Official modeling contest by Paradox Interactive

[wallacet 336]

Hmm... well I definitely am listening. I tend to disagree that the "Featured" section is making it unfair. Some of you are claiming that being featured gives a big advantage, and yet so far being featured isn't equating to most subscribers.

2/3 of the most subscribed entries thus far aren't featured.

Despite claims that the most popular modellers would win (just a popularity contest) there are some previously unknown modellers who are getting a lot of great feedback, followers, and publicity. This is something we wanted to accomplish with this contest.

I understand the dismay of feeling that someone is getting an unfair advantage based on something you can't control (ratings, and therefore views from being Featured). But so far in this contest, I feel like every participant is getting a pretty equal chance at getting most subscribers. It's not perfect. But I don't think anybody is getting shafted because they didn't get upvotes and therefore featured.

There will be other contests. We will use a variety of different methods and processes for competitions. The fact is though, there is no perfect competition system that has absolutely no bias.

Please feel free to share what you think would be a better way to judge/vote future contests =)

 

Also, just for context: We have no control over Steam's Featured Section. And no matter what we do, hosting a contest solely on our forums (where we do have control) results in FAR less participation and viewers

Hmm... well I definitely am listening. I tend to disagree that the "Featured" section is making it unfair. Some of you are claiming that being featured gives a big advantage, and yet so far being featured isn't equating to most subscribers.

2/3 of the most subscribed entries thus far aren't featured.

Despite claims that the most popular modellers would win (just a popularity contest) there are some previously unknown modellers who are getting a lot of great feedback, followers, and publicity. This is something we wanted to accomplish with this contest.

I understand the dismay of feeling that someone is getting an unfair advantage based on something you can't control (ratings, and therefore views from being Featured). But so far in this contest, I feel like every participant is getting a pretty equal chance at getting most subscribers. It's not perfect. But I don't think anybody is getting shafted because they didn't get upvotes and therefore featured.

There will be other contests. We will use a variety of different methods and processes for competitions. The fact is though, there is no perfect competition system that has absolutely no bias.

Please feel free to share what you think would be a better way to judge/vote future contests =)

 

Also, just for context: We have no control over Steam's Featured Section. And no matter what we do, hosting a contest solely on our forums (where we do have control) results in FAR less participation and viewers

It looks like most contests will have to be done through Steam. And that's fine. One of my main gripes though is that basing the competition off of most subscriptions doesn't at all account for whether or not the asset looks good in game/has proper LOD/actually works. If you just see 5000 subscriptions vs. 4000 subscriptions and declare the 5000 subscriptions model the winner without making sure the creator spent the time to do it properly, it defeats the purpose of calling it the best model. I'm not saying it has happened or will, but you can photoshop screenshots if something doesn't look right

[Dubai Skyscraper 129]

Please feel free to share what you think would be a better way to judge/vote future contests =)

Thanks for taking time to browse though the forums here and read all our ramblings.

Maybe pick 5-10 of the most popular ones, grab a nice screen shot of each, send them in a CO wide email, and have them pick their top 3? Tally up the votes, and use yourself as a tie breaker.

I do agree that having an entry on the top page is a huge advantage. People will just download whatever is on there without looking though the other submissions. Its sort of a mob mentality.

Also, I noticed a few that don't appear to be the "worlds largest" of anything.

Anyways, its the awesome $600 USD graphics card that's up for grabs that's causing all this "upheaval". My GTX580 died last month, which pretty much puts an end to my C:SL playing days since I can't afford its replacement. I would love that fancy graphics card, but the fact that the contest was being decided based on subscribers kind of turned me off on the whole thing. I honestly though that Gula or someone else with thousands of subscribers would easily win.

Heh, I can't deny that I'd really like to call it my own... but I'm not broke at all, I could buy it on my own if I really wanted to. And it's not like my 780 wouldn't be enough anyway. So yes, of course I'm also speaking for myself here, but it's not just that. I like my own model but I got to admit there are a few better models in the contest, which deserve to win more than me. But above all I want it to be fair, so that we can look at the aftermath and clearly say: the best model did win.

Btw, what you said about "not the world's largest of anything" is correct too... I also have some concerns about the poly limit on some entries which used a large amount of pretty fancy props... but I put all this aside because the featured asset issue is so many times more concerning to me.

 

Now, I did some investigation about the top 9 "most popular assets of the week", which appear to be the featured assets. Here are two graphs showing the rating/sub/visitor distribution in total and per hour. I find it alarming that these seem to follow no clear pattern; you can't just say "asset_X got the most subs and therefor is ranked 1st". There must be some more complex formula behind all this, taking several numbers into account. The scaling doesn't appear to be linear either, and (even though that's just my speculation) there could be something like a "rating floor" which must be reached before an asset gets featured. Another hint that the whole system is not transparent.

Probably some of you are better with graphs, feel free to correct me.

  Edited July 17, 2015 by Dubai Skyscraper  

[Dubai Skyscraper 129]

Sry for the double post; didn't want to edit the 1st one becasue the images take a lot of space

Here is an analysis of the top subscribed assets so far. Could be that I missed one or two, but the most notable ones should be all included.

Again 2 graphs; total and per hour. Assets sorted after featured? (yes/no) and then by total subs. Again no clear pattern. Why is World's biggest Armilla being featured instead of World's largest World, which outscales in all fields both per hour and in total? This is just... random. Want to be featured? Good luck... because that's exactly what you will need.

Meanwhile, the asset that's been featured for the longest time has more than twice the visitors of the 2nd place. We know that visitors do not equal subs, but more visitors do mean more attention, which brings more subs.

  Edited July 17, 2015 by Dubai Skyscraper  

[Azurespecter 506]

Really good information there Dubai Skyscraper. Thanks for taking the time to formulate this in such a well thought out manner, I appreciate it. I agree, the method of featuring isn't as straightforward as I had previously thought. It's a fairly random amount of bias entered into the contest =/

But how to get around it? Steam Workshop is by far the largest community for the game, and yet I don't think there is a way to avoid the built in featuring system

  Edited July 18, 2015 by Azurespecter  

[Dubai Skyscraper 129]

Really good information there Dubai Skyscraper. Thanks for taking the time to formulate this in such a well thought out manner, I appreciate it. I agree, the method of featuring isn't as straightforward as I had previously thought. It's a fairly random amount of bias entered into the contest =/

But how to get around it? Steam Workshop is by far the largest community for the game, and yet I don't think there is a way to avoid the built in featuring system

Thanks for taking the time to read all of this, it's not like everyone in your situation would do the same. In fact, I bet most won't. As the host of this contest you are kind of in a dilemma. You want the contest to gain the highest possible attention of players, but you also want to satisfy the modelers who take part in it. I can't deny that Steam is the superior tool to cover basically the entire player base. But there's always something that can be improved, and I want to help you with that as much as I can

There could be quite a few ways to host future contests. I've been thinking about it last night and here are two options I consider to be the most viable, without leaving the Steam Environment. Ultimately you as the host have to decide if they are feasible and if the gained transparency justifies some extra work.

1) You could have all modelers send you their entry at a given deadline. Then you, Paradox, view them and ensure that all of them follow the rules of the contest (you would do this anyway). In the next step you could upload all entries on a neutral account of yours, reducing the randomness/bias of the contest quite a lot. Of course, this does not solve the issue of who gets featured and who not.

2) Following the same procedure like my first suggestion, but making the entries some kind of "free download bundle" (DLC?). Within a given timeframe players can download it, which automatically installs all of the contest entries in their game. This ensures they won't miss one. For transparency, there could be an extra menu/tab for that, which lists the entries only. I know that this requires some coding, but this seems reasonable to me. To determine the winner there would be plenty of options. Let's say the players can disable the assets individually, but not deinstall them one by one. You could have some kind of "tracker" which looks up which assets are being disabled the most. Each entry would have a score of 0 - 100% of players who keep it enabled. The winner is the one with the highest score. For transparency, these results could be updated like once per day on the Paradox forums.

These are just theories I came up with. I'm sure you guys at Paradox could come up with even better ones. Just do some brainstorming with your team, discuss the issues the community pointed out. Figure out what you can do, and what is too complicated. Especially my 2nd suggestion, although I believe is the most fair, could be a little tricky to implement, and the voting system might need some improvement.

  Edited July 18, 2015 by Dubai Skyscraper  

[Cool_Z 1,547]

Lolll... I still don't think that the fact an item is featured or not will strongly influence the results, but yeah the system is completely random

 

The gulliver park is now featured... 287 visits, 120 subscribers, not enough upvotes (means 20 or less I think).

 

[wallacet 336]

Even though we all have our disagreements and complaints about the voting system, let's still remember to be grateful for Paradox/CO giving us this opportunity to win free stuff and thankful that they will continue to hold contests. That's the most important thing! 

[Drosovila 86]

Even though we all have our disagreements and complaints about the voting system, let's still remember to be grateful for Paradox/CO giving us this opportunity to win free stuff and thankful that they will continue to hold contests. That's the most important thing! 

Exactly this. The contest is awesome and im so happy right now that I actually am doing quite good :) (Even if I missed the theme a bit...)

Oh and I have some IMO intersting stats for my skygarden hotel. Since ot got featured quite prominently, the number of visitors has increased a lot, while the ammount of new subscribers has decreased proportionally.

I think the "popularity contest" argument kinda shows in the 150 upvotes, I already got. 

Also note the 4 downvotes though. I think my only asset that has more is my first asset, a huge blue dickbutt statue...

And @Azurespecter, thanks for stopping by and reading our rambling. Really appreciate it

  Edited July 18, 2015 by Drosovila  

[Dubai Skyscraper 129]

Drosovila, would you mind sharing the sub/visitor/rating graph which can be seen only by the asset creator? I'd like to see it on July 22 / 23 (when the voting ends).

Btw, I got 2 downvotes (5.7% of the toal votes) too, which is a lot for me. 

  Edited July 18, 2015 by Dubai Skyscraper  

[Drosovila 86]

Drosovila, would you mind sharing the sub/visitor/rating graph which can be seen only by the asset creator? I'd like to see it on July 22 / 23 (when the voting ends).

Btw, I got 2 downvotes (5.7% of the toal votes) too, which is a lot for me. 

Oh, I actually wanted to add that, but forgot about it. I'l edit it into the other post

[Dubai Skyscraper 129]

Even though we all have our disagreements and complaints about the voting system, let's still remember to be grateful for Paradox/CO giving us this opportunity to win free stuff and thankful that they will continue to hold contests. That's the most important thing! 

Exactly this. The contest is awesome and im so happy right now that I actually am doing quite good :) (Even if I missed the theme a bit...)

Oh and I have some IMO intersting stats for my skygarden hotel. Since ot got featured quite prominently, the number of visitors has increased a lot, while the ammount of new subscribers has decreased proportionally.

I think the "popularity contest" argument kinda shows in the 150 upvotes, I already got. 

That seems to be true... on the first sight. But if you have a closer look and compare featured with non featured assets, the picture changes. Here are two of my non featured assets, Dubai Wheel and Growable Residential 01:

As you can see the, the subs and visitors raise to a certain niveau and hold it, with small fluctuations.

Now, here is my Growable Office 03 - featured for 2 - 3 days now:

Notice how the views and subs get a significant boost as soon as the asset is being featured

To draw a comparision to your contest entry, it got the same amount of attention as the increase of views prove. But we both know it is a pretty special design which does not appeal to everyone. Therefor the boost for your asset is not as significant as for my Office 03.

  Edited July 18, 2015 by Dubai Skyscraper  

[Mr_Maison 2,426]

I think the Workshop should be one PART of the judging process along with a panel of judges as stated earlier. I like it better when knowledgable judges have the last say rather than a voting system. I've seen contests where talent meant nothing because one contestant had more votes beating the talented one who earned  perfect scores from judges. The popularity thing is usually used for marketing analysis. It should not be used to determine talent and whether something is done right.

[Shelltoe 15]

dubais idea is pretty great actually. CO would make sure the models meets the minimum requirements and the public can vote.

after all i think it's a creativity contest. if someone just started to modell you can't expact him to create perfect models like dubai or zed, they should still have a change to win from a creative point of view.

i'd always prefer a public vote over a few judges for this kind of contest

  Edited July 20, 2015 by Shelltoe  

[Cool_Z 1,547]

dubais ideas is pretty great actually. CO would make sure the models meets the minimum requirements and the public can vote.

after all i think it's a creativity contest. if someone just started to modell you can't expact him to create perfect models like dubai or zed, they should still have a change to win from a creative point of view

And so they might get help fixing what was wrong 

 

(some workshop posters still propose broken things and won't correct them... it would be nice that a contest helps people understand how to build properly)

[Azurespecter 506]

I'm wondering if hidden uploads to the Workshop can still be accessed via direct link... If so, I like the idea of having entries submitted to Paradox for review, and then we upload them (as hidden) and then put all the links in a forum thread. Pretty sure that hidden uploads can't be featured, but I don't think a hidden upload has ever been popular enough so who knows? haha

Cursed Steam and your bias! ruining our contests!

Now, one of the things I really like about this contest is that I feel a lot of less-well-known modellers, and new modellers, were encouraged to give it a shot. Having more restrictions and quality control can intimidate people into entering contests like this. Even though there are a lot of models in the contest that don't really belong, or aren't super high quality, maybe it's the first stepping stone for a modeller to get better?

That was one of our goals for this contest, was to make it widely accessible and available to as many people as possible. The fact that 3 entries got featured shows that the contest got a lot of popularity.

However, I'll repeat myself and say again that we will have other contests that will cater to more professional modellers, and even to more applicable themes =)

[Dubai Skyscraper 129]

I think the Workshop should be one PART of the judging process along with a panel of judges as stated earlier. I like it better when knowledgable judges have the last say rather than a voting system. I've seen contests where talent meant nothing because one contestant had more votes beating the talented one who earned  perfect scores from judges. The popularity thing is usually used for marketing analysis. It should not be used to determine talent and whether something is done right.

We used to have design contests at Skyscrapercity. The winner was determined by a simple public poll that listed all entries with 1 - 3 images for each. I think the best designs did always win, because the system was entirely transparent  I didn't win a single one (only took part in one or two though), but I never felt "screwed" or in disadvantage to the other contestants. Sure, our community was much smaller than the C:S playerbase on Steam, and most of us had what one would call "acquired taste". Still, I think that the public voting system is not to be blamed in general. I don't know much about the average C:S player but it's not a typical game for kids. I assume most players are at least in their mid-teens and therefor I assume they got a reasonable portion of taste 

 

I'm wondering if hidden uploads to the Workshop can still be accessed via direct link... If so, I like the idea of having entries submitted to Paradox for review, and then we upload them (as hidden) and then put all the links in a forum thread. Pretty sure that hidden uploads can't be featured, but I don't think a hidden upload has ever been popular enough so who knows? haha

Cursed Steam and your bias! ruining our contests!

Now, one of the things I really like about this contest is that I feel a lot of less-well-known modellers, and new modellers, were encouraged to give it a shot. Having more restrictions and quality control can intimidate people into entering contests like this. Even though there are a lot of models in the contest that don't really belong, or aren't super high quality, maybe it's the first stepping stone for a modeller to get better?

That was one of our goals for this contest, was to make it widely accessible and available to as many people as possible. The fact that 3 entries got featured shows that the contest got a lot of popularity.

However, I'll repeat myself and say again that we will have other contests that will cater to more professional modellers, and even to more applicable themes =)

Azure, remember when I could not make my models public anymore (the reason why I left for 6 weeks or so)? No one could access it, not even you with workshop privileges  Only people on my friends list were able to view and download it.

I too think that the contest attracted many new (and actually good) modelers though that was probably mostly for the awesome 1st prize  (Of course this is an absolutely legit reason to join)

[Dubai Skyscraper 129]

(Sry for the double post; again, I'm trying to make my posts not too long)

Just in case I didn't convince everyone about how important it is to get featured, here is a final graph to prove my point:

It's my Office 03 again. Notice how the views and subs collapsed immediately after it wasn't featured anymore  

  Edited July 20, 2015 by Dubai Skyscraper  

[Mr_Maison 2,426]

Yesterday I must of spent a half hour subscribing to some commercial buildings to spice up my city. They look so great in the screenshots on the workshop but I ended up chucking about 75% of the buildings because I didn't like them once I saw them in game. Subscriptions do not = likes. I would say ratings should be more important but then it's not a perfect world where people would try things out first then go back and rate.

But I'm glad this contest happened. I hope more building modeling contests are coming up because I found out for myself yesterday that there's a lot of crap (IMO) on the workshop with some high ratings and 5 stars. Contests like this will push better stuff out like it did the past week.

  Edited July 21, 2015 by Mr_Maison  

[Shelltoe 15]

yeah the old problem. there are 50k entries in the workshop right... only 950 mention the word LOD (including "no lod included" )

will require some time to change this.

 

  Edited July 21, 2015 by Shelltoe  

[Dubai Skyscraper 129]

Yesterday I must of spent a half hour subscribing to some commercial buildings to spice up my city. They look so great in the screenshots on the workshop but I ended up chucking about 75% of the buildings because I didn't like them once I saw them in game. Subscriptions do not = likes. I would say ratings should be more important but then it's not a perfect world where people would try things out first then go back and rate.

But I'm glad this contest happened. I hope more building modeling contests are coming up because I found out for myself yesterday that there's a lot of crap (IMO) on the workshop with some high ratings and 5 stars. Contests like this will push better stuff out like it did the past week.

That's true but at least we can assume that "bad" contest entries would be sorted out by most players in the given timeframe. A rating contest is always better but as of now it's impossible to do in the workshop. A simple thumb up/down system is not designed for that. And even worse, the 5-star rating is based on the number of votes, total nonsense.

yeah the old problem. there are 50k entries in the workshop right... only 950 mention the word LOD (including "no lod included" ;))

 

Well you will never find my models this way, simply because I don't mention they have LODs. A custom LOD is something so basic and crucial, I assume every model has one until proven otherwhise.

[Shelltoe 15]

i assume the opposite until proven otherwise

EDIT: actually i find some of your buildings, cause of your changelog

eg:

- new LOD

 

  Edited July 21, 2015 by Shelltoe  

[Cool_Z 1,547]

Same for me, unless there is a lod mentioned I assume there is none.

[merc187 46]

I'm just now learning how to do this entire modeling thing the past few months and I didn't realize until recently how important LOD's were to make your building really seamless in the game.  I'm going to make it a project to go back to all the shitty ones I've made so far and make sure they have the LOD's included.

[Cool_Z 1,547]

I still don't understand how some people can do such good work on their models and not care about the lod.

And (I'm not the only one) how few people react about it when it's missing, or very bad.

 

I mean... some of the most downloaded (and sometimes featured, not only in the 9 top of the week but in the overall top) workshop items are ok from up close but will kill every wide shot of a city.

[Shelltoe 15]

I still don't understand how some people can do such good work on their models and not care about the lod.

yeah but you have to consider, lots of those almost great assets aren't actually original models, but extracted from other games or downloaded somewhere. (so many GTA cars in the workshop)

those models won't have a lod of course and the creator might not be able to create one even if he tries to.

 

imho whats even more disturbing are ppl with actual knowledge creating vehicles with 13k polys and lods with 1k+ polys

  Edited July 24, 2015 by Shelltoe  

[Shelltoe 15]

#1 - Skygarden Hotel (AKA Worlds Largest Tree) (6676 subscriptions)

#2 - World's Biggest World (6,202 subscriptions)

#3 - World's Tallest Rollercoaster (4932 subscriptions)

#4 Runner Up - Ulm Minster (4709 subscriptions)

#5 Runnner Up - World's Largest Water Tower

https://forum.paradoxplaza.com/forum/index.php?threads/tempted-worlds-biggest-winners.872592/

most assets crushed the 4k tris rule and had a bad lod.

actually models not fitting the theme was smallest problem for this contest imho.

 

  Edited July 30, 2015 by Shelltoe  

[Cool_Z 1,547]

 

most assets crushed the 4k tris rule and had a bad lod.

actually models not fitting the theme was smallest problem for this contest imho.

 

You don't say !

Some of the entries had ok lods but most look VERY bad when zoomed out 

[wallacet 336]

 

most assets crushed the 4k tris rule and had a bad lod.

actually models not fitting the theme was smallest problem for this contest imho.

 

You don't say !

Some of the entries had ok lods but most look VERY bad when zoomed out

I think we all complained enough about there being no testing requirements that next time there will be something like that....

 

I hope

  Edited July 30, 2015 by wallacet  

[Dubai Skyscraper 129]

 

I know the game is scaled down, but I would've expected the "worlds largest tree" to at least be similar in scale to the massive Tokyo Skytree.

 

 

Actually it's not. People, cars, streets... all of this is scaled 1:1. The buildings are fine too, it's only that 4x4 is too small for real highrises, so the game has only mid-rise growables. Many people seem to confuse them for highrises and then called 1:1 scaled skyscrapers "too tall" for the game... which is bullsh*t

Sure, public facilities like schools, hospitals etc. are too small for the number of people they can serve. But that's a gameplay decision.

  Edited July 30, 2015 by Dubai Skyscraper