On the APE Voting

This topic has 1 voice, contains 21 replies, and was last updated by The Janitor 3338 days ago.

Viewing 15 posts - 1 through 15 (of 22 total)

1 2 →

Author	Posts
Author	Posts
March 30, 2004 at 7:44 am #38109
brickbrock	Hi everyone, First of all I want to make clear that I am very happy with the results and my nomination (and relative success) of the APE contest. However, the detailed rankings leave me somewhat puzzled. Thus, have a few remarks on the voting math: From the Table in the events section you can see that a number of 300 points was awarded per category. With at least 20 voters (on the News page 21 are mentioned) this is max. 15 points per voter. The Events page suggests a voting procedure that works like this: Once the nomination committee has completed their task, the nominees will be announced and the winners will be chosen by a voting system. All contestants whose film was not nominated, along with all 24fps or greater Patrons without nominated films will be asked to rank the nominees in each of the categories listed below. A film ranked at the top of a category will receive a number of points for that category equal to the number of nominations, a film ranked second will receive one fewer point, and so on. Each of the categories will be weighted and an overall score for each film will then be calculated. With 10 nominees this suggests 10 points for the first 9 for the second and so on in each category. Clearly the sum of the numbers 1 to 10 is 55, which naturally should be the number of points awarded per voter per category. So there is a difference. I assume that the voting procedure had been changed to a more competitive one, where the winner recieves some points the second a few less the third a few more less and so forth. Films below a certain score receive no points. With this post I do not suggest, that the ranking of the APE contest should be redone. BUT I would like RevMen to give a clear anouncement how it was done (which leaves me less puzzled with less then 20 points in some categories for some entries with more than 20 voters). Still happy with this great contest and its great results lurking around on my harddisk brickbrock
March 30, 2004 at 9:19 am #38113
hali	You just beat me to this. It only hit me at work toady that something did not seem right with ‘the numbers’. I was a bit confused during the ceremony… but I was also tired + excited… Rev? Hali EDIT: Rev, I’ve worked out the tallies based upon the marking scheme listed on the entries page… (doing things like awarding 1st 10 points… an equal 2nd 9points… then 4th 7 points [taking into acount 2x 2nd place cutting out third]) I am not going to post it here. I think this needs some looking into. EDIT 2: On the events page it lists me as 3rd in story, but holgor came 3rd… the numbers reflect the actual placing, but it does need updating.
March 30, 2004 at 11:12 am #38118
richardfrost	“hali” wrote: You just beat me to this. It only hit me at work toady There’s no need to call him names. I think there’s a conspiracy here. Where’s Jarrah?
March 30, 2004 at 11:30 am #38120
LowweeK	I was thinking of this too. And on another subject, I do think that weighting the Story by 50% is a bit too much. Maybe we should later diminish the overweighting of the story and give those extra points to the other criteria. I do think that story is THAT important, but here its weight was quite directly conditionning the final result, no matter the others ratings were.
March 30, 2004 at 11:31 am #38121
hali	“richardfrost” wrote: [quote="hali"]You just beat me to this. It only hit me at work toady There’s no need to call him names. I think there’s a conspiracy here. Where’s Jarrah?[/quote] LOL! That is a classic typo! Obviously I meant ‘tiny little disease spreading gnat’
March 30, 2004 at 12:57 pm #38131
brickbrock	Loweek wrote: And on another subject, I do think that weighting the Story by 50% is a bit too much. Maybe we should later diminish the overweighting of the story and give those extra points to the other criteria. I do think that story is THAT important, but here its weight was quite directly conditionning the final result, no matter the others ratings were. This also is a result of the ‘steeper’ distribution of points within one category. If there are differences up to 25 and more points the diffrence in a 10% category had to be 250 points to make up for that. at Hali, luckily, it is not possible to recalculate the results from the table that was posted in ‘Events’ as this is already the _result_ of the combined votes. The points for the individual votes would have to be adjusted to get it right according to the posted voting scheme. But I think you already realized that by yourself… brickbrock
March 30, 2004 at 1:16 pm #38135
hali	“brickbrock” wrote: at Hali, luckily, it is not possible to recalculate the results from the table that was posted in ‘Events’ as this is already the _result_ of the combined votes. The points for the individual votes would have to be adjusted to get it right according to the posted voting scheme. But I think you already realized that by yourself… brickbrock I’m not sure I agree with you. Seems to me the number in each box is either the total tally of votes for each movie in each category (of which the numbers are totally weird based upon your comments above) or some reduced (out of 100 maybe? a factor?) value. [a third option is that they have been decided upon in a new way, which you expressed above that you'd like to know] Now if the values are simply reduced or factors, then the numerical order (whatever the points) will remain the same, and following on, the rankings do not change. Thus applying the “First gets 10 points. Second gets 9″ etc etc allows you to set up a table. The big issue is that using the system described in the rules ensures the difference between places (apart from tied places and the next one down) is ALWAYS 1 point. This is what can make huge differences when you apply the set weightings of 50% 30% 10% and 10%. Not using the rankings (10, 9, 8, 7, etc) when applying the weightings has caused much more of a skew in the story category as the differences between 1st and 2nd (20) as well as 2nd and 3rd (17) are huge. [/i]
March 30, 2004 at 1:53 pm #38138
brickbrock	Hali wrote Now if the values are simply reduced or factors, then the numerical order (whatever the points) will remain the same, and following on, the rankings do not change. Thus applying the “First gets 10 points. Second gets 9″ etc etc allows you to set up a table. I have to admit that the idea that the numbers could simply be scaled down by a factor never occured to me . (Although its seems a miracle that the result is only integers) The “First gets 10 points. Second gets 9″-scale applies in my reading only to to the individual rankings by the voters, not to the combined ranking. Lets wait for RevMen sorting this out brickbrock
March 30, 2004 at 3:29 pm #38141
boettcher	Yes, the votingsystem has changed compared with the Events-page. For details see the “APE contest – how to vote” thread … http://www.brickfilms.com/phpBB2/viewtopic.php?t=3503 … Every voter has named the best five films in each category and the point where given that 5 points go to the best and 1 for the 5th best film. That makes 5 + 4 + 3 + 2 + 1 = 15 points by each voter in every category. The sum of each category is 300 points, so we had 20 voters. The only question is, why there are 21 voters listed on the News-page? Perhaps The Sisters have voted together … Dirk.
March 30, 2004 at 4:31 pm #38147
brickbrock	sometimes it pays to read first…
March 30, 2004 at 4:52 pm #38153
Cometgreen	Aye, Rev changed the numbers a few weeks ago when he asked us to vote. And I’m guessing there were more than 20 members, because there are 21 on the front page, and some of us didn’t vote until a few days before the deadline (like me ). And richard: Cometgreen
March 30, 2004 at 5:23 pm #38155
RevMen	Controversy!! Conspiracy!! Get Jarrah in here immediately! I think I’ve discovered the problem. There WERE more than 20 voters, there were 22. I used an MS Excel spreadsheet to calculate the votes and I did check my work before announcing the winners, but not in enough ways. I checked the list of voters on the left hand side of each page of the spreadsheet to make sure they were all there (22). I checked the individual totals for each voter to make sure that all of their votes were adding up across each of the category pages correctly (to 15 points). What I did not check, however, was the overhead totals for all votes, assuming that Excel was adding them correctly. As it turns out, I failed to notice that I had exceeded the boundaries of my SUM() function, which needed to be stretched an additional 2 rows to capture everyone’s votes. The good news is the winners are still the same. The bad news is the order for 2nd and 3rd has been reversed. Hali is now in second and Mark B is in third. Boy, do I feel awful about this. Next time I’ll be sure to have someone check my work before announcing the winners. I’m going to send Mark B something better than the headgear set to try and make up for it. As for the heavy weighting on story, yes it was very heavy, and no, we will probably not do it that way again. I made it very clear from the beginning that a good story would be the only way you could win this competition. A real problem that we did have was nepotism, however. There were 2 ballots that were blatantly favoring a particular film (1st in each category), and 1 that was strongly suspect (1st in 3 categories, 2nd in 1). It doesn’t look like these ballots had a lasting effect on the results, as none of the ‘favored’ films won anything, and the other films on the ballots seemed to be in order. It’s unfortunate, especially after the amazingly honest voting in the WOW competition. Now we’ll have to find a new way of scoring the films. It wouldn’t do for me to have to evaluate each ballot for fairness and reject those that I thought were unfair. The next contest is going to have a new dimension added to it, so the judging is going to have to be different anyway. Now it’s going to have to be very different. My apologies to Mark B and hali for screwing this up. I’ll do my best to make it right. Like I said, Mark will be getting something better than the headgear (which I want to keep for myself anyways ). Check the events page for the “real” scores. Geez, you’d think an engineer would be able to get simple addition right….
March 30, 2004 at 5:40 pm #38157
RevMen	I’ve updated the events page to reflect the correct numbers. Please, feel free to check them. As it turns out, other than Mark B and hali swapping places, none of the rankings have changed. It was a mistake that should not have been made, but at least it didn’t have far-reaching consequences.
March 30, 2004 at 6:15 pm #38161
boettcher	I want a recount of the first voting! I’m sure I had to be nominated …
March 30, 2004 at 6:19 pm #38162
LowweeK	Geez, you’d think an engineer would be able to get simple addition right…. I’m engineer too, and I’m the first to jump on my calculator to do an addition… Geez, where’s that world going to ? All this had somewhat a good effect, Hali gains one place… hum, no, I meant : we should have at least 2 guys to check the sum. On the other side, it could be interesting to do a simulation based on the rankings, in order to detect the overweight in a category… And why not, if the contests are being ambitious as it turns to be in the future, have several awards by category… like Oscars (link to an old topic…) :wink