Re-made the univariate model with Percent IVB Change as the dependent variable instead of IVB Z-Score. Coefficient for temperature is 0.002, or 0.2% for every degree change in temperature, 2% for every 10 degrees. Correlation is a bit weaker for IVB percent difference than IVB z-score.

The initial goal of this was to see if Walker Buehler was substantially weather/season-average predicted 4SFB iVB, suggesting regained life on the pitch. More work needed here, but seeing 3 outperformances to end the year is mildly encouraging.

Contrasting two games of similar pitchers / weather. We see that the Z-Scores are far crazier in Citi Field than Yankee Stadium. The next step here would be to try and model iVB for an individual ballpark, based on weather.

In looking at Freddy Peralta's first start of the season (per @lancebroz.bsky.social suggestion, also at Citi Field), we see a similar Z-Score outperformance by most of the game's pitchers. This suggests that ballpark structure could play a greater role in pitch movement than initially understood.

Our model will look like the following:
(Predicted 4SFB iVB Z-Score for game) = B_0 + B_1 * (Wind Speed) + B_2 * (Temperature)

We find both temperature and windspeed have a significant effect on 4SFB iVB Z-Score, and that they account for ~22% of the variance in this Z-Score.

Walker Buehler had a strong outing in G3 of the NLCS. His 4S Fastball averaged 20.2" of iVB, outperforming his season average of 16.6". Other pitchers in this game saw similar outperformances.

Digging into Walker Buehler's playoff performances. Found the fairly strong relationship between a pitcher's IVB and temperature to be interesting:

See your point re: highest bidder, but does this part not imply he might shop the offer around for a match from a team he prefers?

(Personally hoping that Henry’s bid of 15/750 does not have a chance to be matched)