April 15, 2026
By Braden Eberhard
One of the coolest things the UFA has done in recent years is build the data infrastructure for deeper analysis. Since 2021, every team has been generating play-by-play data that can be used for more than a standard box scores, making it possible to ask not just who scored or got the assist, but how possessions unfolded, which actions made a score more likely, and which players were driving that process. That is where Shown Space comes in.
Shown Space is an advanced stats project built on UFA data and based on work presented at the 2025 MIT Sloan Sports Analytics Conference, where we introduced a machine learning approach to estimating player value and decision-making in professional ultimate. At its core, it uses the league’s play-by-play data to describe possessions, player impact, and team style in a way that goes beyond the box score while staying grounded in the game itself.
The core idea: Contribution
The central metric in that effort is adjusted Expected Contribution, or aEC. Colloquially, we refer to it simply as Contribution. Contribution estimates how much a player’s actions added to their team’s expected scoring output over the course of a game or season. In practical terms, it is trying to measure how much value a player created through their throws and catches, while taking context into account.
That matters because not every completion is equally valuable. A reset to the middle of the field can do a lot for an offense even if it loses yards. The same is true of gains that look identical in the stat sheet. Two throws might each gain ten yards, but one could move the disc into an advantageous space while the other changes very little about the possession. Contribution is meant to capture that difference.
The easiest way to think about the scale is in points. A player with 1.0 Contribution has added about one point worth of value. A good example is Walker Frankenberg’s game against the Salt Lake Shred last season. He finished with 12.5 Contribution in a 25-21 loss, the highest single-game mark of the year. Put simply, that means Frankenberg added almost 13 points worth of value through his throws and catches in that game. Even before getting into the model in detail, that gives a much clearer sense of how much of Oakland’s offense ran through him and how much he created from possession to possession.
A new way to describe ultimate
This is the goal of Shown Space for 2026. We want to start with one core idea and use it to make stories around the league easier to understand. Contribution gives us a better way to talk about player value, roster turnover, team identity, and season-long trends.
Over time, there will be more to build from there. Contribution can be broken down further into the value players create as throwers and receivers. Along with other new metrics, it can help describe player archetypes more clearly, separating players not just by raw production, but by the ways they shape possessions and generate offense.
More than anything, this is about matching the sport’s analysis to the quality of the information now available. The UFA has spent years building toward a richer statistical picture of the game. Shown Space is one way to meet that moment and turn that data into something useful, readable, and connected to what we are already seeing on the field.
If you want to learn more, you can visit shownspace.com, read more at the Shown Space Substack, or find the original Sloan paper, “A Machine Learning Approach to Throw Value Estimation in Professional Ultimate Frisbee”
