Quote:
Originally Posted by turtle4499
Can you please provide clarity on exactly what is being multiplied out for dynamic runs per win and split starter/rp. Can you also provide clarity of just the forumla being used to turn park factors into run adjustments. I am not really sure why FIPs run adjustment takes into account the parks doubles factor.
I don't think anyone is actually asking you to calculate park factors based on seasonal data. I am really just asking for the park factors to be applied to the actual games correctly. Applying them broadly cannot possibly work out the correct numbers.
Simple example from my reddit post. I had a starting pitcher in coors with 50 innings at home and 100 on the road (because he was shelled at home) the adjustment is being applied like he played half his innings at home. There is no reason to assume starting pitchers do this at all. Especially in any hitter friendly park in an era where pitchers aren't just pitching 9 innings.
I am really struggling to understand what amount of data do you think is needed that isn't currently being stored right now? If we are able to look up home/away splits and vs team splits there is clearly enough data being stored to apply the park adjustments to the actual parks the games occurred in.
|
A lot of that data could be available in the current season, but we may not necessarily store it for past seasons. So you can't look up a player's home splits from 3 seasons back, we'd have to store a lot more information to be able to adjust those calculations to factor that in. So I mean, yeah, arguably we could give you a more granular breakdown of the current season, but it would look weird to see a guy as 1.8 WAR on October 30th, but then on November 1st when the data rolls over, we're have to recalculate to the generic formula and he might be 1.6. Even on the batters side, when you look at the splits WAR values, we don't include defense in, because we don't actually store the game by game defensive values for players. Never mind accounting for it in previous seasons.
Now, could we adjust a few of the data points more? Maybe, sure. I guess it's always a balancing act. Especially once you get into like double-splits, we won't always track exactly what split we're looking at. Even if we special-cased some of the home/road splits, it would look weird if you saw different WAR values when looking at a split one way vs another, just because it can be complex to track in. So yeah, that's one of those assumptions where we basically treat every stat as a 50/50 home/road mix when applying park factors.
In terms of a FIP specific park factor, to be honest, I didn't even know that Fangraphs split them out. That's probably something we could split out, and give a different park factor adjustment to FIP and fWAR, but use the "traditional" park factor calculation for stuff like wRC+ or rWAR.
For the actual park factors, we tend to take the modifiers you have and see what that does to a typical stats line's wOBA, more or less. i don't recall if that's how they actually do real park factors, but I know it seemed to do reasonably. On the sortable team stats page, you can see the calculated park factor there. It may break down in some extreme cases, if it does look off, feel free to report that and we can see about fixing or updating that. It's always possible that something in a calculation somewhere is off as well. Again, sometimes we do make some assumptions or adjustments (ie. in some cases maybe we apply a basic modifier on league stats to remove pitcher batting lines, but in other cases we might actually have that data and can calculate exactly)
The rest is basically us trying to follow the BB-Ref or Fangraphs calculations as close as we can, with the data and stats that we have. We always aim to be as accurate and faithful about things as we can, but certainly there can always be some room for improvement. Especially with complex calculations like these - even those who understand them have trouble following along.