| |
Richard
Registered: Dec 2001
Posts: 619 |
Event id #2941 : Lockdown 2020 Fun Compo
As we already have a CSDB Emotional Music compo and there have also been some themed intros, etc. I thought it would be fun to launch a short mixed competition for the coders. The LOCKDOWN 2020 Fun Compo.
Compo rules:
No major strict rules. Basically create a fun short demo, intro or game based on COVID-19. The only major rule is that we don't want anything that would do other people harm, and make upsetting content. We want something to cheer us up during these hard times. So the program should be fun.
NO SEUCK, NO GAMEMAKER stuff. It has to be programmed in BASIC or Assembly!
Entries can be submitted directly on to CSDB, as .D64, .PRG or .TAP format. No size restriction :) I'll also be joining in the fun.
Voting will be based on an online voting form, commencing MAY 2020.
Have fun!
No prizes. It's a bit of fun. Also expect to see something from me during this time. |
|
... 19 posts hidden. Click here to view all posts....
|
| |
Richard
Registered: Dec 2001
Posts: 619 |
Okay. Duplicate removed. |
| |
CyberBrain
Administrator
Posts: 390 |
Richard, may I ask how you calculated the final scores?
It looks like the score is calculated as a sum of all the votes (since the scores are greater than 10, whereas an average would be between 1 and 10).
I remember that you told me over a PM that voting zero for your own production would not lead to downvoting yourself.
According to the voting rules on the voting website we had to vote 0 for our own production.
So if the final score really is calculated as a sum, rather than for example an average where the zero-scores have been excluded from the average, then voting zero for your own production will actually substantially downvote it.
(The reason it will downvote your own production, is that you give it 0, but you give all the other productions a score greater than 0).
If more than one person from your team votes, each of the team members will downvote your production.
If none of the people on your team votes, your production will not be downvoted at all.
Hopefully, this is not how it works. :)
In our case, 6 of us voted, which I hope doesn't mean that we accidentally massively downvoted ourselves :) (6/28 =~ 21% of the votes on our demo would be zero-votes.)
So how were the final results calculated and how were zero-votes handled? |
| |
chatGPZ
Registered: Dec 2001
Posts: 11100 |
While at it - explain how the score at csdb is calculated? =) |
| |
Ghostrider
Administrator
Posts: 43 |
=) Groepaz Only Perff knows that |
| |
chatGPZ
Registered: Dec 2001
Posts: 11100 |
CSDb Voter Pro 1.0 proves that wrong - but they dont explain it either, so it'd still be interesting :) |
| |
Richard
Registered: Dec 2001
Posts: 619 |
First of all, I checked for spoilt votes. There were two, votes that had not been counted and automatically binned. These had everything marked as 0, and their own group entries as 10. That was totally unacceptable.
The scores were all imported from the voting forms and saved on to an Excell spread sheet. I then added all of the total scores. Just like I did with the official SEUCK Compo. Then got the overall result.
That's the way I have always done a voting system. |
| |
CyberBrain
Administrator
Posts: 390 |
Thank you for your answer and sorry for delay! As I understand you, you didn't account for the zero-votes and just took the sum of every vote for each release.
Unfortunately that means that you calculated the scores such that anyone who voted and also participated in the compo will downvote their own release. I realize you didn't intend this :)
The reason a zero-vote leads to a downvote here is basically the same reason that a 1-vote is really bad and a 10-vote is pretty great - but 0 is even worse than 1.
Let me try to illustrate how zero-votes impacts the scores: (I only have data for our release, so that's what i'll be using)
According to your results, our average is: 182/28 = 6.5. But if you exclude our own 6 zero-votes from the average, our average is: 182/(28-6) = 8.27.
The last number would be the correct result based on your rules (everyone must vote 0 for their own release but these zero-votes won't lead to self-downvote), and the first number is what you presented as the final score (except you didn't divide by 28).
For fun we could try to scale the two averages back up to something that looks like a sum again, by multiplying them by 28: Your result: 6.5 * 28 = 182. A result with zero-votes excluded: 8.27 * 28 = 231.6.
So the zero-votes makes a big difference to the scores and they really must be accounted for if you don't want them to cause self-downvoting.
What I think you should have done was just to calculate the average with zero-votes excluded: result = sum_of_votes / (num_total_votes - num_zero_votes)
Yes, yes, I know this was just a small little fun compo, but still a kind of anti-climatic way of ending it, to have downvoted yourself so aggressively due to a calculation error and not really knowing what people have voted, imo.
I also think it's a sad precedence to set for future compos, that if you vote you'll be punished by downvoting yourself and if you don't vote you'll be ahead :)
My hope is that either the mistake is corrected or that you will at least release the number of zero-votes for each release, so we can calculate the results ourselves.
Anyway, this stuff is not life and death, so no matter what, thank you for arranging the compo! :) |
| |
chatGPZ
Registered: Dec 2001
Posts: 11100 |
The sane way is not to vote in a compo you took part in anyway. Also who cares :) |
| |
CyberBrain
Administrator
Posts: 390 |
Arh, damn, I didn't mean to sound obsessed. I think it's fair to point out, but maybe it could have been done in fewer words. I just wanted to explain it.
Reading my post again, I can see that I might have made it sound like a bigger issue and less friendly than I wanted it to.
Sorry, Richard, I should have done this in a PM. |
| |
Richard
Registered: Dec 2001
Posts: 619 |
Apology accepted. |
| Previous - 1 | 2 | 3 - Next |
