Welcome to Shown Space
Here’s What You’re Looking At
Shown Space is a new analytics website for ultimate frisbee, powered by UFA data and statistical modeling/machine learning (whichever sounds less daunting to you). We go beyond surface-level stats, breaking performance into four key areas—Defense, Throwing, Receiving, and Team Impact—with custom metrics that reveal deeper insights. Check out the website at shownspace.com.
The Models Under the Hood
At the heart of Shown Space are two core models that make all of this possible which are adapted from and extensions of our recent [2nd place] research at the MIT Sloan Sports Analytics Conference (found here).
The first model evaluates field value (FV)—It considers positional and time context to calculate how much a given position and time contributes to the chance of scoring. This gives us a baseline for understanding the value of movement and positioning across the game.
The second model estimates completion probability (CP) for each throw. It takes into account factors like throw distance, angle, spacing, and game clock to predict how likely a throw is to succeed. This lets us quantify the risk of decisions, not just their outcomes.
Together, these models allow us to attach meaningful, situation-aware information to every event in a game. That means every pass, cut, or turnover can be evaluated not only by what happened, but also by what should have happened:
Value Added: How much a throw or action changed the team's chance of scoring.
Risk Level: How difficult or aggressive a throw is
Performance vs. Expectation: Did the player outperform what the model expected, or fall short?
Impact Metrics: How a player’s choices contribute to team success, not just individual stats.
These models aren’t meant to tell you what to think; instead, they give a more nuanced lens for understanding why a play worked, what could’ve been done differently, and who’s—maybe quietly—making the biggest difference.
From Numbers to Edge: Applying the Insights
The data and visuals on Shown Space aren’t just there for observation—they’re there for action. Here’s how they can be applied to enhance your understanding of the game and give you a competitive edge:
Measure Actual Value with aEC: Not all yards are created equal—and Adjusted Expected Contribution (aEC) is built on that principle. aEC is our best single-number estimate of how much a player helped their team score.
It works by measuring the change in scoring probability across every event a player is involved in, then summing those changes over the course of a game. The model is scaled so that +1 aEC equals the value of scoring one goal. That means if a player ends the game with an aEC of +4.77, their actions added roughly that much scoring value to their team—whether they threw assists or just consistently made plays that moved the offense toward scoring. This goes far beyond traditional stats like yards gained or plus/minus. Our underlying FV model takes into account more than just distance:
Resets and backward throws still get credit—because sometimes a dump swing is a positive path forward.
Central field positioning is rewarded—because the model has learned that being in the middle often leads to better options.
Turnovers are weighted by location and game context—so throwing away the disc 10 yards from your own end zone with 30 seconds left is (correctly) penalized less than a reset miscue in the middle of the quarter.
aEC doesn’t reward extra yards after the goal line—a goal is a goal, whether it’s caught at the front cone or the back. The model is designed to reflect actual impact on scoring, something missed by yardage metrics.
In short, aEC captures value the way it really exists on the field: not in yards, but in value.
A quick aside—we use the term “Adjusted” because the unadjusted metric, Expected Contribution, does not scale the field value so that a goal scoring possession equals one. For more information on Expected Contribution, feel free to read the Sloan Sports Analytics Paper.
Evaluate Thrower Performance with xCP and CPOE: Just like yards, not all completions are created equal—and Expected Completion Percentage (xCP) and Completion Percentage Over Expected (CPOE) are designed to capture that nuance.
xCP estimates the likelihood that a given throw will be completed, based on historical data and contextual factors like throw distance, angle, and field location. Each throw is assigned a probability, from lower-percentage hucks to easy resets.
CPOE then compares a player's actual completion rate to what we'd expect given the difficulty of their throws. A positive CPOE means a player is completing more passes than expected. A negative CPOE suggests the opposite: missed opportunities or overly aggressive decisions that fall short.
Taken from nfelo. Why does this matter?
CPOE rewards precision under pressure—not just raw completion rates. Two handlers with 90% completions might look the same in the box score, but if one has a CPOE of +6.5%, they’re executing tougher throws with greater accuracy.
xCP adjusts for difficulty—throwing 30 yards across the field shouldn’t be judged the same as a 5-yard dump. xCP helps normalize that context.
Risk and reward come into balance—There’s no single correct way to make decisions—but xCP and CPOE quantify the risk and reward behind each one, offering a fuller picture of a player’s decision-making style and effectiveness.
Together, xCP and CPOE give a better picture into throwing performance that goes beyond just completions. They show how well a player executes in context, and whether they’re truly elevating the offense.
Player Impact: Evaluate how balanced or top-heavy a team is with our Player Impact (PI) metric. This metric quantifies how much players contribute to the team relative to their team’s overall contribution.
Scout Player Tendencies: Dive into player scouting reports to analyze their throwing and receiving tendencies. Find out where a player operates best on the field, what types of throws they prefer, and how they fit into their team’s offensive strategy.
More to come, but this is currently explored through radar charts that show the directions and relative frequency of the direction that a player throws to and get receptions from.
Track Player Efficiency Trends: Visualize how a player’s total aEC changes over time. This plot highlights performance consistency, breakout games, and potential slumps—giving coaches, analysts, and fans a tool to assess growth, impact, and momentum across a season or career.
Follow the MVP Race: Use our leaderboards to track the advanced stat-based MVP race in real-time, identifying who’s truly making an impact across various game facets—whether it's scoring, defense, or overall team contribution. Watch as the rankings shift with each game and season.
Coming Soon: The Next Level of Ultimate Analysis
We’re just getting started. We’ve got some ambitious goals, so here’s a look at what’s on the horizon:
Game Center
Your go-to destination for live and historical game tracking. Think ESPN’s Game Flow but for ultimate. Follow possessions in real time, view summary stats, see how momentum shifts, and track the game without having to have access to the stream.
Team Pages & Rankings
We’re launching full team hubs with season-long performance breakdowns, team-level metrics, and rankings that go beyond wins and losses. Expect deep dives into team buildups, roster efficiency, head-to-head comparisons, and more.
Scouting Reports & Player Trends
Get the edge with data-driven insights into opposing players—track usage, involvement, and impact across games and matchups. We’re building tools to make scouting smarter and more efficient.
Interested in team scouting reports for your team? Reach out—we’d love to collaborate.
Advanced Filters & Historical Access
Dive deeper into the data with more detailed filters, such as game phase, player role, quarter, and more. Explore full-season archives and unlock past leaderboards and rankings for the historical frisbee nuts.
…this is so cool