Data is powering the future of sports. 📊
“Every tenth of a second, the NFL’s Next Gen data chips provide information for where every single player is positioned on the field — the direction they’re moving, the speed they’re moving."
CMU researchers are tackling how to use the data at our disposal.
Carnegie Mellon research is helping #NFL teams see what game film can’t. Smarter #data means smarter draft picks.
By measuring traits that aren't always obvious, from shiftiness to arm angle, we can study their impact on the game. Analyzing this data is shaping the way teams draft.
@cbsnews.com
People responding to statistical computation diagnostic warnings.
Many thanks to @realfrankbrank.bsky.social for getting this started! #CMSAC is excited to work with FTN data - with multiple student #sportsanalytics projects already underway!
FTN is partnering with the Carnegie Mellon Sports Analytics Center to provide advanced charted data to students to help prepare them for a career in sports data and analysis.
Special thanks to @stat-ron.bsky.social for being a leader in this partnership.
And check out Quang Nguyen @qntkhvn.bsky.social 's latest work that builds off this!
arxiv.org/abs/2603.17866
Official publication of ' #NFL ghosts: A framework for evaluating defender positioning with conditional density estimation' is now available in the Annals of Applied Statistics projecteuclid.org/journals/ann... #BigDataBowl #sportsanalytics #CMSAC
Kermit the frog screaming with excitement
We have summer internships y'all! Come work at Posit on the PyData, tidymodels, shiny, or Connect teams: grnh.se/tigz810a3us. You will have an awesome time, learn a ton, and help advance our open source and pro tools 🧰 #rstats #pydata
This framework enables comparison of observed movement against a distribution of hypotheticals at each frame within a play
We focus on evaluating RB movement after handoff in run plays, and present examples of play valuation breakdown & player performance metrics
Simulation details:
Given an observed path, at each frame along this path, generate a local distribution of hypothetical next steps (only simulate the immediate next movement, 1 frame ahead)
(work in progress: full path sim by repeatedly drawing posterior predictive steps & turns)
Full framework:
1. Movement characterization with tracking data features (step length & turn angle, inspired by animal movement literature)
2. Bayesian multilevel step-and-turn models
3. Posterior predictive simulation of player movement
4. Ghosting/hypothetical evaluation
Check out the latest from Quang Nguyen in the #CMSAC research lab #sportsanalytics #NFL #BigDataBowl
New paper by myself & @stat-ron.bsky.social
arxiv.org/abs/2603.17866
We introduce a generative modeling framework to evaluate continuous-time player movement against a distribution of frame-level hypotheticals
#cmsac #nfl #bigdatabowl #sportsanalytics #ghosting
🔔 New preprint w/ Catalina Medina
❓ We ran a quasi-experiment in an introductory stats course to see if giving students a choice of real data context on homework actually has impact
📄 Engaging students with statistics through choice of real data context on homework
🔗 arxiv.org/abs/2603.04541
👍 Our recs based on our findings are
1️⃣ use real data with authentic contexts,
2️⃣ select contexts students care about
3️⃣ incorporate variety across data contexts
4️⃣ consider choice as a pedagogical tool.
A screenshot of an RStudio window. On the left-hand side is a new pain called Posit Assistant. The Posit Assistant had recently run code making a lat-lon plot of Washington state, colored by whether the point had been marked as forested or not.
Today we're releasing AI for RStudio. It's really, really good—I'd encourage you to point it at the messiest data sources you have and see what it can do.
www.simonpcouch.com/blog/2026-03...
Teamworks—which calls itself as "The Operating System for Sports"—is closing in on acquiring Pro Football Focus per
@arif.bsky.social
. PFF's valuation in the deal is between $130M-$140M.
Cris Collinsworth purchased PFF for $6M in 2014. www.wideleft.football/p/breaking-t...
The list of why I should switch to Positron keeps getting longer.
Summary: Peak BBCOR same for all 3 bats but shifted to inside for Torpedos. Simple shift for Torpedo-1 but wider sweet spot for Torpedo-2. See baseball.physics.illinois.edu/ppt/Nathan-T... for my @saberseminar.bsky.social talk. This is an on-going study with additional research in progress.
ICYMI 🧵1st results of the Torpedo bat study by Lloyd Smith, @drussellpsu.bsky.social, and me. Measurements done at WSU Sports Science Lab and PSU. Will be presented at ISEA Conference in June. BBCOR, a surrogate for exit velocity, measured for 1 Standard and 2 Torpedo bats via high-speed impact.
So far, overturned ABS Challenges...
ABS Challenge report so far:
- Fielders challenging 113 to 56 for batters (or only 33% challenges are by batters)
- Fielders got 61-113 overturned (54%), earning 0.15 runs per overturn, 0.08 per challenge
- Batters 27-56 (48%), earning 0.20 runs per overturn, 0.10 per challenge
Just published in JOSS: 'SC2Tools: StarCraft II Toolset and Dataset API' https://doi.org/10.21105/joss.08889
Got to join the Wharton Moneyball podcast yesterday to talk about the state of the NBA and why @deanolytics.bsky.social belongs in the basketball hall of fame. podcasts.apple.com/us/podcast/w...
Cade is the best, you always want to hear him on the draft.
Ron is at least medium.
Appreciate anyone sharing with media that plan to be in town for the draft! @sethwalder.bsky.social @minakimes.bsky.social @benjaminsolak.bsky.social @billbarnwell.com @dannyheifetz.bsky.social
Very excited to host Cade Massy when the #NFLDraft is in Pittsburgh this year! If you're able to make it to Pittsburgh that week, then join us for FREE to see Cade's talk on the 'The Loser's Curse'! w/ virtual option for students #sportsanalytics
