With Spring Training games beginning this week, the first stats of 2007 will start to roll in soon. How much should weight should we give to these numbers, though? Major league pitchers will be facing raw AA level talent, and top-level hitters will face pitchers with no MLB experience and very little "stuff". The inconsistency of talent in the lineups is quite significant.
Every year you'll hear a few managers insist that the performance of players in March won't win them a job. You'll also hear a few proclaim that the competition for a roster spot will depend on who performs in the spring. Who is right here, and who'd wrong? I think I can at least shed a little light on the subject.
I'm far from the first person to consider this question. Joe Sheehan summed it up fairly well in a Baseball Prospectus article from 2003. He concluded that stats in March are largely meaningless. You could break it down and see the quality of players that a specific pitcher or hitter faced, which would give some legitimacy to the stats, but you're still looking at rough numbers. But to what level should we disregard the stats from March?
If we can't trust the numbers at face value, can we at least trust them to serve as barometers of early season trends? Will a player who's hitting hot in March carry that through to April? Will a slumping hitter continue to slump once the season begins?
I took a look at all players who got at least 50 at-bats in Spring Training of March 2006 and accumulated another 50 in the regular season's first month of April. My list was 135 players long.
After comparing the stats from March with how each player began the year in April, I reached a couple of conclusions:
- One, that was more time-consuming than I anticipated.
- Two, Google Docs is very, very handy.
- Three, there seems to be a very small correlation between performance in March carrying through to April. It's so small, it's almost not real.
If you'd like a look at the numbers, here's a link to the Google Spreadsheet.
For the more visually inclined, here's a chart of the differences between AVG, OBP and SLG between March and April of 2006. The total at-bats in March and April increase from left to right, to help spread out the data so you can see the grouping.
I should probably provide some numbers, too, like correlation coefficients. Here they are:
Average R = -0.04667
On-Base R = 0.1308
Slugging R = 0.1513
OPS R = 0.1235
As you can see, that's not a very convincing case that players who hit well in March will stay hot, or those that start cold will stay cold.
For every Robinson Cano, who hit .326/.359/.465 in March and .330/.350/.450 in April, there are ten players like Bobby Crosby, who hit .277/.360/.492 and then .220/.260/.320 the next month.
I'd encourage you to go check out the spreadsheet I linked above. There are some interesting cases, like Casey Kotchman, Rondell White, and Ben Broussard, where their Spring Training was absolutely nothing like their April.
If you're not yet convinced that Spring Training statistics have little, if anything, to do with how a player is going to hit in the regular season, take this into consideration: Last March, Doug Mientkiewicz slugged .627. To quote Forrest Gump, that's all I have to say about that.
But, this is just my opinion. What do you think?
---
Credit where credit's due: Without David Pinto's wonderful Day-by-day Database, this wouldn't have seen the light of day.