qntkhvn
@qntkhvn.bsky.social
data; sports; memes; #RStats
qntkhvn.netlify.app
qntkhvn.netlify.app
if you're in Chicago (or close by), come to this show this coming Thursday (Aug 14)
more show info + tickets: eventbrite.com/e/uncontrolled-variables-tickets-167491434389
more show info + tickets: eventbrite.com/e/uncontrolled-variables-tickets-167491434389
August 9, 2025 at 11:28 PM
if you're in Chicago (or close by), come to this show this coming Thursday (Aug 14)
more show info + tickets: eventbrite.com/e/uncontrolled-variables-tickets-167491434389
more show info + tickets: eventbrite.com/e/uncontrolled-variables-tickets-167491434389
Congratulations to the one and only @statsinthewild.bsky.social for winning the 2025 ASA Section on Statistics in Sports Significant Contributor Award. #truehero #jsm2025
August 4, 2025 at 7:33 PM
Congratulations to the one and only @statsinthewild.bsky.social for winning the 2025 ASA Section on Statistics in Sports Significant Contributor Award. #truehero #jsm2025
Excited to be speaking at the US Olympic & Paralympic Performance Innovation Summit (Aug 27–28, Colorado Springs), where I'll present recent work in football analytics using player tracking data.
Register for the event here: teamusaevents.teamusa.org/trncl/PI_Sum...
Register for the event here: teamusaevents.teamusa.org/trncl/PI_Sum...
July 21, 2025 at 6:34 PM
Excited to be speaking at the US Olympic & Paralympic Performance Innovation Summit (Aug 27–28, Colorado Springs), where I'll present recent work in football analytics using player tracking data.
Register for the event here: teamusaevents.teamusa.org/trncl/PI_Sum...
Register for the event here: teamusaevents.teamusa.org/trncl/PI_Sum...
Of course, what could've been another great metric name...
July 9, 2025 at 1:12 AM
Of course, what could've been another great metric name...
We find that ball carriers within the TE & WR positions (w/ more open-field, unstructured movements) differ more in their turn angle variability compared to RBs
July 9, 2025 at 1:12 AM
We find that ball carriers within the TE & WR positions (w/ more open-field, unstructured movements) differ more in their turn angle variability compared to RBs
We fit a von Mises mixed-effects model for the turn angle, modeling both the mean and concentration parameters.
We include ball carrier random effects at the concentration level, with variance grouped by position.
This reveals the shiftiest players with high variability in turning behavior.
We include ball carrier random effects at the concentration level, with variance grouped by position.
This reveals the shiftiest players with high variability in turning behavior.
July 9, 2025 at 1:12 AM
We fit a von Mises mixed-effects model for the turn angle, modeling both the mean and concentration parameters.
We include ball carrier random effects at the concentration level, with variance grouped by position.
This reveals the shiftiest players with high variability in turning behavior.
We include ball carrier random effects at the concentration level, with variance grouped by position.
This reveals the shiftiest players with high variability in turning behavior.
Using #BigDataBowl tracking data, we model the instantaneous turn angle (i.e. the angle a player takes between consecutive steps; inspired by the animal movement literature) for NFL ball carriers.
We focus on RBs after handoff and RBs/TEs/WRs after the catch.
We focus on RBs after handoff and RBs/TEs/WRs after the catch.
July 9, 2025 at 1:12 AM
Using #BigDataBowl tracking data, we model the instantaneous turn angle (i.e. the angle a player takes between consecutive steps; inspired by the animal movement literature) for NFL ball carriers.
We focus on RBs after handoff and RBs/TEs/WRs after the catch.
We focus on RBs after handoff and RBs/TEs/WRs after the catch.
New #CMSAC #SportsAnalytics paper on evaluating change of direction in the #NFL by myself and @stat-ron.bsky.social
arxiv.org/pdf/2507.06122
Spoiler alert: DK Metcalf displays straight-line, less agile movement, while George Pickens exhibits erratic, more variable directional movement.
arxiv.org/pdf/2507.06122
Spoiler alert: DK Metcalf displays straight-line, less agile movement, while George Pickens exhibits erratic, more variable directional movement.
July 9, 2025 at 1:12 AM
New #CMSAC #SportsAnalytics paper on evaluating change of direction in the #NFL by myself and @stat-ron.bsky.social
arxiv.org/pdf/2507.06122
Spoiler alert: DK Metcalf displays straight-line, less agile movement, while George Pickens exhibits erratic, more variable directional movement.
arxiv.org/pdf/2507.06122
Spoiler alert: DK Metcalf displays straight-line, less agile movement, while George Pickens exhibits erratic, more variable directional movement.
Second lap: (got gassed and could barely do anything during the last 10 min)
bsky.app/profile/libb...
bsky.app/profile/libb...
June 13, 2025 at 6:36 PM
Second lap: (got gassed and could barely do anything during the last 10 min)
bsky.app/profile/libb...
bsky.app/profile/libb...
Last night was amazing + exhausting. Competed in a vizbuzz double header!
Here are my vizzes from last night.
First lap:
Here are my vizzes from last night.
First lap:
June 13, 2025 at 6:34 PM
Last night was amazing + exhausting. Competed in a vizbuzz double header!
Here are my vizzes from last night.
First lap:
Here are my vizzes from last night.
First lap:
January 17, 2025 at 12:15 AM
Register now for the 2025 Connecticut #SportsAnalytics Symposium (April 11-12 @ Yale) statds.org/events/csas2...
More information about speakers, accommodations, etc. can be found on the conference website.
And FYI - I will be leading a workshop on player tracking data using #BigDataBowl data.
More information about speakers, accommodations, etc. can be found on the conference website.
And FYI - I will be leading a workshop on player tracking data using #BigDataBowl data.
January 10, 2025 at 12:26 AM
Register now for the 2025 Connecticut #SportsAnalytics Symposium (April 11-12 @ Yale) statds.org/events/csas2...
More information about speakers, accommodations, etc. can be found on the conference website.
And FYI - I will be leading a workshop on player tracking data using #BigDataBowl data.
More information about speakers, accommodations, etc. can be found on the conference website.
And FYI - I will be leading a workshop on player tracking data using #BigDataBowl data.
Next up, Trench Chess (kaggle.com/abhishekvaradarajan/trench-chess) by undergrad Abhi Varadarajan
Trench Chess explores how defensive lines change their looks from pre-snap to post-snap and how it relates to pressures, focusing on two pass rush schemes: disguised pressures and stunts
Trench Chess explores how defensive lines change their looks from pre-snap to post-snap and how it relates to pressures, focusing on two pass rush schemes: disguised pressures and stunts
January 7, 2025 at 1:29 PM
Next up, Trench Chess (kaggle.com/abhishekvaradarajan/trench-chess) by undergrad Abhi Varadarajan
Trench Chess explores how defensive lines change their looks from pre-snap to post-snap and how it relates to pressures, focusing on two pass rush schemes: disguised pressures and stunts
Trench Chess explores how defensive lines change their looks from pre-snap to post-snap and how it relates to pressures, focusing on two pass rush schemes: disguised pressures and stunts
First up, CHASE (kaggle.com/lauerjames/chase-a-new-metric-for-receiver-spatial-impact) by 3 MADS students (James Lauer, Nicco Jacimovic, Jason Andriopoulos) and 1 neuroscience PhD student (Larry Jiang)
CHASE (Convex Hull Area Strength Estimate) quantifies a receiver’s impact on defensive spacing
CHASE (Convex Hull Area Strength Estimate) quantifies a receiver’s impact on defensive spacing
January 7, 2025 at 1:28 PM
First up, CHASE (kaggle.com/lauerjames/chase-a-new-metric-for-receiver-spatial-impact) by 3 MADS students (James Lauer, Nicco Jacimovic, Jason Andriopoulos) and 1 neuroscience PhD student (Larry Jiang)
CHASE (Convex Hull Area Strength Estimate) quantifies a receiver’s impact on defensive spacing
CHASE (Convex Hull Area Strength Estimate) quantifies a receiver’s impact on defensive spacing
Resharing our #BigDataBowl entry as the submission deadline is today kaggle.com/tindata/down...
TL;DR—We use a Bayesian hierarchical modeling approach (w/ a Gamma distribution) to model QB snap timing variability and find that higher variability relates to facing less havoc generated by the defense
TL;DR—We use a Bayesian hierarchical modeling approach (w/ a Gamma distribution) to model QB snap timing variability and find that higher variability relates to facing less havoc generated by the defense
January 6, 2025 at 2:26 PM
Resharing our #BigDataBowl entry as the submission deadline is today kaggle.com/tindata/down...
TL;DR—We use a Bayesian hierarchical modeling approach (w/ a Gamma distribution) to model QB snap timing variability and find that higher variability relates to facing less havoc generated by the defense
TL;DR—We use a Bayesian hierarchical modeling approach (w/ a Gamma distribution) to model QB snap timing variability and find that higher variability relates to facing less havoc generated by the defense
ICYMI — Check out our #BigDataBowl entry:
kaggle.com/code/tindata...
substack.com/note/p-15308...
We model QB snap timing variability and find that higher variability relates to facing less havoc events generated by the opposing defense.
Yet again, Patrick Mahomes is really good at football...
kaggle.com/code/tindata...
substack.com/note/p-15308...
We model QB snap timing variability and find that higher variability relates to facing less havoc events generated by the opposing defense.
Yet again, Patrick Mahomes is really good at football...
December 16, 2024 at 3:03 PM
ICYMI — Check out our #BigDataBowl entry:
kaggle.com/code/tindata...
substack.com/note/p-15308...
We model QB snap timing variability and find that higher variability relates to facing less havoc events generated by the opposing defense.
Yet again, Patrick Mahomes is really good at football...
kaggle.com/code/tindata...
substack.com/note/p-15308...
We model QB snap timing variability and find that higher variability relates to facing less havoc events generated by the opposing defense.
Yet again, Patrick Mahomes is really good at football...
December 14, 2024 at 4:48 AM
Our contribution also includes a simple approach using derived tracking data features for identifying the types of motion via a Gaussian mixture model (which is then fed into our model as a covariate).
December 13, 2024 at 10:41 PM
Our contribution also includes a simple approach using derived tracking data features for identifying the types of motion via a Gaussian mixture model (which is then fed into our model as a covariate).
The results are fun. Patrick Mahomes displays the highest variability in snap timing, with Daniel Jones at the bottom. We also observe a relationship with havoc rate across all passing plays, not just those with motion! kaggle.com/code/tindata...
December 13, 2024 at 10:41 PM
The results are fun. Patrick Mahomes displays the highest variability in snap timing, with Daniel Jones at the bottom. We also observe a relationship with havoc rate across all passing plays, not just those with motion! kaggle.com/code/tindata...
We model snap timing using a Gamma distribution (for both mean and shape parameters ) with a Bayesian multilevel model, where the QB shape random effects capture differences between QB snap timing variability on passing plays with motion.
December 13, 2024 at 10:41 PM
We model snap timing using a Gamma distribution (for both mean and shape parameters ) with a Bayesian multilevel model, where the QB shape random effects capture differences between QB snap timing variability on passing plays with motion.
Excited to share my #BigDataBowl submission this year, joint work with the great Ron Yurko: Down, set, hut! Explaining variability in snap timing on plays with motion www.kaggle.com/code/tindata...
December 13, 2024 at 10:41 PM
Excited to share my #BigDataBowl submission this year, joint work with the great Ron Yurko: Down, set, hut! Explaining variability in snap timing on plays with motion www.kaggle.com/code/tindata...
my #adventofcode day 1 solution #Rstats -- doing it the old school base R way; was aiming for as few lines as possible
December 1, 2024 at 3:06 PM
my #adventofcode day 1 solution #Rstats -- doing it the old school base R way; was aiming for as few lines as possible
@statsinthewild.bsky.social looking up how to make a stacked bar chart live during the competition
November 29, 2024 at 6:22 AM
@statsinthewild.bsky.social looking up how to make a stacked bar chart live during the competition
Me this viz buzz szn
November 29, 2024 at 12:26 AM
Me this viz buzz szn