In advance of the Baseball Writers’ Association of America announcing the winners of the American and National League Rookie of the Year awards on Monday afternoon, SB Nation released its own awards based on the votes of its writers.

Not all that dissimilarly from our own Getting Blanked Awards (The Blankies), SB Nation’s fine scribes voted for Michael Pineda to receive their version of the American League Rookie of the Year award. However, the handing out of the pretend silverware wasn’t without its share of deep thought.

From Jeff Sullivan of Lookout Landing:

As I understand it, the Rookie of the Year award is supposed to go to the league’s best rookie. Consensus seems to be that “best” is some combination of performance and playing time. This is why Brett Lawrie doesn’t show up at the top of many lists. But why should playing time be that important? Brett Lawrie came to the plate 171 times and hit .293/.373/.580. That is an outstanding performance. An outstanding performance over a limited sample, sure, but a more outstanding performance than any other AL rookie, as far as I can tell. Why shouldn’t he get more consideration for the award? It isn’t the AL’s most valuable rookie. It’s the AL’s best rookie.

There does seem to be more attachment to playing time when it comes to the Rookie Of The Year award than other awards, and when you think about it, it makes sense. The idea in any award is to recognize the display of some level of true talent, and the more examples of that individual using his true talent, the easier it is to award him for it.

Rookies don’t have the same history that other players have and so the total amount of games that they played in their initial season is important in that, the more there is, the more likely it is that we’re seeing a true talent level. It’s easier to award the MVP to Josh Hamilton when he only plays in 133 games because we already know that Josh Hamilton is an excellent player based on the 335 games he played previously.

As Tom Tango explains:

If you had two plate appearances for Lawrie, then we attribute very little of the OUTCOMES to Lawrie.  We just don’t know if he was being a good driver, or just happened to be in the driver seat at that moment in time.  If he had 20 PA, then we attribute more of those outcomes to Lawrie.  If he had 20,000 PA, then we’d attribute 99% of each of those outcomes (including those first 2 outcomes) to Lawrie.

It seems that the dependence on games played when it comes to the Rookie of the Year voting is one of the rare occurrence when the BBWAA has got it right. Playing time might seem like the sort of thing that’s erroneously considered around this time of year, but for Rookie of the Year specifically, it makes sense to have as wide a sample as possible.

Comments (10)

  1. I think the cogent counter-argument would be that if a player is essentially disqualified because the sample size is too small, he should still be eligible next year.

  2. But the thing is, I don’t think there’s an exact number. It’s just a comment that it’s easier to say this guy is talented after he has success in 700 plate appearances versus someone else who puts up better rate numbers in 250 plate appearances. It’s unfair to the actual player, but that’s all we have to go by.

  3. But there should be some way, a formula to say if a player should be eligible or not. Like if you’re trying to determine if a gambling six sided die is loaded or not, you throw it X times and then based on the results you can say with Y certainty that it’s loaded. How good would Lawrie have had to be to win the award in his number of PAs?


    500PA’s perhaps? Would basically mean a more or less full season.

    • John,

      That misses one wrinkle, that in this case we’re not looking for a reliable sample size for an accurate assessment of the player. We’re looking for a point where we can say that player A was better than player B. That depends both on the sample size and the level of performance taken together.

  5. I don’t entirely buy this argument. If you’re familiar with null hypothesis significance testing and the idea of type I error, you know that there is generally a magic sample size number at which the odds of getting a false positive (that is, attributing an outcome to random chance when it is inappropriate to do so) is incredibly low (less than 5%). So, what constitutes an appropriate sample size if you’re trying to quantify who had the better year and not who projects as the best career player? I’d guess that from a purely statistical perspective, there’s not much difference between 200 PAs and 400 PAs.

    If you’re looking at WAR, how do you then compare Lawrie with Dustin Ackley, who put up the same fWAR over almost double the playing time? Does consistency over a larger sample win out, or do you look at the fact that it took him twice as long to put up a (roughly) similar line? What about comparing Pineda to, say, Craig Kimbrel who was also around a full season but pitched far fewer innings?

    I agree that the sample size for Lawrie is probably a little too small, but guys have anomalous years all the time.

  6. Dammit Parkes, you should know better. WAR is a cumulative stat, and is measured against a replacement player baseline; unless we think that the best rookies are likely to regress below replacement level, WAR is a conservative measure for evaluating who the best rookie is.

    A guy with 3 WAR in 400 PA has the same production as a guy with 3 WAR in 200 PA + 200 replacement level PA. The better player (rookie, whatever) is clearly the second guy.

  7. I gotta say that I did not think Mr Cosby had the time to study statistics when he was becoming a comedian, actor, and obstetrician. Bravo for this ananlysis.

  8. fuck off parkes

Leave a Reply

Your email address will not be published. Required fields are marked *