NL vs. AL: Which league is better?

No brainer, right?

Eric Karabell said yesterday (7/7/08 ) on his podcast that the AL was obviously the better league — I think he used the word “clobbered.”

And why not?  This year’s interleague record was 149 for the AL, to 102 for the NL.  That gives the Senior League a 0.396 winning percentage.  Friends, that’s close to Washington territory. (*)

*By the way, why did Jim Bowden think that the 2003 Reds would be any better in 2008 than they were back then? Dmitri Young, Kearns, Lopez, Aaron Boone — playing the Nationals is like Old Home Week.  Also, I’m stealing this footnote technique from Joe Posnanski.

Add to that the fact that the AL has won 8 of the last 13 World Series, and 10 straight All-Star games.  Case closed, right?

Not so fast.  When you look at the underlying numbers, it’s much less clear.

League	Runs	HR	OBP	SLG	OPS
American League	409	85	0.332	0.410	0.742
National League	401	90	0.329	0.410	0.739

Gee, that looks a little closer, doesn’t it?  Those numbers are per team, to compensate for the fact that there are 16 NL teams and only 14 AL teams.

It turns out that the difference is barely statistically significant.  In runs, it’s about 1.1 standard deviations; the NL advantage in HR is about 1.5 standard deviations.  The AL numbers are better, but only slightly better — not the kind of difference you would expect would generate the lopsided interleague results.  Furthermore, we would expect the AL to be better because of the DH.  Shaving 1/9 off the AL run numbers would put them lower than the NL.

Well, lets look at the players, then.  From Fangraphs.com, the 10 best players this year in WPA are Berkman, Burrell, Holliday, Manny, Bay, Pujols, Mauer, Chipper, Uggla, and Carlos Lee.  That’s 8 NL players.  The top 10 (actually 12, because of ties) in Runs Created from baseball-reference.com:

Berkman-HOU	95
Jones-ATL	88
Kinsler-TEX	86
Utley-PHI	83
Pujols-STL	80
Ramirez-FLA	79
Burrell-PHI	76
Holliday-COL	74
Sizemore-CLE	74
Bay-PIT	72
Bradley-TEX	72
Hamilton-TEX	72

That’s 8 National Leaguers of the 12.  Certainly looks like the NL can pull its weight here.

Maybe it’s the pitching?  Here’s Adjusted ERA+:

Duchscherer-OAK	194
Volquez-CIN	190
Lee-CLE	172
Lincecum-SFG	171
Danks-CHW	170
Haren-ARI	160
Sheets-MIL	156
Marcum-TOR	153
Zambrano-CHC	151
Billingsley-LAD	141

Only 6 of 10 here for the NL. But still! Around half of the best pitchers are in the NL.  And Santana, Sabathia, and Harden aren’t on that list (yet).

Jayson Stark had a chat about this topic after the first interleague weekend.  He took the position that the NL was at least closing the gap, if not even.  Of course, after that the NL really got clobbered in interleague play.

So, how can we possibly explain why the AL becomes unbeatable in interleague play?

Here’s the possibilities I’ve come up with.  Maybe you can add more.

1. The AL pitching and hitting are both better than the NL by precisely the same amount.  Thus, they cancel each other out in AL games, but show up big time in interleague.  But then how do you explain the preponderance of NL players atop the leader boards?
2. The NL has the stars, but the AL teams are better balanced.  The balanced teams win.  Okay, but is there a reason why we think a balanced team is better than a stars-and-scrubs team?  Also, apparently the stars and scrubs just happen to balance out to give PRECISELY the same overall stats as the even-keeled AL.  Sounds fishy.
3. AL teams make better use of the DH, because they actually have a good hitter on the payroll to play the position.  The NL teams are using utility infielders.  Two words: Jose Vidro.  Besides, that should only matter in half the games where the AL is at home.
4. The AL has more big-name aging stars.  The kids in the NL are simply awed, and fold.  C’mon, it’s the major leagues.  They get over it pretty fast.
5. The leagues are actually pretty equal; it’s just been a string of bad luck for the NL.  149-102?  That’s pretty unlikely — the binomial calculation puts it at 0.15% chance if the true odds are 50/50.
6. AL Three-True-Outcomes ball beats NL small-ball. Except that the NL has the lead in HR, BB, and K.
7. Wait, I mean that AL small-ball beats NL Three-True-Outcomes ball.  Um, now you are reaching.

It doesn’t make any sense to me.  The NL looks roughly equal to the AL.  If I didn’t know the head to head outcomes, I’d give the NL a slight advantage.  What gives?