Sebastian Carl
@mrcaseb.com
1.4K followers 270 following 430 posts
Football 🏈 in the #nflverse | #rstats | #dataviz | #engineering | Justin Herbert is Elite https://mrcaseb.com/
Posts Media Videos Starter Packs
Pinned
mrcaseb.com
Since many new users are looking at my profile I summarize some of my more recent #rstats dataviz here.
Thanks for following and/or sharing!
A table of implied NFL betting market probabilities of season outcomes, team ratings and strength of schedule after the 2024 preseason. NFL point differential by game second of NFC North teams after their hot start into the 2024 season. Win probability over the course of the 2024 week 6 NFL game Commanders @ Ravens Does unit investment pay off? How much #NFL teams spend on their units (in 2024) vs. their respective PFF grades through week 2. 
What stands out to me here is that the average defense grade is lower than the average offense grade although we see a decline in offensive performance so far
mrcaseb.com
@hadley.nz is there an option in pkgdown to list reexports in the reference index? I am aware of the reexports site, e.g. dplyr.tidyverse.org/reference/re...,
but those don’t appear on the index.

I would like to have those on the index for cases where I outsource functions to separate packages.
Objects exported from other packages — reexports
These objects are imported from other packages. Follow the links below to see their documentation. magrittr %>% pillar type_sum tibble add_row, as_data_frame, as_tibble, data_frame, lst,...
dplyr.tidyverse.org
mrcaseb.com
When rewriting it, I realized it could be useful to do things like standings scenarios in the final 1-2 weeks of the season. That's why I added points tiebreakers. We were SO close to see tied SOS in the NFC West last season.
mrcaseb.com
Yeah it defaults to coinflip after SOS because the main use case are season simulations. And most sims do not simulate actual scores. Just game outcomes. So applying point tiebreakers doesn’t make any sense.
mrcaseb.com
Pretty sure week 1 came down to either points scored or coin flip. My code returns the information about the applied tiebreaker.
Worth to be noted that nflseedR computes standings as "if season ended today". The are other approaches where upcoming games are incorporated, e.g. for SOV, SOS etc.
Reposted by Sebastian Carl
mrcaseb.com
Has Chase Daniel already complained that the Chiefs lost because the kickoff landed half a yard short of a made-up landing zone?
brianmfloyd.bsky.social
don’t want that hilarious game-winning drive to overshadow that harrison butker booting the ball out of bounds set the whole thing up
mrcaseb.com
Has Chase Daniel already complained that the Chiefs lost because the kickoff landed half a yard short of a made-up landing zone?
brianmfloyd.bsky.social
don’t want that hilarious game-winning drive to overshadow that harrison butker booting the ball out of bounds set the whole thing up
Reposted by Sebastian Carl
rbsdm.com
Updated the look of the box score site (with thanks to @mrcaseb.com)
mrcaseb.com
Whoops I have some ties in there for games about to start soon, sorry. Doesn't affect point differential tho
mrcaseb.com
The Indianapolis Colts currently lead the league with a point differential of +77 points
mrcaseb.com
And most of them haven't had 17 regular season games or 7 playoff spots per conference
mrcaseb.com
Stroud having a day against the Ravens backup defense?
mrcaseb.com
Testing new styles in boxscore app
mrcaseb.com
Oh nice, thank you! I swear I searched Github without result. Which means I have to work on my GH search skills haha.
mrcaseb.com
@davisvaughan.bsky.social I see that you use {cross} in PRs for benchmarks across branches. Is that an internal posit package? I would love to use this for performance PRs but I have trouble finding the source.
mrcaseb.com
Haha I ran some sims today as well because I was wondering what draft pick the Falcons will hand out to the Rams. Close to 30% a top-10 pick…
mrcaseb.com
Yes, I planned to update the strings and forgot it. I can use it as synonym, can I?
mrcaseb.com
Absolutely agree. Sometimes we do see offenses where epa is heavily inflated by big plays, though.
mrcaseb.com
Continuously moving the ball is just as important in the NFL as explosive plays.
This chart shows offensive EPA per play (x-axis) and success rate (plays where EPA > 0) split by dropback and rushing.

Imagine the Patriots fixing their rushing game!
A NFL logo scatter plot shows offensive EPA per play (x-axis) and success rate (plays where EPA > 0) split by dropback and rushing.
mrcaseb.com
But the broadcast told me the kick was "a 1000% the right decision", Seth
mrcaseb.com
I have a package that exports a method for a ggplot2 generic. ggplot2 v4 made me look up and parse the call stack to identify from where the method is called because I couldn’t figure out a better way to solve a problem.
mrcaseb.com
There’s nothing wrong about that
mrcaseb.com
I always/ use "We" because it’s me and the person reading the comment
mrcaseb.com
Yeah that’s fair. I guess it’s just a matter of time to get used to the way the code looks. I never felt the urge to try it but I understand the idea.