3 winners, yeah...but where is the ranking of all the others???

[sjm]

I'm the other person who voted for Flame.

[tanabarbier]

Me too! And I use it in my new live-set too, I really like it and look forward to see where it is gonna go!

[Caco]

I'm shocked, Flame finished very low in the voting and nobody really mentioned it. I am really pleased some people are using it

[highkoo]

I thought Flame was the beginning of something great.
Good stuff, just not quite there in time for the DC maybe.

I think the center panel needs a GUI "concept" for the routing. That is the key.
Something to make the routing itself easier to see and more intuitive, and a way for the routings to visually tie to the units on the perimeter.

[sinkmusic]

I'm glad to see Transient and Space360 so close to the top of the list.
I'm surprised to see Flame and Tapatiki not as well ranked as i'd hoped...
Anyway, thanks and congratulations to all the entrants !

[AXP]

I've found Azurite done very well and voted high for it. Good job!9

[Ichad.c]

If the numbers are correct, out of the 15 votes available to each member, the average member only cast 3.6 votes.

Someone please check my arithmetic.

And if I'm right - what was going on?

Hmm... I checked the math too - very odd indeed:

4,319 votes were cast by 1,198 members.

If 1,198 members voted - and used all 5 slots - it would equal 5,990 votes. That's 5,990 - 4,319 = 1671 missing votes! Statistically - if a member votes only for 1 plug-in - then that top 5 points - is actually more points on average - because the user's last 10 points (slots 2 to 4) - is not distributed into the total sum. In musical terms - it can cause a "peak analysis" and not an "averaging analysis" - in simpler terms - it can cook the system.

Not that I'm b**ching - but something that might be needed to be taken into consideration for DC'13.

[AXP]

If 4319 is the number of votes then it's not as bad as if it's the number of points.

[Ichad.c]

If 4319 is the number of votes then it's not as bad as if it's the number of points.

The number of points could be anywhere from ~1,671 to ~4,177.5 that is "un-accounted" for.

*~1,671 - if only the last slot is missing - which is untrue - only 1,198 members voted.

*~4,177.5 - worst case scenario - some members only used the top slot - every 4 slots (slots2->5) would then equal 10 points - so:
(1671(votes) / 4(slots))*10(points) = 4177.5.

Sounds worse in points.

Except if I messed up the math

If the total points are available - the total "peak vs. average" could be roughly estimated.

*Edit* if the 1,198 members that voted - used all their slots -> the total points would be 17,970.

[Caco]

I thought Flame was the beginning of something great.
Good stuff, just not quite there in time for the DC maybe.

I think the center panel needs a GUI "concept" for the routing. That is the key.
Something to make the routing itself easier to see and more intuitive, and a way for the routings to visually tie to the units on the perimeter.

Yes, good points highkoo, perhaps the the routing needs to be simplified or more user-friendly for future versions. Will have a think about that.

[Compyfox]

4,319 votes were cast by 1,198 members.

If 1,198 members voted - and used all 5 slots - it would equal 5,990 votes. That's 5,990 - 4,319 = 1671 missing votes! Statistically - if a member votes only for 1 plug-in - then that top 5 points - is actually more points on average - because the user's last 10 points (slots 2 to 4) - is not distributed into the total sum. In musical terms - it can cause a "peak analysis" and not an "averaging analysis" - in simpler terms - it can cook the system.

Not that I'm b**ching - but something that might be needed to be taken into consideration for DC'13.

This kind of confirms my fear with the symphathy/follower/fan votes. I even remember that just a couple of minutes (hours) after the DC start, that certain people already "made up their mind".

This needs(!) to be adressed by the next Developer Challenge. And I'm happy that I got a certain backup by users like highkoo.

Again, I do not(!) hold a grudge towards anyone of the devs - but the (official) posted results literally scream "something went wrong!".



I mean, has anyone even really tested Christian Budde's tool or actually wrote him (placed 35th place!) as he asked to (feedback)? Granted, the DC entry was buggy and basically abandonned right from the start. But it was just as revolutionary as Priximity.

Or what about rsmus7's preset pack (placed 37th)? Did someone even take a closer look or did the people just vote 1 point just because it's there?


Overall I think that both Proximity and Space 360 are among the most revolutionary tools in the DC this year. The Chorus was something different as well, Philteroid can be massively enhanced, Sleepy Time Records already works on upgraded versions of his tools as well. Flame was awesome (I liked the simplicity) but had nasty bugs (noisees) while being idle, Radio Junk suffered issues from the Nebula Engine ACQUA, ...


We definitely need new rules. And/Or maybe different challenges to keep KVR interesting and less chaotic or predictable than this year.


Example:
We have a ton of sound designers on KVR. Instead of pulling of a DC every 2-4 years with now added sound designers, why not have a yearly Sound Designer Challenge. Rule is to provide samples in most common formats (SFZ, Kontakt, WAV, REX) or for certain selected synths and no crazy players. We would then have two categories: Samples and Synths. People can burn 5 votes again, we have that out of the DC and people can actually focus on the provided content.

Then we could pull off a DC bi-yearly (read: every two years), along with Windows, Macintosh, Linux, Portable Devices (which will be more interesting the next years).

Additional to that, we have the "One Synth" challenges, currently no prices.

Recently, there was this "Mastering Challenge", which could be pulled off once a quarterly period (once every 3 months) or better every half year - with a small price fund maybe.


We have ton of possibilties, we just need to finetune certain rules first. And one of them is to somewhat rule out symphathy votes.

I usually don't like such ideas, since I had bad experience with it, but I think we need a judges panel that has the power to shift votes. And by that, I don't mean moderators and admins of KVR. Though here I'm just throwing around ideas.

[Ichad.c]

Again, I do not(!) hold a grudge towards anyone of the devs...

Me, neither!!!

We definitely need new rules.

Yeah, my calculations give a worst case scenario error-rate of ~23%, that's a bit much.

I can suggest 2 rules that *might* lessen some irregularities:

1) Use all 5 slots or the votes do not count - this will encourage people to test properly. This will give a more averaged view of plug-ins tested, because as I said before - if you only vote for the top1 - that 5points actually count more - than if you used all slots.

2) A 1 week "cooling off" period for voting, i.e. only open voting after a week has past, this will also encourage people to test properly, and diminish "spontaneous voting".

I usually don't like such ideas, since I had bad experience with it, but I think we need a judges panel that has the power to shift votes. And by that, I don't mean moderators and admins of KVR.

That won't work - everybody is biased (including you and me) - to a certain extent.

Andrew

[bk]

I can suggest 2 rules that *might* lessen some irregularities:

1) Use all 5 slots or the votes do not count - this will encourage people to test properly. This will give a more averaged view of plug-ins tested, because as I said before - if you only vote for the top1 - that 5points actually count more - than if you used all slots.

2) A 1 week "cooling off" period for voting, i.e. only open voting after a week has past, this will also encourage people to test properly, and diminish "spontaneous voting".

Those are great ideas, and should be implemented next time! In fact, I'd say two weeks cooling off period might be appropriate. I found it hard to adequately test that many plugins in a month, let alone a week!

[Syntilla]

Simpler still ... (I like simple) ...

Leave DC exactly as it is.
Free software for all.
A month's 'interaction' with the devs and general discussion for all members.

Except:

(It ain't gonna happen, but I'd like to know your theories about why.)

Voting costs $15.

To be allocated to the devs of your choice.

Would anyone honestly say their favourite plug in the competition isn't worth $5? Then it can't really be worth 5 votes/points.

(Every time Bootsie releases anything, someone always posts on his site that it's worth $100s - and I sometimes even agree with them. Be interesting to see how many would actually pay even $5 for it.)

If only a quarter of the people who voted this time do so next year, the prize fund would be bigger.

There would still be a 'winner' (the one with the most money).

And all the runners-up and also-rans would get their fair share of the spoils. (And a pretty good idea of what people really think of their entry!)

Nothing says 'I love you' quite like cold, hard cash.

[Syntilla]

My first impression when I saw this plug was: "...and we have the winner!".

Me too.

A 1 week "cooling off" period for voting, i.e. only open voting after a week has past, this will also encourage people to test properly, and diminish "spontaneous voting".

Then I heard it.