Notes on Minor League Statistics

4787522686_6bdb6d5403_b== As we embark upon the Spec66 rankings ==

If you followed our rankings of the “most valuable” MLB players of 2013, we linked you up to our philosophical approach to baseball statistics: the mild protest against one-number stats and the specifics of our focus on the “Venn Diagram” overlap of “avoiding (hitters) or getting (pitchers) non-random outs” and “producing (hitters) or denying (pitchers) non-random offense.

And thus the “three numbers” approach:

  • Nelson Cruz: 98-115-113 (average at avoiding outs; above-average at producing offense; above-average overall)
  • Robinson Cano: 136-125-161 (well above-average in every way)
  • John Buck: 70-81-51 (bad at avoiding outs; below-average at producing offense; below-average overall)


For minor leaguers we use the same approach, focusing on the same “three number” concept, but with a few wrinkles.

1) It is very important when looking a minor-league stats to keep an eye on the “age-arc.”  Whether or not a player is at an “age-appropriate” level is very important.  The stats of a player who is considerably older or more experienced than his opposition are suspect.

The “standard” age-arc that I follow is:

  • Rookie or Short Season-A — 19
  • Low-A — 20
  • High-A — 21
  • AA — 22
  • AAA — 23

If a player is more than a year older than the above chart, I generally discredit the stats.  As such, once a player is over age 25, I generally stop looking.  Players who hang on in the minors at age 26 and above can start putting up impressive stats, but, in light of their experience, they don’t necessarily mean much (unless they dovetail with what came before).

On the other hand, stats posted by a player who is a year or more younger than his “standard” level are considerably more valuable.

I am going to try this year to indicate by a “+/-” notation whether a player is above or below “standard” age level.

2) A minor-leaguer needs to out-perform (in stat terms) the typical major-leaguer in order to show MLB potential.  What I mean is: stats that would, for a major-league player, indicate a league-average performance, are not really sufficient to indicate MLB potential in a minor-leaguer (even when produced at an age-appropriate level).

I go by a “rule of thumb” which is based on observation of hundreds of cases of minor-league stats, but which I did not systematically derive.  It is this: a minor leaguer in an age-appropriate league needs to demonstrate at least 115% of MLB average production in order to be indicating a reasonable likelihood of MLB success.

In other words, a guy putting up “average” numbers in the minors is not a good bet.  Guys who succeed in the majors were not “average” in the minors — they were above-average.

Thus, for minor-league stats, I set the “100” mark at 115% of the 10-year MLB average for that category.  So a player who hits exactly the 10-year MLB average in the majors would have a 100-100-100 result, a player posting the same numbers in the minors would have a line of 85-85-70 (it’s 70 because the 15% difference in each category cumulates in the composite score).

This is so that it can be “easy” to spot the guys with MLB potential: their scores will be over 100.

Guys scoring less than 100 have not really shown MLB potential (statistically) yet.

3) Hitters who go on to long-term MLB success almost always show that ability in the minors by age 23.  The same is not true for pitchers, by the way.  But I have yet to find a hitter who ended up with long-term success in the majors who didn’t show hitting ability in the minors by his age-23 season.  When a guy starts hitting well at 24 or 25, I don’t take it as seriously.

4)  All of the exotic splits and what-not that folks have come to enjoy for major-league players can be found, but they are not necessarily accurate.  The minor-league box scores are pretty accurate, but the “Gameday” data that goes into the splits and exotic stats is not.  Even at the AA level, we’ve found some pretty major issues.  So apply skepticism to any splits or stats that are dependent on non-traditional record-keeping like “line-drive %” or “swing-and-miss %.”


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s