The Dota 2 Polycount Contest began on October 8th and officially ended this past Monday December 3rd. Over that two month period the Polycount modeling community submitted more than 1,000 entries to the Dota 2 Workshop. We aren’t exaggerating when we say that the sheer magnitude of high quality submissions has blown us away and that judging entries has been difficult because so many sets could have been winners. We have no doubt that a very large number (if not a large majority) of the sets contributed in the course of the contest will eventually be incorporated into the game. If your favorite entry didn’t win, don’t despair. There are enough great items here that we will see entries ship for the foreseeable future.
Incorporating feedback into a design is an important part of game development that a contest can’t easily capture. Many of the contest submissions demonstrate this challenge with items that are 95% of the way to excellence, but need a few rounds of adjustments to be true winners. Some common flaws we discovered while judging were:
- An overabundance of detail in small areas. In Dota, patches of small detail can look good in the profile view of a hero, but tend to become hard to read during the normal course of the game. Where and how to use detail, what color values, levels of contrast to apply on that detail are critical considerations when creating items for Dota.
- Not taking advantage of value gradients that could enhance a hero’s readability. If the brightness and tone of the colors at a hero’s feet are as intense as those at the head, the character can become difficult for the eye to break apart at a glance. More on this in our Dota 2 art guide.
- Some sets are extremely well executed and quite beautiful, but at the same time don’t push on the character design enough. This is always in a delicate balance. Items should be recognizable as unique and enhance the hero’s identity without being so radical as to shift the lore into too alien a territory.
Winners and losers can’t be judged by the Workshop presentation alone. Many of these flaws only became apparent when we examined the items closely in game, under real game lighting and motion. As a result, some of our picks may surprise you. When relevant we will describe the tough decisions we faced.
Moving forward, we want to create an open discussion with the community about the techniques that enhance items to top tier status. To the extent that we can help teach, we will. To the extent that we can improve our tools to make it easier to iterate, we will. Initially this may manifest as blog posts from the art team, but longer term we expect to see Workshop features aimed at making iteration easier and technical feedback more visible to the community.
We will announce 3 winners today, 3 Wednesday, 4 Thursday, with the grand winner being unveiled on Friday. Trying to rank the top ten sets in order isn’t our goal. They do not all succeed or fail in the same way. All winners will be available for purchase or as drops in next week’s update.
Our selection of winners, then, is about capturing the contest entries that we feel represent the highest quality on all possible dimensions. In order to recognize the broadest range of set designs, we chose to not select two winners for the same hero. This doesn’t mean these other sets won’t be sold in the Dota store or appear as drops in future updates. Expect to see more sets ship in January.
Let’s get on to the first set of winners.
Our second winner…
And today’s last winner…
Remember to check back tomorrow for another set of winners!