Math and sports are pretty cool || TCNJ 27’
Joined December 2022
- Tweets 32
- Following 49
- Followers 4
- Likes 144
9 Photos and videos
Shane Sullivan retweeted
May 27
Did a deep dive into CFB rest advantages (s/o to @SharpFootball for the inspiration) & here's what I found:
1) The schedule this year is more condensed than usual with how the calendar falls, w only 13 weeks between Labor Day & Thanksgiving (ie only 1 bye for most teams)
1/5
2
5
28
1,931
Shane Sullivan retweeted
Apr 26
Apr 26
god forbid a man wants to be named after a programming language
1
9
2,203
Final plot of draft pick # vs predicted wAV/G.
UNDERDRAFTED: Eli Heidenreich
OVERDRAFTED: Jadarian Price, Kaelon Black
Spreadsheet (docs.google.com/spreadsheets…) has undrafted RB. Surprised to see Noah Whittington and Roman Hemby go undrafted.
Project I’ve been working on for the Draft, using athletic testing and PFF data to project RB prospect performance. Used drafted RBs drafted since 2015, measuring performance with Pro Football References weighted Approximate Value (per game). My 2026 results: (more in replies)
325
Shane Sullivan retweeted
Apr 24
#Rutgers Nation, time to step up.
For each repost/like of this post through tomorrow I will donate $52 to @RUAthletics and @R_Fund.
Please join me, spread the word and #MaxOutTheMillionMatch.
Donate here: give.rutgersfoundation.org/s…
Rutgers Giving Day is LIVE—and starting early! 🎉
Inspired by the Hall of Distinguished Alumni Awards Ceremony and the leadership of Greg LC'82 & Anna Brown, whose $1M match challenge is motivating alumni and friends to give early and give boldly.
🔗 rfund.scarletknights.com/giv…
11
384
812
86,228
Project I’ve been working on for the Draft, using athletic testing and PFF data to project RB prospect performance. Used drafted RBs drafted since 2015, measuring performance with Pro Football References weighted Approximate Value (per game). My 2026 results: (more in replies)
1
371
I cross validated by draft class in testing, looking to maximize the mean of each draft classes correlation between predicted wAV/G and actual wAV/G.
Mean R^2 between Draft Pick # and wAV/G: 0.383
Mean R^2 between predicted wAV/G and WAV/G: 0.410
(More in replies)
2
36
And finally, here’s a spreadsheet with wAV/G and predicted wAV/G for all drafted RB dating back to 2015, along with 2026 prospects.
docs.google.com/spreadsheets…
21
Looking back at the initial tweet, you’ll see my model has Jeremiyah Love as the best RB prospect in the past 11 draft classes, in the projected All-Pro tier along with Jonathan Taylor, Ezekiel Elliot, Saquon, JK Dobbins, and Bijan. Then a massive gab to 2nd Eli Heidenreich.
155
Shane Sullivan retweeted
Mar 17
Build a model with me!
In this video, I walk through how I built a tight end evaluation model in Dropback. I cover:
-> metric selection
-> regression types
-> best practices
-> distributions
-> player benchmarking
@DropbackHQ
1
5
64
16,202
Been doing some modeling for the NCAA Tournament, my favorite upset at every seed line
9 Utah St* over Nova
10 Santa Clara* over UK
11 VCU* over UNC
12 Northern Iowa over SJU
13 Hofstra* over Bama
14 Kennesaw St over Gonzaga
15 Tennessee St over Iowa St
16 Howard over UM
1
302
66/68 in the field correct
47/68 on correct seed line
66/68 within 1 seed line
Not terrible, but was one of the easier 68/68 years in recent memory. Still pretty happy with how this went with limited time. Lots of room for improvement!
Fully Modeled Bracketology!
Thought of this somewhat last minute and only based on data from last year, but thought this would be fun. Definitely some things I disagree with and probably a few things wrong with the actual bracketology. But still excited to see how this does!
27
Fully Modeled Bracketology!
Thought of this somewhat last minute and only based on data from last year, but thought this would be fun. Definitely some things I disagree with and probably a few things wrong with the actual bracketology. But still excited to see how this does!
143
That’s a wrap for us at #SABRanalytics! Thanks to all of our panelists, presenters, attendees in person at @ASUCollegeOfLaw and online, and partners at @MLB, @RawlingsSports, and @Teamworks. Stay tuned for more highlights and videos at sabr.org/analytics
ALT Clockwise from top-left: Arizona Diamondbacks Front Office 360 panel, Tyrone Brooks of Major League Baseball, SABR Analytics Conference Yoseloff Scholarship and Sarah Langs Scholarship recipients, Siddharth Ramakrishnan, Jessica Fackler.
3
21
2,705
Shane Sullivan retweeted
Feb 27
Registration for the 2026 #SMTDataChallenge is OPEN! We want to see what you would put “On the Big Screen!” We want to see how you can tell stories like @SMTlive through a metric, visualization, or interactive fan experience that could be displayed on a MiLB scoreboard!
2
11
18
4,894
Data on new Rutgers DC Travis Johansen from his time at South Dakota.
docs.google.com/spreadsheets…
83
For the 2026 @ShrineBowl x @SumerSports 2026 Analytics Competition, I created a metric for evaluating DB performance in limiting effective separation in the WR vs DB 1 on 1 drills.
Additional info on Separation Allowed Under Coverage Expectation, or SAUCE for short, below!
1
4
124
Also just for fun here were the best and worst reps I tracked with an animation I used throughout developing this project.
Red = WR
Dark Blue = DB
Light Blue = Expected/Model DB
Dark Green = QB
Lime Green = Expected Catch point
1
44
And if anyone has any questions as this was a pretty brief breakdown (not helped by tweet character limits), I’d love to talk about this more!
28