THE BOOK--Playing The Percentages In Baseball

Related article from John Walsh at THT (January 2007):

http://www.hardballtimes.com/main/article/americans-defeat-nationals-in-pitchers-duel/

I am not sure I get his methodology.  In fact, I have no idea what it is.  I wish he would explain it and give an example.  Since we don’t know about the parks (and other external things that affect run scoring), when we do league comparisons, we have to use metrics that are relative to league average.  And we also have to account for familiarity factors.  I am sure he did all that, but I have no idea what this means:

To answer I looked at all pitchers and all hitters who appeared in each league within a decade and figured the RC/game for hitters in each league, and RA/game for pitchers, matching by the lessor total of plate appearances.

I am a little surprised that there are such big differences across decades.  I would have expected near parity over a decade even though the balance may fluctuate a lot from year to year.

Also, college teams go 1-161 versus a MLB team?  Does not sound right to me!  I guess it depends on whether you mean all colleges or DI (or DI and DII) only.  16-1 average score?  That also does not sound right.  Some of the top colleges have top pitchers who are close to major league ready, or at least near replacement level.  They would be pitching every 5 games and winning occasionally.

Maybe I am wrong, but I found it humorous that Rally uses the 1-161 record as an example of “passing the smell test.” If 1-161 passes the smell test, what wouldn’t (at one end) pass it?  Losing 200 games in a 162 game schedule?

A team of average MLB pitchers, as hitters, would score 1 run per game.

I’m not sure what a typical college player is, but that probably is around reasonable I guess.  Are college pitchers around a league average hitter?  That is probably as good a baseline as we can use I guess.

Since Rally presumed a typical (median as opposed to average I’d say) player, this would knock out the David Prices of the world from the competition scene.  It’s not like we are choosing to pitch David Price one time per 1000 starts, and the second best college pitcher one time, etc.

And even if we did, if they are only being supported with 1 run per game, it’ll be hard for them to win a game.

I agree with MGL’s basic point about the smell test.  A team that scores 2 and allows 10 will win 5 games or so.  That’s a huge difference in performance numbers, but tiny difference in wins.  So, I’m not sure that really tells us anything.

In early spring training teams will play exhibition games against college teams.  I don’t have a comprehensive record but going by my (admittedly flawed) memory the MLB team usually wins by 15-20 runs or so.

I think your college team is too low compared to low A. A good college team is going to be relatively close to low A ball.

I looked at the Baseball Cube for the SEC for 2002, figuring that was enough time to see who went pro.

http://www.thebaseballcube.com/leagues/2002/SEC.shtml

12 teams - 127 guys played at least A ball - 10+ per team - with 4 in majors, 14 in AAA, 20 in AA.

Age adjust, but you got a league that is much closer to .50 than .25

You probably need to define college teams. There are 100s of Division 1 teams. For reference point, why don’t you try to scale the SEC and scale it from there. Maybe they are .40, the Big Ten .30, Ivy .20, etc.

I had always assumed that a decent college team was equivalent to A ball as many players in A ball are high school players, and not necessarily good ones at that.

If indeed ST teams are beating college teams by even 10 or 15 runs (on the average) then Rally is right.  The ST team that is on the field against a college team is probably close to a replacement major league team.  It’s not like they are going to pitch CC Sabathia for 8 innings in ST versus San Diego State.  But I doubt that is the case (that the average score differential is 10 or 20 runs).  But you never know.  Like Bill O’Reilley, I am occasionally wrong.