CJ Turtoro Profile picture
Hockey Analytics Researcher | Devils Blogger | Insufferable Pedant
Jan 4 20 tweets 6 min read
Okay so I've talked to several of the modelers about this and I want to just give a sort of public report for posterity on the state of the public xG models and the NHL shot tracking changes (2023) that have influenced them as I understand it now. 🧵 Okay so what happened? To understand that, I think we need to know which models did NOT fall out of tune. I examined xG models from Hockeyviz (McCurdy), Evolving-Hockey (Younggrens), Moneypuck (Tanner), HocketStats (Bacon), NaturalStatTrick (Timmins), and Hockeyanalsis (Johnson)
Jan 25, 2025 12 tweets 5 min read
Okay let's go.

We'll start with @HockeyViz.

Interestingly, Necas/Rantanen are negative play-drivers while Hall/Drury are positive.

The difference is all from production (finishing/setting) and penalty±.

In all. It's two back-end top-liners and two middle-6ers. Even trade. Image
Image
Image
Image
On @EvolvingHockey, however, there's a clear star and it's Rantanen. Grades or as an offensive dynamo and clear overall upgrade on Necas.

Similar to HVs viz, Drury also grades out as a clearly useful defensive 5v5 piece.

2 3rd liners, one 1L and one star. CAR📈 Image
Image
Image
Image
Nov 8, 2024 7 tweets 3 min read
People acting like it's way more complicated than it is. Job growth is great but costs, particularly housing, are insane. Everyone's working their asses off but aren't getting the American Dream they were promised in return. Look at where the biggest Dem drops were... Dem vote went UP among 65+ because they already had homes. Historic lows among young adults who are trying to buy. Image
Image
Mar 3, 2021 9 tweets 3 min read
I think one reading this tweet could take two big ideas away: one of which is an essential caveat to a term that's been a mainstay in the analytical zeitgeist since its inception, the other of which is throwing the baby out with the bathwater. Warrants a nuanced comment IMO (1/?) One possible takeaway is that PDO is made up and should be replaced with better metrics. This is true. IMO PDO should've never been created -- it is composed of two metrics that are complementary on the surface, but reflect completely independent and unrelated underlying skills.