Bobby Witt ZiPS: How to think about a projection

Dan Szymborski put out the Royals’ ZiPS projections at the end of last week. I have been waiting for them, so of course the Royals were the 27th team to be covered. ZiPS is one of the best projection systems around, but I feel like people misunderstand projections systems and how to read them. This is partially a problem of how they are presented. I would like to dig into the Bobby Witt Jr. projection in particular and go into some depth about expectations we should have for him based on this, and other, projections for 2025.

Most people see this graphic, and think that is their entire interaction with projections for the year. It is easy to look at this and say that Dan thinks Bobby will have 6 fWAR and that is that. Luckily, Dan is much more thorough than that, read the writeup linked in the first sentence of this article to get all the details. He lists out the 20th and 80th percentile projections for each player in that article, a much better way to think about projection. In fact, I would prefer all projections be laid out as confidence intervals, but I understand why they are not.

Witt’s 6 WAR projection might be disappointing coming off of a 10.4 fWAR season. Here is what the uninitiated are missing, the 2024 ZiPS projections going into last year had only one player in all of the MLB projected for more than 6 WAR, Ronald Acuna at 7.4 wins. The 6+ win threshold is reserved for MVP candidates only. This year, Bobby is being joined by Ohtani (9.2 wins), Judge (7.9), Henderson (6.1), and Lindor (6.0) make up the entire group at or above 6 wins in the projections this year unless someone on the Padres surprises as they have not dropped yet. Only five players, no Juan Soto or Elly de la Cruz. Bobby is now in the elite.

Do I think 6 wins is too low for Bobby’s projection? Yes I do, but let’s talk about the range of outcomes. The 20th and 80th percentile projections for him have a fWAR between 3.9 and 7.7 wins. His slash line ranges are .261/.315/.462 and .314/.371/.582, which puts his 80th percentile projection as a hitter almost exactly where he was last season minus just a little in each of the three slash categories. So if he hits almost as well as last season, the projection says his WAR will drop by over 2.5 and that it is undervaluing his defense.

ZiPS uses , and Witt has some defensive struggles in his rookie campaign. Since then he has been one of the best defenders in baseball, and I expect that to continue. But his defensive prowess won’t be fully acknowledged by ZiPS until next season. Also, his hitting is still reflecting a lot of his early struggles. Bobby has had a split career so far:

He has hit at the 80th percentile projection for the last thousand plate appearances. There is no guarantee that next year he will have another 160 wRC+ season, but I would be surprised if he fell all the way to the 137 that is the midpoint of the projection. His worst month last year was at 126. Then again, even a seeming machine of output like Jose Ramirez has had his wRC+ drop to 108 in 2019 and 122 in 2023, so even the best hitters have those 20th-percentile outcomes sometimes.

It is fun to have the three best shortstops all on top of each other in the projection. Francisco Lindor, Gunnar Henderson, and Witt are all coming off of MVP-caliber seasons, though none of them won the award. Lindor has been a model of consistency, producing year after year since he showed up in 2015. He does derive a lot of his value from defense though, and has never reached the heights of the other two with his bat.

Henderson is the opposite, defensively he is okay, and his value is mostly in the bat. Bobby of 2024 matched Lindor defensively and Henderson offensively and I think this year will become the clear answer for best shortstop in baseball. Elly De La Cruz should get the dark horse mention here, though everyone is aware that he is a special player so maybe not a very dark horse. Lindor and Henderson’s 20/80 is narrower than Bobby’s at 4.2 to 7.2 fWAR for Lindor and 4.3 to 7.8 fWAR for Henderson. I understand why the low end on Witt is lower than theirs, but his high end should be higher and is not. I say that because his low and high years are below and above both players. More volatility and upside should make his confidence intervals wider on both ends and yet his intervals seem to be constrained at the top. I do not know enough about how Dan sets the simulation up to explain this to my satisfaction, but I do think Bobby’s 80th percentile projections are too low across the board. The low end seems pretty reasonable.

One of the other little nuggets in Szymborski’s writeup is top near-age offensive comps. Bobby Witt’s comps now stand as Francisco Lindor, Alex Rodriguez, and Dickie Thon, which is kind of awesome. He is a better hitter than Lindor and just needs one more year of proving it to move him out, but if Bobby is Francisco-esque over the next six years you will hear no complaints from me. A-Rod, on the other hand, is one of the greatest hitters to ever play baseball and only moved off of shortstop because of Derek Jeter’s ego. If Bobby hits anywhere near Rodriguez levels while playing Gold Glove shortstop for the next several years, Witt will probably be considered the best player in baseball non-Ohtani division. If you are unfamiliar with Dickie Thon, at age 25 he looked like a HoFer and then fell off pretty hard only to put up one more year above 2 wins. Bad things can, and do, happen.

What you want out of a projection system is a well thought out and consistent approach. That is to say, the Bobby Witt Jr. Projection ZiPS is kicking out is what you want a projection system to kick out. It should not be trying to look at the first year struggles and writing them off as casually as I have in this article. If you use three years of data, then you keep using three years of data as long as that is the amount that is sufficient from a predictive standpoint statistically. Eventually those years will drop off and if the player has reached a higher true talent level, it will show even stronger next year. If players distribute across the 20 to 80th percentiles at the end of the year in the right way, the fact that almost none of them hit the projected WAR total midpoint exactly is actually a feature not a bug. Life and baseball just are not that predictable.