Last week Jordan Bastian, MLB.com’s Cleveland Indians beat reporter, explained his projection methodology for the team’s offense in a blog post title “Projecting the offense.” His system, as he explained it, involves taking a player’s average production over the last three seasons, drawn out to create a 145 game sample for each year, and then averaging this production against the player’s most recent season. He used career numbers, instead of three year averages, for players with less than three years of experience. After doing some number-crunching for the nine players likely to see the most playing time, he presented his projections alongside those of Bill James.
Bastian’s goal was to come up with a system for landing “within the ballpark of realistic expectations,” and in the aggregate I was happy he found reason to expect an improvement over last year. However, I wish he would have used these numbers as a platform for discussing why certain players are likely to stay within or transcend the realm of past performance. This step is what makes preseason baseball commentary so engaging; all the more so when the step is taken by somebody that has had the opportunity to follow the team day in and day out and get to know the players. After all, Bastian’s projection system isn’t sophisticated enough to consider how player production over the past few years may have been influenced by things like sample size, consistency of playing time, or the effects of injury. If a player is going to get better from one year to the next, by definition nothing in the statistics of yesteryear is going to indicate it.
I’m not a statistics-shunning old school baseball sentimentalist, but in my opinion any discussion of player expectations heading into a new season that is entirely formula-driven is severely lacking. Except when there are statistically insignificant samples, past performance should be the foundation upon which predictions are formulated, but it shouldn’t be repackaged and offered as a harbinger of things to come. This is of especially questionable value where there is no explanation of why the system has any predictive value at all. Many baseball commentators don’t ever make predictions, except in special circumstances where doing so is safe, or when their outlet provides enough space for qualifying explication. Perhaps they worry that going out on a limb will backfire and impact their reputation if and when the prediction doesn’t materialize.
Bastian’s system yields projections that are ultimately not distinguishable from James’ in any meaningful way. On the whole, they may be slightly less optimistic. Bastian’s average home run totals and batting averages for the lineup are slightly lower than James’. In fact, Nick Swisher is the only player Bastian projected would hit more than .002 higher than James projected. In addition to my earlier prediction that Lonnie Chisenhall would hit 20 home runs to go along with a .270/.325/.435 line (Bastian has him at 17 home runs and .264/.303/.426), there are at least three noteworthy instances in which I disagree with Bastian’s projections.
There are at least three noteworthy instances in which I disagree with Bastian’s projections.
- 1. Michael Brantley: Here are Bastian and James’ respective projections for Brantley:
Bastian: .280/.337/.390, 6 HR, 34 2B, 5 3B, 58 RBI, 67 R, 14 SB, 50 BB, 67 K
James: .279/.344/.379, 7 HR, 29 2B, 3 3B, 55 RBI, 78 R, 19 SB, 54 BB, 60 K
I think Brantley will hit .300 this year. His playing time has increased in remarkably consistent increments over the past three years, and he’s improved each step of the way. This past year, he had over 600 plate appearances for the first time, and he put together a .288/.348/.402 line. He’s tough to strike out (just 9.2% of the time in 2012), and FanGraphs shows he’s swinging at more pitches in the zone and fewer pitches out of the zone. His line drive percentage is also on the rise. I was impressed that he didn’t hit under .250 in any month last year. Ultimately, however, this is a gut level prediction based on having watched Brantley’s approach at the plate over the past couple years. His focus in the batting box is unshakeable.
- 2. Jason Kipnis: Bastian projects Kipnis will steal 30 bases; James projects 28. I’d be surprised if either comes close. I think he’ll steal about 20. Sure, he swiped 31 bags last year, and was caught just seven times. But in 254 games over parts of three minor league seasons (2009-11), he stole a total of 24 bases. Some commentators say that what we saw this past year was the real Kipnis, that until now his running game was affected by his transition to second base. While I think he’s a “dirtbag” and a tough out on the basepaths, I don’t think it makes sense to count on this kind of production yet – notwithstanding the effect Terry Francona may have on the team’s running game.
Last year’s success can, in part, be explained by the fact that teams didn’t know to expect this kind of aggressiveness from Kip. His scouting report didn’t emphasize his running game, since he hadn’t yet established himself as a stolen base threat. With this in mind, it makes sense that he was 20 for 21 in stolen base attempts in the first half, but just 11 for 17 in the second half. Bastian and James both have Drew Stubbs stealing about 30 as well, but Stubbs has done this three years in a row. I do think that the projections are right, however, in that we can expect between 15 and 18 home runs from Kipnis.
- 3. Mark Reynolds: Bastian’s projection system has Reynolds hitting 27 home runs and striking out 174 times. James has Reynolds hitting 32 home runs and fanning 201 times, a projection I find more likely. Reynolds had a poor spring last year that carried over into the regular season. He didn’t hit any home runs in April, and when it seemed he might have found his power stroke in early May, an oblique strain forced him to the disabled list. He ended the first half with just seven home runs, his lowest total in any half of a season since 2007. But he broke out of his power drought to hit 16 in the second half. On a year to year basis, Reynolds is actually very consistent from one half to the next; I think it is unlikely he’ll slump in the first half again this year.
Of course, he led all of baseball in strikeouts from 2008-10. In 2011, he led the American League, and was second in the majors (behind new teammate Stubbs). So I guess some would say his whole career is a slump. While it’s true Reynolds’ strikeout percentage last year was the lowest of his career (at 29.6%), I’m not convinced this is a trend. It’s possible his first half power outage inspired him to temporarily shift his approach at the plate to just try and make contact, but the decrease really isn’t significant enough to get excited about. We’re going to see plenty of home runs and plenty of swings and misses from Reynolds this coming year.