Yo
Hey guys. Im Tony Cannon (Ponder), the Tournament Director and one of the founders of Evo. Thanks for supporting Evo in record numbers again this year. I know the experience was disappointing for many of you; especially those of you who made it to top-32.
Our goal at Evo is to hold a great tournament for the community. Ive seen reports that Evo is trying to exploit the community for profit, or skate by with just a good enough tournament. Over the next year, we aim to prove to you that this is absolutely not the case. We want everyone to leave the experience thinking wow, that was an awesome event! I cant wait to come next year!!. Obviously we did not achieve that this year.
But to hold a great Smash 4 event, we need your guys constructive feedback and cooperation. Ive read through all the feedback online and wanted to respond to your specific pain points, explain how they happened, and let you know how wed like to improve upon them for next year. In no particular order, heres what Im hearing from you about your biggest pain points. Please let me know if Ive missed any; Id be happy to respond to those, too.
The Convention Center Felt Cramped on Friday and Saturday
There was definitely a lack of space in the convention center at peak hours, which caught us by surprise. We believe we had booked more than enough room for the event. Evo 2015 was held in two ballrooms at the Ballys Event Center totaling 79,251 sq ft. We were greatly concerned that this would not be enough space for 2016. Evo attendance has historically grown by about 25% year-over-year. We also knew this would be the first year for Street Fighter V, which would certainly spur additional growth.
Last last year we cancelled the contract with the planned venue and booked the 150,000 sq ft venue in the Las Vegas Convention Center, nearly double the space from 2015. Unfortunately, the event grew faster than our most aggressive estimates. Registrations for the Street Fighter tournament grew by over 100% year over year, and Pokken (the first time at Evo) interest was much more than we anticipated. Pokken also required twice as much space as other games, further exasperating the space problem. Finally, over 4x as many people bought day-passes to Evo this year compared to 2015, greatly increasing the number of people actually in the event hall at a time.
The result was that the event felt cramped in spite of our best efforts. We are currently investigating ways to fix this for 2017, including booking a single hall larger than 150,000 sq ft, booking additional tournament halls as we did in 2015, or making better use of the space we have.
Viewing Experience in the Convention Center was Bad
Smash 4 did not have a projector for a majority of the tournament. This caused a lot of crowding around the tournament stations which made the Smash area feel more cramped than any other tournament. Even when Smash 4 was on the main screen, there was not enough seating for everyone who wanted to watch, resulting in crowding of the isles leading between the two tournament sections.
Many of these problems are caused by the aforementioned space problems, but even in a space-constrained environment we need to do better. For example, we could have setup a dedicated stream-viewing room for spectators in an adjacent hall or conference room.
One thing we certainly need to do for 2017 is to ensure all tournaments have a big-screen somewhere (ideally in the tournament hall, but perhaps in another room if were space constrained) where they can view their tournament. This is one of my highest priority issues to fix for next year. Everyone at the event should be able to watch their game of choice without having to restore to looking at a game station monitor.
Finally, games which had their finals in the Las Vegas Convention Center did not have the level of in-house production they had come to expect from Evo. The big screen was much smaller and the main stage much more modest. The seating area for the main stage was also smaller than what we had in 2015. Most of these changes were the result of increased complexity on the production side. This was the first year where Evo produced the Sunday finals in an arena setting and a live television broadcast, simultaneously. Rather than stretching ourselves too thin trying to do everything (and potentially failing at everything), we elected to focus on Sunday to make sure we did a good job there. In hindsight, that was a mistake. Sunday went off well, but peoples Friday and Saturday experiences were too greatly compromised. Next year, were going to increase the staff and budget for Tournament Hall production to ensure the level of spectacle is what you have come to expect from Evo.
Friday Night Pools Ran Late While Stations Were Available for Friendlies
This should not have happened. The plan going into Friday night was to use every available station to accelerate Round 2 brackets to finish the tournament as soon as possible. Friendlies were to be allowed only on stations which were not being used for the tournament. Ive spoken with Bear (the head TO for Smash 4) and have confirmed that they executed that plan to the best of their ability, and that he was personally unplugged joysticks of players who refused to stop their friendly match to make room for a tournament one.
In Bears opinion, the biggest reason we ran late was because of poor scheduling and delays caused by paper brackets. Round 2 matches were scheduled to start at 8:00 PM, immediately after the 6:00 PM Round 1 matches. This meant when Round 1 brackets ran late, a significant number of Round 2 brackets were delayed as well. The manual process by which we moved qualifiers from Round 1 into Round 2 delayed the start of some Round 2 brackets even further.
We certainly need to do a better job of scheduling. In hindsight, it would have been much better to start Round 2 on Saturday morning instead of Friday night. This would have given people who finished Round 1 pools early the rest of the day off and given the TOs time to propagate changes from Round 1 to Round 2 overnight. If Round 2 had started Saturday morning instead of Friday night, I dont think we would have had this problem.
In addition, we need to switch from paper brackets to a fully-online system to eliminate the delay in propagating players from one round to the next. We plan to do this for next year.
The Schedule for Event Competitors was Oppressive
Many Mortal Kombat XL and Smash 4 players were required to play their last match late into the night one day followed by an 8:00 am match the next. Thats just not acceptable and was not our intent. Had everything gone according to plan, both MK XL and Smash 4 would have finished by 10:00 PM and players would have had 10 hours of between matches.
Our biggest mistake here is that we need to do a better job of scheduling to make sure competitors have an ample opportunity to rest even if things do not go according to plan. Even if our schedule is running late, players should have at least 12 hours after they finish their games on one day before we ask them to play again the next. This is something we will absolutely fix for next year.
The Bracket Seeding System Was Limited and Cumbersome
This year we asked each fighting game community to provide the top 32 seeds for their tournament. Once the initial brackets were created, we put them up for public review and asked representatives from the community for feedback to make them better.
By putting the brackets up for review, the community caught an egregious bug in our software which resulted in double jeopardy in late rounds when creating brackets for tournaments featuring between 2048 and 4095 players. This affected both Smash 4 and Melee, and we were able to fix the bug and re-create the brackets before the tournament started.
We also asked for a swap-list, which was used to swap two players in the bracket to provide better structure (e.g. fixing seeding problems). The initial swap-list was limited to 32 players, but we continued to accept player swaps throughout the preview week before brackets were locked down. In total, we processed 97 swaps for the Smash 4 tournament, including swaps to realign the seeds after the Round 2 brackets were fixed.
By processing all 97 swaps volunteered by the community, we collectively were able to seed the bracket much deeper than we had originally requested and fix some issues which happened after seeding; including swapping Ally into a seeded position after we accidentally forgot to include him, and swapping PPMD out when we learned he wasnt going to attend.
Overall, the system we had in place had good results. After all the swaps are processed and the Round 2 issue were fixed, Bear and the community leaders he worked with were satisfied with the structure of the brackets.
The biggest problem here is that the swap system is painful. Its very, very painful, and the brackets dont actually look good until the very end of it. This makes it difficult for players to know exactly who their opponents will be, and for community to generate pre-EVO content and hype (e.g. top 32 fantasy brackets). Next year we want to revamp the system to take a much longer list of seeded players and streamline the bracket tweaking process so that the brackets look good from the very beginning.
Brackets Take Too Long to Update Online
Evo uses paper brackets to run the event (if you want to know why, scroll down). This means the online brackets are updated much much later than systems like smash.gg or challonge.com. This is bad for players who are scouting their next matchup and bad for people who are following the event; especially those watching on stream. We are planning an overhaul of our bracket making software to address some of the other issues (e.g. seeding beyond top 32). Adding realtime bracket updates is something we want to put in as well.
Best of 3 Is Meh. Best of 5 Would be Better. So would Sunday Finals.
There are 8 other games running at Evo besides Smash 4, and all of them need to make compromises to finish within the allotted time. Wed like to work with you to figure out what the best compromises are. For example, we want all Finals brackets to fit into about 2 hours. Smash 4 finals are currently bo3 until Winners finals to fit that schedule. Is that the correct solution? What about changing the stock/timer rules to make matches run faster? What about going to 3/5 and running some losers bracket matches off stage (but on a secondary screen)? What about only doing top 6 instead of top 8? You probably hate all those ideas. What are some ideas you dont hate which can help us have a good event and still meet our time constraints? We want to reach the best format for you guys that fits the constraints of the event. To do so, we need to establish a stronger line of communication between the Evo tournament staff and the Smash community leaders much farther in advance of the tournament.
Outro
Once again, thanks for coming out to Evo again. Hopefully this post has given you a greater understanding of how we want to make things better for next year. Wed like your help in making sure we do the right things for your community. Quite honestly, Tom, Joey and I are not Smash players, which makes it impossible for us to know how best to cater to your event without a lot of input from you. Until now, a lot of that communication has happened on social media (Reddit, Twitter, etc.) and has not been effective. Id like to find a better channel for us to work on to make sure we deliver an event youll be happy with. We can do this in whatever forum youre most comfortable with (Facebook, Discord, etc.), just let me know. Im looking forward to making Evo 2017 better for you guys.
Thanks,
Tony Cannon (Ponder) Evolution Tournament Director
Addendum: About those paper brackets
The Evo tournament software was originally written 15 years ago to automate the process of making paper brackets to run the event. In that era, running brackets on paper was the only way to run an event of Evos scale; smart-phones didnt exist, and even if they did internet in a hotel ballroom was extremely spotty.
As technology advanced (say, in 2008), we continued to use paper brackets. They were reliable, flexible, and we were absolutely, 100% sure we could finish the event with them. Hotel internet was spotty, and the risk of a loss of internet causing the tournament to grind to a complete halt was too great.
Around 2012, smart phones were prevalent enough and hotel internet reliable enough that it seemed possible to run a tournament online with a smartphone per bracket, but by now the tournament software was over 10 years old. Paper brackets were still working great, and the value from shifting to all online just wasnt there.
Fast-forwarding to 2016, Evo is finally at a scale where paper brackets are no longer a viable way to run the tournament. There are now hundreds of thousands of people potentially interested in following the tournament online in real-time and the manual process of moving players between rounds is starting to become a bottleneck. Changing Evo to use online brackets first (falling back to paper brackets only in extreme emergencies) is one of my biggest priorities for 2017.