Jon Shepherd: The first half does not matter

It is natural to think of the most recent occurrences to be the product of a new realization. That is, if a player all of a sudden has a monster first half or a terrible first half, then that player is defined by that play. However, if the only bits of information you have are projections from the beginning of the year and how a player has performed so far, then you can ignore the latter and focus solely on the former.

In other words, it is more likely that Nelson Cruz’s preseason projection (ZIPS, .259/.316/.469) is more accurate than what he has done so far this season (.273/.344/.541). That is a pretty obvious statement. Three or more seasons of data will tend to better describe a player’s talent level than what 100 games would describe, even though you might feel different.

Perhaps an easier example would be one that has already happened. Last year, the Orioles made a splash and acquired Scott Feldman. Some folks thought his addition in exchange for Jake Arrieta, Pedro Strop and some international signing bonus money was a great idea based on his 3.46 ERA along with some imagining that he had finally turned the corner.

That is something that you might hear a good bit about fringe trade candidates. They might be re-experiencing past success or maybe they discovered something new and added it to their mix. Sometimes, these narratives are true. Sometimes, we simply like to have a good story to go along with data to make sense of it.

We want patterns and sometimes those patterns do not exist. We want to believe that everything means something in baseball. That the best players play well. That the best teams win. Alas, baseball is dirty with observational noise and sometimes data science can cut through that. Back to Feldman, ZIPS last year thought Feldman was a 4.31 ERA pitcher, which was quite different from his solid performance before he was dealt to the Orioles. Once in Baltimore, he pitched to the tune of 4.27.

Now, sometimes a change in performance is in line with a change in ability or talent. That can happen. It is just that it typically does not. Mitchel Lichtman, a data science baseball consultant for the past 25 years who is probably known best for taking existing ideas and developing Ultimate Zone Rating, wrote columns on hitting and pitching that illustrated that exceptionally good or bad performances in the first half are less predictive than preseason projections. In other words, a player is more likely to be what we thought he was as opposed to what he is presenting himself as. That is a pretty interesting idea.

A couple of weeks ago, the Depot highlighted how Baseball Prospectus’ playoff odds saw Baltimore as having the worst second-half record in the American League East. This was expected to be met with derision because it goes against human nature. We tend to believe what we have in our hand as opposed to what we are supposed to have in our hand. The lottery winner is happy he bought his ticket. He tends not to think that buying lottery tickets is a bad idea, which it is.

What this all brings us back to is trying to figure out the best way to improve this club. If the Orioles’ front office has special information to believe in players who are experiencing cold or hot streaks, then great. That said, they whiffed on Feldman last year, so they appear imperfect at being able to do that with pitchers from other clubs. That whiff on second-tier pitching last year makes it somewhat concerning that (outside of Jon Lester) they have been tied to more second-tier starters this year. When we discussed what each available second-tier pitcher could provide, none of them provide much in terms of improving the club over what the club already possesses.

But it is good to remember that the Orioles are indeed in first place no matter how they have gotten there. They have survived Chris Davis being the team’s worst hitter in July. They have survived Nelson Cruz being outhit by Caleb Joseph in June. The record may have a dash of “Orioles Magic” in it. That is OK. For the second half though, hopefully the club can rely more on consistent performance as opposed to clutch performance.

Jon Shepherd blogs about the Orioles at Camden Depot. Follow him on Twitter: @CamdenDepot. His thoughts on the O’s appear here as part of’s continuing commitment to welcome guest bloggers to our little corner of cyberspace. All opinions expressed are those of the guest bloggers, who are not employed by but are just as passionate about their baseball as our roster of writers.

blog comments powered by Disqus