clock menu more-arrow no yes

Filed under:

Pointless but maybe fun, Part 2: guess the IWAG distribution for pitchers

New, 1 comment

Esoterica, part 2

Atlanta Braves v Baltimore Orioles Photo by Julio Aguilar/Getty Images

I guess folks had fun with the first part of this pointless exercise, for position players (see here), and so, we’re following up with the second (and final) part today.

With the IWAG projections for pitchers being largely complete at this point, I’ve pulled together another anonymized set of distribution projections for your quizzing pleasure. Unlike the position player array, though, this one requires a bit more explanation.

MLB, at least at this point, appears to be rushing headlong down the slope of “no starters, no relievers, just pitchers.” (I find this a welcome change, but my preferences are beside the point.) To get ahead of this, I’ve re-done the pitching module of IWAG this year to be able to handle three types of players: (1) your traditional, go-as-long-as-he-can starter; (2) your traditional, one-inning-ish reliever; and (3) everyone else, including swingmen, multi-inning relievers, opener candidates, “starters” that should really only be limited to one or two times through the order, or even guys who may end up having their workloads curtailed for various reasons, whether health-related, performance-related, or just plain there’s-no-room-in-the-rotation-for-you.

As such, the quiz below (more on the modeling once the actual projections post goes up, but this is just about the quiz) mixes the first group (starters) and the third group (mishmash). There are no pure relievers in the quiz, because projections for those would be pointless. You can tell apart the starters (there are four of them) and the mishmashers (there are nine of them) by whether the legend at the bottom identifies the black line as WAR/200 (starters) or WAR/140 (mishmash). Given this, you should keep in mind two things for the mishmashers. First, each mishmasher’s WAR projection is oriented around his own projected usage, which isn’t consistent. Therefore, a “better” run preventer working more heavily as a starter will achieve greater WAR-per-inning than one working more heavily as a reliever. Second, the actual WAR (non-rate-basis) figures vary widely based on innings expectations and potential, so you will be best served by making educated guesses based on the gold line, rather than the black one. The black line could be very misleading for players who are generally expected to only be able to throw 30 frames or so, but the potential for those frames to be high-quality could lead to a very strange WAR/140 mark.

Anyway, happy guessing: https://www.sporcle.com/games/striderremix/iwag-2020-pitcher-distributions-copy


Have absolutely no idea what’s going on here? Well, this is a four-sentence explainer: IWAG is a projection system I made a half-decade ago to try to better understand how projection systems work, as well as their strengths and weaknesses, by the process of building one myself. While current projection systems definitely generate outcomes probabilistically, they generally don’t share the probability distributions they generate (for good reason). Each year, I put together a series of posts detailing the IWAG outputs, both distributions and point estimates, for the Braves in their upcoming season. I’m still doing that this year, but since I had all the distributions generated anyway, I threw them together into a quiz on a lark, and now you can waste your time on that quiz as well. Good luck!