One Synth Challenge #159: Surge XT (j5v Wins!)

[Juergen]

Voted!
Good luck everyone.

[rilderec]

(Reverso traduction)
I put a lot of 1 and 2, a little less than 3, 4 and finally very little 5. This time I only put a track with 5 points.
It’s the best way to bring your favorite songs to the top of the leaderboard.

There shouldn't be anything wrong with that but realistically someone will see that only you gave them a 1 and remember next month when voting. Sorry to suggest this negative thing. I like to assume the best of people. Perhaps electronic music makers are better than most other people.

It’s wrong, people who vote "revenge" are not good people.
Some people commented on me suggesting that they find some of my leads bad, but that didn’t stop me from giving them good grades when I found their sons good.
Or the other way around.

[Double Tap]

(I know this because I copy my notes into a .txt file before voting and keep this file)
It would be preferable for the hearings to be public.

The votes are public - look at the results post and at the bottom you'll see a link to a Google Doc which shows the votes that have been cast. There is even a 'generosity scale' which shows whether people gave more generous scores or harsher scores.

[emikaela]

(Reverso traduction)
I put a lot of 1 and 2, a little less than 3, 4 and finally very little 5. This time I only put a track with 5 points.
It’s the best way to bring your favorite songs to the top of the leaderboard.

I thought the Idea is to judge tracks by their merits, rather than try and push one's favorites. But maybe I'm wrong and will have to adjust scoring for your track.

(reverso traduction)
Judging the songs "on merit" and "pushing your favorites" is the same thing. We vote for the songs we prefer and not for anything else.
...
(I know this because I copy my notes into a .txt file before voting and keep this file)
It would be preferable for the hearings to be public.
...

votes are made public after the round ends! (i think this is what you're asking for? translation is weird)

here are the votes from #158 for instance: https://docs.google.com/spreadsheets/d/ ... edit#gid=0

if the entire point was to maximize the chance of your favorites getting to the top, it would indeed be the best strategy to give a few very high scores and otherwise all very low. i think the fact that we have more than two scores to choose between indicates that this is not the entire point.

people have their different methods of scoring, but it seems most will interpret a low score as the track being of low quality (or at least below average).

a score of 1 suggests to me that not only is there something seriously wrong with the track, but that you have now actively made the world a worse place by creating it and you should be ashamed of yourself. slight hyperbole but yeah, i don't give out a lot of 1s. good to know that this isn't what you mean by it!

[baconpaul]

a score of 1 suggests to me that not only is there something seriously wrong with the track, but that you have now actively made the world a worse place by creating it and you should be ashamed of yourself.

See now I’m interested in a different type of one synth challenge where the idea is use whatever you want but make tracks that everyone scores as a “1”

[emikaela]

a score of 1 suggests to me that not only is there something seriously wrong with the track, but that you have now actively made the world a worse place by creating it and you should be ashamed of yourself.

See now I’m interested in a different type of one synth challenge where the idea is use whatever you want but make tracks that everyone scores as a “1”

what if we just start giving out a prize to the lowest scoring entry here? omaha high/low poker style, sort of

[rilderec]

@emikaela
"votes are made public after the round ends! (i think this is what you're asking for? translation is weird)"

Okay. I wasn’t paying attention, but you can’t see who’s voting who, I wouldn’t mind.

I don’t think score 1 means that the track is "bad", for me it’s just that I don’t feel much emotion in that track or that I don’t like the style.

"if the entire point was to maximize the chance of your favorites getting to the top"

My goal is not only to maximize that my favourites reach the top, my goal is to transcribe by my vote my feelings on the tracks I listened to.
I’ve listened to OSC twice each time, the first time I listen, I find my favorites and I rate them at 4, the second I distribute them according to my tastes, and by replaying my favorites, I put 5 on my top.
But once I didn’t put any 5, for example at OSC #158

[FrogsInPants]

I try to divide the entries into groups of roughly similar quality, and don't often have much use for the lower point bins. If I would rate a track significantly below the average for the month, but there are enough people rating it highly for it to make a difference whether I rated it a 3, 2, or 1, it must have some merit that I do not recognize. I'm not so invested in my picks winning that I will be upset if a piece I didn't care for as much pulls ahead. Intentionally rating most pieces lower so that your votes for the pieces you find outstanding count more than others' votes strikes me as immodest. Also, giving mostly low scores and few or no 5s looks like strategic voting to raise one's own chances. As long as it's consistent and not actually strategic, I support people voting their conscience according to whatever criteria they use, but I always look askance at the people who give scores substantially below the average.

[FrogsInPants]

@emikaela
"votes are made public after the round ends! (i think this is what you're asking for? translation is weird)"

Okay. I wasn’t paying attention, but you can’t see who’s voting who, I wouldn’t mind.

Look at the bottom of the score page. The spreadsheet also has a "Data" tab, which shows each individual score given.

[rilderec]

@FrogsInPants

No, since we score ourself with 1 point, at the end it’s an average, no matter how we votes, so that’s fair. Most tracks I note are between 1,2,3

Ah ok thank, I had never seen that there was this "data" tab, and so much the better, so that people can see those who note with 4 or 5 and those who are wider.

[J.Ruegg]

Voted

[diaseis]

@FrogsInPants

No, since we score ourself with 1 point, at the end it’s an average, no matter how we votes, so that’s fair. Most tracks I note are between 1,2,3

Ah ok thank, I had never seen that there was this "data" tab, and so much the better, so that people can see those who note with 4 or 5 and those who are wider.

Actually mathematically, if you score lower than average, you do increase your chances of winning, especially on instances with not a lot of participants. Here is an example : let's say there are 9 participants (except you) who give an average of 4 points (including the 1 point self rating). Let's now say you give an average of 2 points (including the 1 point self rating). The average rating on you will be (4*9)/9 = 4.0 and average rating on other people than you will be (2 + 4*8)/9 = 3.78

You can generalize this example to some mathematic proposition that says essentially that if you rate under average, you WILL get some statistical advantage (the opposite is also true, but I've always rated way higher than average hhahahahah ).

[EDIT : a few [EDIT a lot of] math mistakes ]

[diaseis]

Also I have voted, congrats to everybody !

[bh9090]

Voted. Musicality and creativity are at an all-time high this month. Tons of impressive work.

[Double Tap]

@FrogsInPants

No, since we score ourself with 1 point, at the end it’s an average, no matter how we votes, so that’s fair. Most tracks I note are between 1,2,3

Ah ok thank, I had never seen that there was this "data" tab, and so much the better, so that people can see those who note with 4 or 5 and those who are wider.

Actually mathematically, if you score lower than average, you do increase your chances of winning, especially on instances with not a lot of participants. Here is an example : let's say there are 9 participants (except you) who give an average of 4 points (including the 1 point self rating). Let's now say you give an average of 2 points (including the 1 point self rating). This means that in total, 38 points where given. So average rating on all participants is (4*9)/10 = 3.6 . However, average rating on you will be (4*9)/9 = 4.0 and average rating on other people than you will be (2 + 4*8)/10 = 3.4

You can generalize this example to some mathematic proposition that says essentially that if you rate under average, you WILL get some statistical advantage (the opposite is also true, but I've always rated way higher than average hhahahahah ).

Great analysis

What I find interesting about the scoring is that usually the winner gets more 5s than everyone else, and in fact you can usually rank the top 10 or so according to the number of 5s they get. Occasionally someone will get an unusually high number of 4s which pushes them up a place or two, but that's quite rare.

Which tells me that the voting system is working pretty effectively at discovering which track the most people here like the most. So I don't think it matters all that much.