - 2012/04/16 07:59:26
|
devdog51
ACX Member
- Total Posts : 389
- Reward points : 0
- Joined: 2011/02/21 23:58:17
- Location: Lafayette, IN
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 07:57:50
(permalink)
Ill get back to you later on the choices for name change.
|
Re:Chimp Challenge 2012
2012/04/16 09:27:12
(permalink)
Punchy
theGryphon
I just some ran the numbers. I took the folding data from CC 2010 as a reference for growth rates. I know in 2010 it was a different format (till 20 mil), but I took the daily PPD as the basis and multiplied it by 10 (as in 10 days of folding in CC 2011). This is not an accurate reference point but it works to show the system.
Results are attached below, ranked according to "Final Points". Don't get caught with who won, or should have won. The point is that this system works very well.
Edit: Sorry, table didn't paste well.
I'd suggest using the geometric mean of the "relative" values in each column rather than the sum, as the geomean should help to even out the differences in weighting between the columns. Otherwise total points will always have the largest deltas between lowest and highest relative values, giving the bigger teams an extra advantage.
I understand your concern and I would like to remedy that but I don't know what you'd like me to do with the geometric mean. As a remedy, I took the square root of the raw points, and calculated the relative points performance using the square rooted points. The results are very interesting and much better balanced in that I think it does not favor larger teams anymore. For example, TPU rises from 3rd to 1st place, OCN goes from 1st to 2nd place, EVGA goes from 3rd to 5th place. TPU wins because they appear to do extremely well in Growth and Conversion, although they don't do well at all in Points. Take a look at the attached table.
post edited by theGryphon - 2012/04/16 09:32:20
Attached Image(s)
|
Re:Chimp Challenge 2012
2012/04/16 09:36:03
(permalink)
Just to add, if anybody still thinks that the above scheme (with square root) still favors the larger teams, you're basically arguing that "Points" should not be included in the categories.
|
Punchy
CLASSIFIED Member
- Total Posts : 2872
- Reward points : 0
- Joined: 2010/02/06 09:33:05
- Status: offline
- Ribbons : 14
Re:Chimp Challenge 2012
2012/04/16 09:53:46
(permalink)
geomean(a,b,c) = (a*b*c)^1/3 From Wikipedia : A geometric mean is often used when comparing different items- finding a single "figure of merit" for these items- when each item has multiple properties that have different numeric ranges. For example, the geometric mean can give a meaningful "average" to compare two companies which are each rated at 0 to 5 for their environmental sustainability, and are rated at 0 to 100 for their financial viability. If an arithmetic mean was used instead of a geometric mean, the financial viability is given more weight because its numeric range is larger- so a small percentage change in the financial rating (e.g. going from 80 to 90) makes a much larger difference in the arithmetic mean than a large percentage change in environmental sustainability (e.g. going from 2 to 5). The use of a geometric mean "normalizes" the ranges being averaged, so that no range dominates the weighting, and a given percentage change in any of the properties has the same effect on the geometric mean. So, a 20% change in environmental sustainability from 4 to 4.8 has the same effect on the geometric mean as a 20% change in financial viability from 60 to 72.
post edited by Punchy - 2012/04/16 09:56:43
|
Re:Chimp Challenge 2012
2012/04/16 10:11:49
(permalink)
Punchy
geomean(a,b,c) = (a*b*c)^1/3
From Wikipedia : A geometric mean is often used when comparing different items- finding a single "figure of merit" for these items- when each item has multiple properties that have different numeric ranges. For example, the geometric mean can give a meaningful "average" to compare two companies which are each rated at 0 to 5 for their environmental sustainability, and are rated at 0 to 100 for their financial viability. If an arithmetic mean was used instead of a geometric mean, the financial viability is given more weight because its numeric range is larger- so a small percentage change in the financial rating (e.g. going from 80 to 90) makes a much larger difference in the arithmetic mean than a large percentage change in environmental sustainability (e.g. going from 2 to 5). The use of a geometric mean "normalizes" the ranges being averaged, so that no range dominates the weighting, and a given percentage change in any of the properties has the same effect on the geometric mean. So, a 20% change in environmental sustainability from 4 to 4.8 has the same effect on the geometric mean as a 20% change in financial viability from 60 to 72.
You just didn't quote Wikipedia on geometric mean, did you Punchy?  I have a Ph.D. in Management Science, so I know what it is. The question is, how do you suggest I use it for the purpose at hand, which is to reduce the range of relative points so that large teams are not favored? I just did that above, using square root. Please take a look. I experimented with base 2 logarithm, and it works similarly well but with somewhat different results. I'm attaching the table with logarithm (base 2) adjusted points.
Attached Image(s)
|
csm725OCN
New Member
- Total Posts : 39
- Reward points : 0
- Joined: 2012/04/14 11:28:12
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 10:12:46
(permalink)
I think it's a mistake giving conversion and growth as much weight as points.
|
Punchy
CLASSIFIED Member
- Total Posts : 2872
- Reward points : 0
- Joined: 2010/02/06 09:33:05
- Status: offline
- Ribbons : 14
Re:Chimp Challenge 2012
2012/04/16 10:35:14
(permalink)
@theGryphon, you said you didn't know what I wanted you to do with geometric mean, so I explained how to use it. How else was I supposed to interpret that statement? I'm suggesting (rel points * rel conversion * rel growth)^1/3, in case you care to try it; otherwise I'll do it when I get home. Geometric mean is used for exactly the situation we have here, so I don't see the need to create a new method rather than using the most widely-accepted statistical method of combining numbers of different scales in a way that gives them equal weighting.
post edited by Punchy - 2012/04/16 10:45:46
|
Re:Chimp Challenge 2012
2012/04/16 10:49:22
(permalink)
Punchy
@theGryphon, you said you didn't know what I wanted you to do with geometric mean, so I explained how to use it. How else was I supposed to interpret that statement?
Geometric mean is used for exactly the situation we have here, so I don't know why you would create your own method rather than using the most widely-accepted statistical method of combining numbers of different scales in a way that gives them equal weighting.
You throw the idea of geometric mean but you don't specify the way how I should use it. But, I think I figured what you're trying to say. You're saying, instead of summing the relative performances across categories, we can take the geometric mean. I just tried that (see the attachment). It's giving very similar results. The problem I think is that it still favors large teams, because the the range (maximum minus minimum) in Points (even in Relative Points) is much larger. I have an idea to solve this issue. I'll post it up right away...
Attached Image(s)
|
devdog51
ACX Member
- Total Posts : 389
- Reward points : 0
- Joined: 2011/02/21 23:58:17
- Location: Lafayette, IN
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 11:02:22
(permalink)
But is it still going to favor the small teams too much? That's the thing that needs to be fixed above all else. Oh yea, zodac is pushing for a reschedule on the date for CC this year, at least to allow us time to figure all this out. Adak seems convinced that we should go ahead with May 5th, but Im thinking that this is nowhere near being ready for us to start that soon.
|
devdog51
ACX Member
- Total Posts : 389
- Reward points : 0
- Joined: 2011/02/21 23:58:17
- Location: Lafayette, IN
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 11:04:20
(permalink)
csm725OCN
I think it's a mistake giving conversion and growth as much weight as points.
But if you gave points the highest weight it would in turn favor only the large teams. Of course OCN and EVGA are going to come out on top for total points, if they run like they should.
|
Re:Chimp Challenge 2012
2012/04/16 11:09:02
(permalink)
Alright, there was a problem all along with my proposal in that the range (maximum minus minimum) within different categories is too different such that when you some things up, it's like adding apples to oranges. I implemented an idea to remedy this. It's very simple. Instead of calculating relative performance, you find the range within each category, and determine the category with the smallest range. Then, you bring the range of each category to this same range by a simple division (by the ratio of ranges). At the end, what you have is a set of categories with the same range in itself. Think of it as bringing different measures to the same scale. I'm attaching the results. This is my final proposal as it completely solves the issue of imbalanced weighing, as long as we agree on what categories to include.
Attached Image(s)
|
csm725OCN
New Member
- Total Posts : 39
- Reward points : 0
- Joined: 2012/04/14 11:28:12
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 11:22:04
(permalink)
I still think conversion points weight should be decreased. It makes no sense that for annihilating TPU points wise EVGA get 40 points and for having a better conversion rate for such a small team, TPU get 40 as well.
|
devdog51
ACX Member
- Total Posts : 389
- Reward points : 0
- Joined: 2011/02/21 23:58:17
- Location: Lafayette, IN
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 11:24:35
(permalink)
theGryphon
Alright, there was a problem all along with my proposal in that the range (maximum minus minimum) within different categories is too different such that when you some things up, it's like adding apples to oranges. I implemented an idea to remedy this.
It's very simple. Instead of calculating relative performance, you find the range within each category, and determine the category with the smallest range. Then, you bring the range of each category to this same range by a simple division (by the ratio of ranges). At the end, what you have is a set of categories with the same range in itself. Think of it as bringing different measures to the same scale.
I'm attaching the results. This is my final proposal as it completely solves the issue of imbalanced weighing, as long as we agree on what categories to include.
Are there any other categories that you could suggest including? Im going to post this to the CC Captains forum, for them to view it so if you have any other suggestions to go with it, Ill include them.
|
shdbcamping 
FTW Member
- Total Posts : 1602
- Reward points : 0
- Joined: 2008/10/10 09:39:41
- Location: Erie, Pa
- Status: offline
- Ribbons : 3
Re:Chimp Challenge 2012
2012/04/16 11:31:31
(permalink)
This is like "way" too complicated, and like last year we will end up shafted (as I suspect [H] will). I personally think our Team should abstain. Edit: I still think multiple "monkey" awards should be the way to go. most points, most improved most points/capita etc, etc, etc
post edited by shdbcamping - 2012/04/16 11:36:15
|
zodac
New Member
- Total Posts : 73
- Reward points : 0
- Joined: 2009/07/10 08:58:25
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 11:32:17
(permalink)
I probably missed it earlier, but how is "Growth Rate" being calculated?
|
Re:Chimp Challenge 2012
2012/04/16 11:38:21
(permalink)
zodac
I probably missed it earlier, but how is "Growth Rate" being calculated?
My understanding was that it's the growth rate compared to last year's CC. In the calculations I made, I used the daily PPD reported for CC2010 as the basis. By the way, devdog51 and all, my proposal is NOT about what categories should be included. It's about how to reasonably put them together. The assumption so far has been the categories would be equally weighed (which makes most sense to me), but if the teams decide to weigh one category more or less than others, it would be very easy to implement.
|
zodac
New Member
- Total Posts : 73
- Reward points : 0
- Joined: 2009/07/10 08:58:25
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 12:16:11
(permalink)
theGryphon
zodac
I probably missed it earlier, but how is "Growth Rate" being calculated?
My understanding was that it's the growth rate compared to last year's CC. In the calculations I made, I used the daily PPD reported for CC2010 as the basis.
By the way, devdog51 and all, my proposal is NOT about what categories should be included. It's about how to reasonably put them together. The assumption so far has been the categories would be equally weighed (which makes most sense to me), but if the teams decide to weigh one category more or less than others, it would be very easy to implement.
Ahh, got you. As for how to weight the categories, I'm fine with total points and relative points being equal, but conversion probably should be weighted a bit less. Simply because if you have low conversion, you're gonna be hit in that category, the total points category, and the relative points category. It should be a factor, but I don't think it's "worth" as much as the other two.
|
zodac
New Member
- Total Posts : 73
- Reward points : 0
- Joined: 2009/07/10 08:58:25
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 12:16:20
(permalink)
theGryphon
zodac
I probably missed it earlier, but how is "Growth Rate" being calculated?
My understanding was that it's the growth rate compared to last year's CC. In the calculations I made, I used the daily PPD reported for CC2010 as the basis.
By the way, devdog51 and all, my proposal is NOT about what categories should be included. It's about how to reasonably put them together. The assumption so far has been the categories would be equally weighed (which makes most sense to me), but if the teams decide to weigh one category more or less than others, it would be very easy to implement.
Ahh, got you. As for how to weight the categories, I'm fine with total points and relative points being equal, but conversion probably should be weighted a bit less. Simply because if you have low conversion, you're gonna be hit in that category, the total points category, and the relative points category. It should be a factor, but I don't think it's "worth" as much as the other two.
|
zodac
New Member
- Total Posts : 73
- Reward points : 0
- Joined: 2009/07/10 08:58:25
- Status: offline
- Ribbons : 0
Re:Chimp Challenge 2012
2012/04/16 12:16:21
(permalink)
theGryphon
zodac
I probably missed it earlier, but how is "Growth Rate" being calculated?
My understanding was that it's the growth rate compared to last year's CC. In the calculations I made, I used the daily PPD reported for CC2010 as the basis.
By the way, devdog51 and all, my proposal is NOT about what categories should be included. It's about how to reasonably put them together. The assumption so far has been the categories would be equally weighed (which makes most sense to me), but if the teams decide to weigh one category more or less than others, it would be very easy to implement.
Ahh, got you. As for how to weight the categories, I'm fine with total points and relative points being equal, but conversion probably should be weighted a bit less. Simply because if you have low conversion, you're gonna be hit in that category, the total points category, and the relative points category. It should be a factor, but I don't think it's "worth" as much as the other two.
|
