THE BOOK--Playing The Percentages In Baseball

I have looked at something like this for my projections.  Take players who play 162 games one year and see how many they play next year.

My theory is that this will give me the upper limit on durability.  The result is about 150 games in year 2.  So I don’t project any player to have more time than that.

Taking anything less than a full season and you’ve got to worry about the player not playing as much because he’s just not very good.  But if a player plays all 162 in year 1, then I assume that the team believes he’s a valuable player, not in danger of losing his job, and any fewer games the next year will be because of injuries.

The 12MM$ man TWICE played 162 games.  Let’s just call him the exception to the rule.

You mean Frenchy, right?

He followed the 162 with another 162, and then 155.  He doesn’t break the rule.  Regardless of what we think of his talent, his team thought he was good enough to play every day.

Juan Pierre: 162, 162, 162, 162, 162, 119.  Benched solely for ineffectiveness.  Ironic, of course, because he was just as worthless as usual.

Any chance we could get error bars on those? Like a box and whisker plot? That would show (I think) that the older a player gets, the bigger the downside (even if the average is only slightly lower).

@Sal - sample size is a little small.  We only have 784 guys - 1995-2009 (excluding 2004.)

For guys 19-29:

Median: 78, Avg 73, Std 12.7

For guys 30-37:

Median: 77, Avg 69, Std 16.6

I could add 1979-1993 (80-game season, little expansion) to the sample, but I think we’d still be a little on the low end...Let me take a look.

Ok, I added 1979-1993:

19-29: (25th%/50th/75th/avg)

70/79/82/73

30-37:

64/75/81/69