Minie
@cylune.bsky.social
Habs fan. Hockey stats enthusiast. I like to fact-check hockey stats.
Thanks for the update on Evans. Appreciate the great work and enjoy the rest as well!
December 24, 2025 at 1:20 PM
Thanks for the update on Evans. Appreciate the great work and enjoy the rest as well!
Reposted by Minie
Ducks continuing to play the silliest games this year.
December 23, 2025 at 3:19 PM
Ducks continuing to play the silliest games this year.
For such a big market, analytics literacy is surprisingly low among the media covering the team.
December 22, 2025 at 2:04 AM
For such a big market, analytics literacy is surprisingly low among the media covering the team.
Thanks 😊 always appreciate your input
December 21, 2025 at 6:54 PM
Thanks 😊 always appreciate your input
Brian, the Corsi numbers you’re using are misleading. Corsi is meant to be evaluated at 5v5, not all situations. All-situations data will obviously make PKers look bad.
At 5v5, Evans is at 46.4%, not 36%, which is reasonable given his usage and deployment. His isolated impacts numbers are solid.
At 5v5, Evans is at 46.4%, not 36%, which is reasonable given his usage and deployment. His isolated impacts numbers are solid.
December 21, 2025 at 5:18 PM
Brian, the Corsi numbers you’re using are misleading. Corsi is meant to be evaluated at 5v5, not all situations. All-situations data will obviously make PKers look bad.
At 5v5, Evans is at 46.4%, not 36%, which is reasonable given his usage and deployment. His isolated impacts numbers are solid.
At 5v5, Evans is at 46.4%, not 36%, which is reasonable given his usage and deployment. His isolated impacts numbers are solid.
Advanced stats guys that don't understand how their own model works bsky.app/profile/cylu...
Reacting to this JFresh post on twitter:
I assume his model is similar to HockeyViz. Kapanen’s numbers look awful, Demidov’s look decent, so the model conclusion becomes “Kapanen sucks and is carried by Demidov.” I don’t buy it.. ⬇️
I assume his model is similar to HockeyViz. Kapanen’s numbers look awful, Demidov’s look decent, so the model conclusion becomes “Kapanen sucks and is carried by Demidov.” I don’t buy it.. ⬇️
December 21, 2025 at 1:58 AM
Advanced stats guys that don't understand how their own model works bsky.app/profile/cylu...
Remember this? Suzuki started playing better at the same time as RHP joined his line so the model attributed Suzuki's success to RHP and this is the result.
December 20, 2025 at 7:38 PM
Remember this? Suzuki started playing better at the same time as RHP joined his line so the model attributed Suzuki's success to RHP and this is the result.
Models are powerful tools - I pay a monthly fee to access hockeyviz charts. But knowing their blind spots, like inseparable linemates and priors results that don’t apply to a rookie improving at warp speed, matters just as much as reading the chart.
December 20, 2025 at 5:16 PM
Models are powerful tools - I pay a monthly fee to access hockeyviz charts. But knowing their blind spots, like inseparable linemates and priors results that don’t apply to a rookie improving at warp speed, matters just as much as reading the chart.