Well if they say that it can work if we drop conversion and use stamina instead, then let's drop the chimp names. Its what the team wants so Im here to represent that. Good job guys, in figuring this out.

zodac

I think a few teams have given their answers. HWC and OCN are for keeping the CC names, OCF are against, Evga are 66% against and OCAU are 50/50. Still no replies from TSC, TPU or CPC&B.

So it's not really clear-cut enough for us to make a majority decision. :/

 
you can get hold of TSC directly thru Bosun
I can post what you want overs at TSC  just send me a PM (contents to send) and I'll post it there
I can also post at CPC&B and TPU if someone hasn't already
 
bit-tech will have a delay post (has to go thru modration)
 
 
 

Xavier Zepherious

zodac

I think a few teams have given their answers. HWC and OCN are for keeping the CC names, OCF are against, Evga are 66% against and OCAU are 50/50. Still no replies from TSC, TPU or CPC&B.

So it's not really clear-cut enough for us to make a majority decision. :/

you can get hold of TSC directly thru Bosun
I can post what you want overs at TSC  just send me a PM (contents to send) and I'll post it there
I can also post at CPC&B and TPU if someone hasn't already

bit-tech will have a delay post (has to go thru modration)

Apparently adak went to TPU and TSC a while back. He says theyre 100% against changing names

devdog51

Lol @ Afterburner. Alright Gryphon, or anyone else can you work up the formula and put it into one post? Everyone at the captain's formula is demanding to see it. I asked for a vote on the date.

 
I'm super busy at work right now, but I'll do that by tonight. I know it's dispersed all over in a number of posts now, so till then you can refer them to mainly posts #502 and #632.
 
Just to clarify, you need both the updated CCC formula and the Stamina formula together, right? Is that what's being voted on now?
 
I'm going to run a scenario for "what if there were no CC names in 2011" but the "Growth" will be inaccurate because all I have for 2010 is CC name PPD's (so it's going to calculate CC-name growth for now, but it will correctly calculate the total team growth if CC-names are dropped in CC 2012). 
 
 

Bosun made a post in this thread a week or so ago. I believe he said that TSC would go with the majority. 

I have made a final cosmetic change in the way I calculate the scores, only so that it's easier to interpret. Below is the whole account of how to calculate a single score from a number of different categories (no matter what they are).
Please refer anyone here to read on this. I will make another post on each of the performance measures (categories).

CCC - Categorized Chimp Challenge  
All teams compete in a number of categories (performance measures). Lets assume there are three categories and let's call them CAT1, CAT2 and CAT3.
At the end of the competition, each team gets their scores in all categories, and the team with the best overall performance is crowned "King of the Chimps" and receives this year's jaded monkey.  

I will now explain how to come up with that "overall performance". Let's work on a 4-Team example:  

CAT1                               CAT2                           CAT3 
Team A: 140                Team A: 50              Team A:  30 
Team B:  60                Team B: 70                Team B:  60 
Team C: 100                Team C: 90              Team C:  10  
Team D:  80                Team D: 80              Team D:  40 

At this point, the scale of scores in each category is completely different, and this poses a serious concern when you want to put them together in a single score by adding them. For example, in CAT1, Team A has an 80 point advantage over Team B and this spread is impossible to close whatever no matter how well Team B does in other categories.

To remedy this, we will bring all categories to the same scale from 0 (zero) to 100, such that the best team in each category will receive a 100, and the worst team will receive a zero. This way, no single category will have a dominance in the overall score and the categories will be completely balanced.
To this end, first, calculate the "range" within each category by subtracting the minimum score from the maximum.

             CAT1                   CAT2                    CAT3 
Range         80                     40                      50

Let's call these ranges, RANGE1, RANGE2 and RANGE3.
Then, multiply the scores in CAT1 by 100/RANGE1, scores in CAT2 by 100/RANGE2, and scores in CAT3 by 100/RANGE3. This is what you get:
 
CAT1                                CAT2                            CAT3 
Team A: 175                Team A: 125              Team A:  60 
Team B:  75                  Team B: 175              Team B: 120 
Team C: 125                Team C: 225              Team C:  20  
Team D: 100                Team D: 200              Team D:  80 

Observe now that the ranges in all categories are equal to 100; exactly what we wanted. 

At this point, the goal is actually achieved (categories are equally balanced); however, for neatness and ease of interpretation, we'll bring all categories to a score between 100 and zero. This is very easy, because all you need to do is to subtract the minimum score in each category from all scores in that category. So, this is what you get:

CAT1                               CAT2                          CAT3 
Team A: 100                Team A:   0              Team A:  40 
Team B:   0                   Team B:  50              Team B: 100 
Team C:  50                Team C: 100              Team C:   0  
Team D:  25                Team D:  75              Team D:  60 

The scores can easily be interpreted as "given the best team is 100, and worst team is 0, how well Team X did in this category". A score of 50, for example means that that team did exactly halfway between the best and the worst teams.

As a final step, all you have to do is sum up the scores for each team:

FINAL SCORES
Team A: 140
Team B: 150
Team C: 150
Team D: 160

Team D wins.

To summarize, what this system does is adjusting the category scores such that each category is between zero and 100. Let's call this adjusted categories ACAT1, ACAT2, and ACAT3. 

For a shorthand notation, I will use ACATi, CATi and RANGEi for category i, and ACATij, CATij for Team j's scores in category i. Finally, let's call Team j's final score, SCOREj.
The formula can be written as:

SCOREj = ACAT1j + ACAT2j + ACAT3j
where
ACATij = CATij x 100 / RANGEi

For the Stamina category
 
Raw scores:
TPU - 100.0
HWC - 99.1
Evga - 92.6
OCN - 89.9
 
Range = 10.1, normalised scores = (raw score*100)/Range
 
Normalised scores
TPU = 990.1
HWC = 981.2
Evga = 916.8
OCN = 890.1
 
Final scores
TPU = 100
HWC = 91.1
Evga = 26.7
OCN = 0
 
The difference for TPU and HWC wasn't even 1% in the category, and yet they get almost 10% less points.
 
Is there a way to reign that in somehow? I can understand the normalisation for the scores when they can be in widely varying ranges, but when they're all bunched together, there's a big difference for a small performance gap.
 
Just a comment on that category - I'm happy to use this format as is this year if the captains agree to use it; it's the most promising system that's been proposed so far, and if we find any flaws in this year's event, hopefully we'll get together a bit sooner than late next March to discuss it. :P

Just a few thoughts, take them for what they are worth:
 
In trading conversion or stamina you are trading a real, measurable metric for something essentially meaningless and difficult to measure in a meaningful way.  You can come up with all the assumptions and statistical tricks you want, that still isn't going to make it meaningful.
 
More importantly, removing the chimps from the chimp challenge makes the contest ordinary and boring.  You do this because you take adak at his word when few teams appear to be actually involved in the process and several teams have said "we're up for whatever."   Now the schict for the contest is essentially: don't really do anything you aren't already doing except pay closer to the stats than usual.  This does not get me excited in the least. 
 
Also, with that proposed format we are going to win this year in a runaway.  The contest will suffer for it and I am sure there will be accusations that we rigged the contest.  To which I say
 
Don't miss the forest for the trees...

zodac
 The difference for TPU and HWC wasn't even 1% in the category, and yet they get almost 10% less points.

Is there a way to reign that in somehow? I can understand the normalisation for the scores when they can be in widely varying ranges, but when they're all bunched together, there's a big difference for a small performance gap.

Just a comment on that category - I'm happy to use this format as is this year if the captains agree to use it; it's the most promising system that's been proposed so far, and if we find any flaws in this year's event, hopefully we'll get together a bit sooner than late next March to discuss it. :P

 
zodac, the purpose is to bring the different categories to an equal scale so that when you add them up it is actually adding apples to apples. Say in Stamina, the difference between TPU and HWC was 1, but in Points the difference is 10,000,000. I can hear you saying 'why not convert the Points to number of millions?' but that would be an arbitrary unit conversion, which still does not guarantee that two categories equally affect the final score. What I'm doing by bringing all categories to the same scale of zero to 100, exactly brings all categories to the same level.
 
One thing you (all participants) should realize is that once the category scores are adjusted, they are on a different scale and they're not comparable to their original values. It's like converting Fahrenheit temperature to Celsius. 212 F is equal to 100 C, right? Consider a typical room temperature, say 68 F, which is equal to 20 C. Question: how many degrees is there between a boiling water and a typical room? The answer is 144 in F, but 80 in C. I hope I'm making things clear.
 

Post #672 is updated with details on Stamina.  All the information there is is now on posts #671 and #672, please refer anyone to those posts. I'll be happy to answer questions or clarify things when I can.
 
By the way, I'll be traveling in a few hours and I won't have internet connection probably till late night ET.

Gryhon, apparently youre not allowed to not fold in the CC.

adak
Sit it out?

B****!

Tell him I like it, and he's folding! <smile>

Well done.

 
Also, Adak agrees to the system with one request:

adak
I really like the stats post you made, three posts up from this one! I would vote for that in an instant.

I'd like to make ONE SMALL SUGGESTION, and it's strictly cosmetic. I'd like to see the lowest possible score 100, instead of a zero. For a team to see a goose egg for their racing efforts, is a cold slap in the face, I'd like to avoid.

I know, it's strictly a feel good measure, but a goose egg is a bit tough to face. Could the range be from 10 to 110 perhaps?

Tell your stats guys, good job!

He even complimented you guys, lol.
 

csm725OCN

I personally think the range should be 50-100 because 100 points less is quite a lot for a change that is minuscule in reality (no more than 15 percent points)...
Or am I not comprehending this correctly?
Edit, @devdog, your badge isn't working since EOC is down.

That would explain it then. Ive been crunching so I haven't looked at my EOC for a while. And apparently WCG is done right now too. It could probably be changed to that but would require a change to the formula that converts the numbers. I assume it would be fairly easy. I would have to through the formula again. But a shift like what Adak is asking for wouldn't necessarily change anything. Its just assigning a different "name" if you will to the place that holds the value. As long as there are 100 parts you can specify any set of numbers you want and the system stays the same. Is this correct Gryphon? I think Im understanding this, but feel free to slap me if Im wrong.

The shift adak is suggesting may help psychologically but it wouldn't affect the outcome. Any scale and any lower limits are possible but a range from 0 to 100 is very nice from a pure mathematical pov.

CSM, please pay attention that the adjusted (scaled) scores are not comparable to original ones and the meaning of the adjusted scores is totally different. They're now relative scores with respect to the max and min in that category. And, again, we're doing this because this is the only way to make sure the categories have the same affect on the final score.

Ok, but here're some numbers I put together. Using current stats from EOC:
 

 
 
The % difference between CPC and OCAU in raw stats is 3.5%, and % diff in final stats is 4.4%.
However, in the Stamina category, a difference of 1% gave a final stats % different of almost 10%.
 
That's what might need to be looked at - there seems to be a lot more "spreading" of the points in the Stamina category, since the scores are so bunched up.

To Gryphon:  Just wanted to say that you have and are still doing an "outstanding job" for everyone in the Chimp Challenge!  No matter if the formula is absolutely flawless or not, your work, interest and seemingly tireless ability to keep on plugging are incredible.  I'm proud to be on the same Team with you buddy.  You deserve so much more than a Blue Ribbon for your efforts, but I think all of us can agree that you have earned that and a lot more with your donation of time and skills.  Great job!!