Designing Draft Champions

I’ve been meaning to do a design retrospective on my features for some time now.  I’ve just been terrible about documenting my process aside from my written notes, and I think it will be a worthwhile exercise to see how the mode started and evolved throughout our development cycle.   Also, I had a lengthy discussion on Twitter the other day regarding providing more information on the development process.  So here we are.  Since Draft Champions is the latest feature I’ve worked on, and it provided some interesting challenges, I’ve decided to start there.

The High Level

I was approached with designing some type of draft experience that would complement the MUT (Madden Ultimate Team) experience in Madden.  Aside from that, I was given little direction which was rather refreshing.  If you want a high level breakdown, you can find it here

Previous Failure

One thing I did discover is that previously, unknown to me as I was buried in Franchise mode at the time, MUT had already tried to do a “draft” mode called Draft Duels that did not have the greatest launch.  The first thing I did was dig into that mode and try to understand what went wrong.  The last thing I wanted was history to repeat itself.  This helped guide what the pillars were for Draft Champions.

Discovering the Pillars

Very early on there was a lot of discussion around building the team, how the choices would be made, how many cards to offer at once, how many rounds, etc.  I mean, a TON of discussion.  Since everything was technically on the table, a lot of input was suggested regarding these boundaries.  I’ve learned that it’s always good to work within boundaries.  Be creative within the box.  With that in mind, a few pillars were established.

·         It had to be a fast, strategic, asynchronous draft experience.

·         Balanced online H2H experience

·         Clear engagement loops for users with a low barrier to entry.

·         Highlight some of the best UT players

MUT is meant to be a longer game experience, and as a result it can take a while to obtain the best players.  We had set the ideal draft time for just a few minutes, with the entire experience including multiple games, ideally completed in an evening.  Thanks to some research we knew we had around two to three hours per night for the average user, and wanted to fit within that window.  That meant trying to make a draft and the max number of games fit within that window. 

At the same time, it had to be strategic.  I cannot convey how much of a battle it was to keep the number of picks below the max number of positions available on the field.  This is something that I discussed, argued, and battled for the entire development cycle.  The holes that having fewer rounds than positions needed creates a sense of urgency and makes each selection even more critical.  It also forces the user into a decision as to how they want to play for the next few games.

We also wanted to showcase the best players that are available in UT while at the same time keeping the playing field balanced for H2H competition.  There are very few players out there that get to enjoy playing with some of the amazing items in UT.  We wanted to provide them with the opportunity to quickly obtain some of the best players, but were confident that the shorter game experience would keep it balanced.  Also, the randomness of Drafting and different strategies will make for different H2H matchups.  No one wants to play against the Seahawks or Patriots every game.

One thing I learned while working on Dynasty in NCAA and CFM in Madden is that clear and concise engagement loops will keep your users returning to the game.  We wanted to ensure the larger experience (Entry/Intro-> Draft-> Play-> Award -> Repeat), was reinforced with much smaller feedback loops.  That means updating team overall, depth chart, and play style after each round and showing the summary.  This is why we show your progress upon exiting game.

First Playable and Prototypes

One thing we identified right from the start is that we needed to get a vertical slice available as quickly as possible.  We started by getting a C# app early on that provided us with a something to playtest.  That was a great way for us to test the content, the number of choices per round, and the number of rounds. This let us test whether the content we were planning was interesting enough, plus also served as a basic tech demo and test.  It also helped us identify what information the user’s found critical to make a decision.

We even prototyped giving a timer on the round, but felt that after a few play tests, this was really against our pillar of being asynchronous.  We wanted users to be able to walk away from a draft if they had to.

While prototyping the back end, we were also iterating upon the UI design for the mode.  When it comes to designing UI, I have adapted a significant portion of the Google Venture design sprint to fit into our processes.  Typically this means I start with some incredibly basic sketches, using their Crazy 8s, and Storyboards and end up with a click through prototype within a very short period of time.

At every step I brought in stakeholders, other people in the office, basically anyone that was under an NDA took a look at the basic UI design and provided feedback at this point.  Same goes for the click through.  I was pulling people aside all the time to look at the UI.  Fortunately, there was already a lot of positive sentiments around the mode, so people were happy to take a look.  It was during this process that I identified the basic three card design with the lineup beneath.  It really has not changed since I settled upon that design early on.  

If you want some more information on Google Venture sprints, check it out here.

The final prototype was an old fashioned paper prototype.  We used this to leverage exactly what users look at when evaluating players.  This prototype was fantastic for me as it confirmed what I had suspected.  Players really look for name recognition, team recognition, and then Overall rating.  Only the hard core players will evaluate the actual ratings of a player.  However, we wanted to make this mode fun for both the new players and the hard core so we had to figure out a way to make the details and the basic information available to the user at a glance.

Making the Right Cuts

There were a couple of additional features that were initially part of Draft Champions but were cut for a few reasons.  I really don’t want to go into more detail into what the features were in case they come back, but I felt them to be extraneous from the start and distracted from the core focus of the mode.  This was a great example of improving the design by cutting out anything that may be superfluous and not true to the pillars.

Personally, I’ve found that if it doesn’t immediately resonate with me, odds are it’s not going to resonate with me in a week, two weeks, or ever. 

Fighting against Groupthink

There were a lot of people with a stake in Draft Champions.  As a new mode in Madden, especially one that is close to MUT, there was always going to be a lot of internal interest in the mode.  While there was a lot of great feedback, there were times when we had to check ourselves to ensure we weren’t just group thinking the concept at hand.

Always question what is being proposed!

Finding the FUN during Playtests

I’m lucky that we have now have a play testing group at Tiburon that will help setup the proper cadence for play testing features.  This is really broken down into a few different stages and types of playtests.

First, we ran guided prototypes and playtests.  Working with the playtest team, I created a handful of decks with different UI options.  This early feedback was used for further iteration.  This continued throughout the development cycle.

Once we had our first playable vertical slice, we brought in testers to look at the experience.  At this point they could get through a complete draft, which meant providing feedback on UI in game in addition to content, and play against the CPU teams.

As more features were completed for Draft Champions, we continued to bring in more play testers with various experience levels to try the mode out.  We were able to fine tune the mode a significant amount due to the feedback we received during some of these playtests.

We also ran internal playtests that were open to the entire team and ultimately the entire studio.  Finally, in order to maintain focus on the mode, we had daily meetings during Alpha to walk through the mode.  This gave everyone an opportunity to primarily provide feedback on balancing content.  I think this portion was critical to getting the content iteration right.

The Future

I wrapped up my involvement soon after we were ready to ship Madden ’16.  I passed on my other designs for the features we didn’t get to this year and a few more cycles worth of improvements in other concepts that were planned. 

Here’s what some of the reviews said about DC:

Polygon 1 2

US Gamer

Kotaku Article