Doitsujin -> RE: INTERMEDIATE CHALLENGE ENTRIES Mp3 (Feb. 28 2009 7:12:30)
|
Indeed a great challenge! Oley guys and gals! Did I mention that I love stats?? Here we go! ;.) Now the missing point of the i of the challenge! ;.) Because I have not much time.. I´ll describe all very shortly and sloppy with many writing mistakes.. Sorry. I´ll correct them later. Here are some statistics which are maybe interesting! As we already know, Peter Pan is (Jimi Hen..xyz) is the winner, followed by Simba and Cheshire cat (A). But when going a bit further into detail, the total points which were given by all judges were sometimes higher and sometimes a bit lower. So I calculated the median and the standard error of the difference of the given points. And we can clearly see, that Peter Pan, Simba and Cheshire cat are not judged significantly different (B). So, they share the first place in the perception of “nerds” haha… The numbers indicate the “real” placement of the challengers. I drew a red line to visualize who is significantly different (B). Than I threw an eyeball on the judges. I changed their given points in the % of given points. I set 100% for the the highest given points for each judge individually to eliminate the fact, that some judges rated overall with higher points and others at a lower level. In other words, I normalized all judges to be able to compare the challengers more objectively. (By the way, I also did that for (B) and it of course doesn’t change anything, just the measuring unit from points to percentage.) And we can see clearly that the judges had more or less the same opinion about the challengers (E left side). Well.. the most disagreement occurred about Cheshire cat (E left side). I think there might be potential.. haha..The judges were all more or less in common about the 2 best challengers, Peter Pan and Simba. Also about Uncle Sam and Baloo… (sorry for that.. ;.)) Well.. so far. I wanted to know which judge thinks most often completely different than the other judges. So, I calculated the “level of agreement” between the judges (E right side) on the basis of the median of given points of all judges and sub ducted the particular points from each individual judge. The numerical value in the right sheet is 0 when a judge hite exactly the average opinion of all judges for each challenger and each category. Negative numbers show that a judge voted something worser than the average of all judges. And if somebody voted more positively, you see a positive number. To have a better overview, I stained numbers which are 15% over or under the average green and red. That shows, which judge thought very different than the other judges… (You see the summary in D) (EDIT: SORRY THE HEADLINE OF D IS FALSE. IT SHOWS HOW OFTEN A UDGE VOTED OVER OR UNDER THE MEDIAN IN TOTAL. SO 0,0001% BETTER OR WORSER THAN THE AVERAGE ALREADY COUNTS. What do we see: Florian and Sean are nice people… they vote most often much much better than the other judges.. They are “pussy footing” good judges.. hehe ;.) ToddK, Norman and me are evil judges (or very critic and clear judges) ;.) We rate much lower than the average. Ron and Jason are maybe the best judges… they are completely tared. Seeing the things clear but aren’t too pedantic. The rest is also fine (D). (E right side only shows the details on which the diagram in D is based) And the last diagram which I want to show is the overall generosity of each judge. In other words, how often they used the 10 points or always showing 5 or 6 as maximum. We can see,.. the judges used all more or less the same amount of points! They are not significantly different. BUT! But we see again… Florian and Sean are the significantly most generous judges (C indicated with an asterisk) They used most often high scores! (You pussys… hehehe) Well.. I did a lot more detailed statistics on other things and on the 3 categories but that would be too much to show here. The most striking thing would be, that the judges disagreed most in the category AIRE. And.. Florian has problems with seeing when something isn’t played in compass.. (Sorry Florian.. that’s what the statistics say.. See it funny.. my post is just for amusement) (^____^) Hahaha...
Images are resized automatically to a maximum width of 800px
|
|
|
|