Z-I-v Summer Contest 2021

So after doing some experimentation with the current system of calculating scores, I have found a flaw.

Finding the outliers in the current way is not effective because it does NOT take into account each voter's average assigned rating.

For example, voter A could rate simfiles according to the scale:
1- below average
2 - average
3- above average
4- very good
5- awesome
6- outstanding
7- literally amazing
8- flawless
9- unbelievable
10- the best simfile ever

But voter B could use a different scale:
1- unplayable
2- trash
3- horrible
4- very bad
5- not good
6- eh
7- okay
8- nice
9- fun!
10- spectacular

Voter A assigns an average vote of around 2.5, but Voter B assigns an average vote of around 7.5. So in this scenario, a 4 from voter 1 is above average and would actually be good for the recipients overall score, better than a 7 from voter B. But the 4 would get removed over the 7 in the current system of removing outliers.

The fix is to base the outliers on which vote has the greatest effect on that simfiles final score? This is mathematically equivalent to asking: which voter rated this simfile the worst/best compared to their average given vote?

The method of doing is what I have listed above, which is subtracting each voter's average given vote from each of their votes. This makes it so that way every voter's average is the same (in this case zero) so we CAN now compare votes for a certain simfile and remove the outliers.

Now you may understand this far, but still not see the issue with the way the Week 1 votes were calculated. Well let me just state some statistics about it.

23 of the 33 lower outliers that got removed can solely be attributed to one voter. That means that voter could have changed any on those votes by increasing them 1 point, or down my any number of points, and it would have no effect on the overall scores except their own because their voter average would change and they would get a different score for their own simfile. I hope you can see that this is an issue, because this voter's votes matter less overall.

Okay, that's all for now.

You all are going through so many hoops to defend someone's votes instead of having that person learn from the consequences of grading everyone poorly...

---

EDIT: I understand that the current system is not going to change for the first 10 weeks, but I just ask that in the future, the method of calculating final scores should be made transparent before the votes are released. That way these issues can more readily be addressed before results are released.

Anyway, as long as people don't intentionally inflate their votes to gain an advantage, there really shouldn't be any issues. The difference is negligible for most people.

I think I'm gonna leave it for now, but I hope this can be addressed later on.

Shock
Beast

https://zenius-i-vanisher.com/v5.2/viewsimfile.php?simfileid=46364

Challenge has 44.25 beats of Freeze Arrows.

3 5 10 11 15 edit 10

Every difficulty has shocks

---

BUBBLE BASS / YOOKiE
BPM: 75-150
3 6 11 13
Download here!

---

---

---

4 7 11 14

Download here!

---

Decided to pass for this week and take it easy while recovering

---

Pulling back on the difficulty with this one.

Spiritbox - Blessed Be (Nightcore)

Singles: 4 9 11 14 16

Link: https://zenius-i-vanisher.com/v5.2/viewsimfile.php?simfileid=46373

---

---

Yeah I think even Doubles exceeded 40 beats' worth of freeze arrows; enjoy.

---

i had a lot of fun charting this one :)

---

---

YEEEEES PAISLEY PARKS MUSIC

Is it too late to join? I've been busy adjusting to a new job, but I'd like to participate if I can find the time for it

You're in! Yeah, I'm not limiting when anyone can join, just as long as it's before the VIP Room

---

Repentance / Abel

4 8 10 14

https://zenius-i-vanisher.com/v5.2/viewsimfile.php?simfileid=46394