---

Thursday, September 2, 2010

Are the 2010 Yankees Better Than the 2009 Yankees?

Today’s win against Oakland was the Yankees 134th game of the season.  They now sit at 84-50, one game ahead of Tampa Bay in the loss column for the AL East division lead.

The Yankees have scored 726 runs this year and allowed 545.  That works out to a Pythagenpat winning percentage of .633, compared to their actual winning percentage of .627.  Their Pythagenpat record of 85-49 is one game better than their actual record.

After 134 games in 2009, the Yankees were 86-48. 

They had scored 773 runs and allowed 627 runs.  That works out to a Pythagenpat winning percentage of .601, compared to their actual winning percentage of .642.  So their Pythagenpat record of 81-53 was about five games worse than their actual record.  Of course, that could just be the Wang/Claggett factor.

Here’s how the two teams compare on a per-game basis through 134 games.

Year	Gms	RS/G	RA/G	AL Avg
2009	134	5.77	4.68	4.82
2010	134	5.42	4.07	4.46

RS/G: Runs scored per game
RA/G: Runs allowed per game

We shouldn’t really do a direct comparison here because there’s a fairly significant difference in the run environments of the 2009 and 2010 American League.  The average AL team scored and allowed 4.82 runs per game last season, compared to 4.46 so far in 2010.  So here’s a comparison of the 2009 and 2010 Yankees through 134 games relative to their respective run environments.

Year	Gms	RS/G	RA/G	AL Avg
2009	134	1.20	0.97	1.00
2010	134	1.21	0.91	1.00

What this shows is the ratio of Yankees RS/G and RA/G relative to league average. 

So after 134 games in 2009, the Yankees were scoring about 20% better than league average and were allowing runs at a rate of about 3% better than average.

In 2010, the Yankee offense despite scoring fewer runs per game this season have actually been 21% better than league average.  But the real difference so far is that the Yankee run prevention has been 9% better than league average.

So if you look at Pythagenpat and runs scored/allowed relative to league, the 2010 Yankees have performed better than the 2009 Yankees.

It’s not that simple though, at least not to me.  To me when we talk about something being better than something else, you’re not just talking about what happened, you’re also talking about ability.  So if we really want to see if the 2010 Yankees are better than the 2009 Yankees, we should look a bit deeper into the offense and pitching/defense to see if the runs scored/allowed are skewed by things that may or may not be repeatable.  This could be due to different performance with men on base or fluctuations in BABIP on offense and defense, or defensive support being better or worse.

Through 134 games the 2010 Yankees have hit .270/.349/.439, and their offensive performance adds up to 704 context-neutral linear weights batting runs, compared to their 726 actual runs scored.  So if you go by linear weights instead of actual runs scored, the 2010 offense has been more like 18% better than league average.

Contrast that with the 2009 Yankees who had hit .282/.360/.479 which was good for 793 context-neutral batting runs compared to their 773 actual runs scored.  That would have made them about 23% than league average.

The difference there is probably due to a slightly better performance by the 2010 Yankees with runners in scoring position.  They’ve also hit into 18 fewer double plays than their 2009 counterparts had through 134 games, although my linear weights formula does factor in double plays.

I don’t think I’d disagree that even if the 2010 Yankees are scoring runs at a higher rate relative to league than the 2009 Yankees did their offense isn’t quite as good, but the gap is certainly smaller than I’d intuitively thought it was.

The big argument for the 2010 Yankees being better than the 2009 Yankees lies in run prevention.  Allowing runs at a rate of 91% of league average compared to 97% of league average is a pretty significant improvement.  Again though, we probably need to look a little deeper into the pitching staffs and their performances to see if they’re actually pitching better.

Year	RA	ERA	FIP	CERA
2009	4.69	4.31	4.46	4.56
2010	4.13	3.86	4.36	4.51

RA: Runs allowed per nine innings
ERA: Earned runs allowed per nine innings
FIP: Fielding independent pitching
CERA: Component ERA

Despite the big difference in RA and ERA between 2009 and 2010, looking at the peripheral stats shows that the pitching hasn’t necessarily been that much better than this year than it was last year.  Here’s how these stats look relative to league average.

Year	RA	ERA	FIP	CERA
2009	0.97	0.97	1.00	1.02
2010	0.93	0.93	1.05	1.08

Interestingly, the component stats show that the 2010 Yankee pitchers haven’t even been as good as the 2009 pitchers were, at least if you trust FIP and CERA.  While we shouldn’t necessarily assume all of the divergence between FIP and ERA is due to luck in BABIP, if component ERA says the same thing as FIP then it’s a bit more likely that there’s been some good fortune in the 2010 Yankees pitching performance, whether it be with BABIP against or in pitching better with runners on base, or in strand rate, or in a myriad of other things.

Of course, that doesn’t mean that the Yankee run prevention as a unit hasn’t been better, because we still need to consider defense.

Year	ZR	DRS	UZR	AVG
2009	-1	8	-5	1
2010	-6	9	1	1

ZR: Run saved compared to average using Chris Dial’s methodology for converting standard zone rating
DRS: Defensive runs saved using John Dewan’s plus/minus methodology
UZR: Runs saved compared to average using Ultimate zone rating

I’ll mention the fact that defensive metrics are still a bit wonky, so even though these numbers say there’s not much difference between the two defenses that’s not necessarily true.  Replacing Johnny Damon in LF with Brett Gardner and adding Curtis Granderson in CF has sure looked like an improvement to me, and I think Nick Swisher’s been better defensively this year too.  The primary areas that have looked worse defensively have been catcher and shortstop, but I think that’s been mitigated elsewhere.  I think this year’s defense is better than last year’s, although I don’t think the gap is that big.

With 28 games left in the year, a lot can change, and we can probably revisit this question.  Should the Yankees make the postseason by some miracle, it’ll be interesting to see how the 25 man postseason roster compares to the 2009 Yankee postseason roster which was stronger than the overall 2009 team.

Right now, I’d still probably say the 2009 Yankees were better, but the gap is smaller than I thought.