Tech/Systems Designer

Storytelling in Madden:

In Madden’s most popular mode, players can make their own custom player, coach, or team and play through the entirety of a career or franchise. This mode is largely popular because players want to be the main characters of stories like the ones they see on TV: they want to experience the rise of new legends, heated rivalries, and to experience the journey to becoming one of the greatest quarterbacks, franchises, or coaches in NFL history.

The problem is that Madden’s approach to telling in-game stories either speak only to the play which just happened, or are specific to the current real world football season. Outside of game commentary, the franchise mode falls even further short. You can see basic stats about what’s happened last week in your league and little else.

My conviction is that information about a single point in time is NOT a story. Stories, at a minimum, have a beginning, middle, and end. Furthermore: videogames are inherently an interactive medium, and therefore a good video game story should be written in collaboration with the player.

The Story of a Game

How can we use in-game data to tell custom stories?

Kurt Vonnegut gave a lecture called “The Shape of Stories” in which he plotted different story archetypes across an X-Y axis where X represented Time and Y represented Good Fortune. Below are four of his examples:

And here is ESPN’s Patriots win probability graph from the 2016 Superbowl against the Atlanta Falcons:

While other genres such as RPGs are more traditionally associated with good storytelling, my conviction is that competitive 1:1 games are actually blessed relative to other genres when it comes to emergent story telling because player fortunes can be summed up entirely with one metric - their win probability.

To move past point-in-time analysis towards actual storytelling, the first and easiest step is to use the win-probability graph to categorize a story type. Categories we used included comeback, upset, slow victory march, nail-bitter etc.

Stitching Together a Narrative

Okay, we should track win probability to track the shape of a story, But how do you turn the shape of a story into an actual story?

Not every play in a game is equally important or impactful and not every quarter in a game is interesting. What we need to do is identify “story beats.”

If a game is the story of who is going to win, then the story beats should, generally, be the points in time where the win probability changes. Therefore you can structure a story beat as a three part mini story like so:

This may seem like an over simplification, but this sort of cause-effect story telling is exactly how much of sports journalism works. Here's a real world example from Cris Collinsworth and Taran Killam writing for ESPN:

How do we get from the first picture to the second one

First you need to pick a time period to analyse the win probabilities before and after. To do this, simply look for periods of dramatic win probability change. Failing that, we can look at things like Q1 vs Q2 or the win probability coming into the game vs the probability at half time.

That gives us the first and third parts of the story beat. To fill in the middle, we look at three categories of information.

Big Plays

This is the Malcolm Butler example. One big play that changes the win probability - an interception, a long pass, sometimes even just a 4th down stop. It’s easy to track these metrics. If the win probability changes wildly on one play, we just look and see what happened on that play and that’s the story.

KPI Change

When the win probability changes more gradually and we down have a specific event to point to, we instead want to look at KPIs. Here, KPIs refer to stats which are always good or bad. So, for example, Yards per Pass Attempt, Pass Completion Rate, 3rd down conversion rate, etc, are all metrics which you (nearly) always want to be high.

Decision Metrics

These are metrics around decisions players are making, which may be strategic in nature but not inherently tied to success or failure. Examples are Pass vs Run play %. Time to Throw. Man vs Zone Percent. Blitz frequency. Etc

These story beats can be stitched together across a game using a bank of flavor text to create more complex stories. Here are some examples:

Expanding Stories to Seasons and Careers

The same design and principles can be extended across a season by replacing “Win Probability” with “Probability to Win The Super Bowl,” across a career by replacing it with “GOAT score” (a custom formula which would rank the player PC against real world players of the same position), and across a franchise by replacing it with “Legacy Score” (another custom formula for a team.)