Snow Day Nats Stats: 2009 Run Expectancies

While we wait for pitchers and catchers to report, here are some long-overdue stats to wrap up last season.  That's right, more run expectancy matrices! Yay!  What are run expectancy matrices?  Despite the intimidating name, it's basically a table that that tells you how many runs you score in various combinations of outs and men on base.  Here's the 2009 run expectancy for the NL:

What does this mean?  Well, every time an NL team had no outs and no one on base, they scored an average of 0.48 runs before the end of the inning.  Similarly, when an NL team came to bat with 1 out and runners on 1st and 3rd, they scored 1.13 runs on average--and so on, for all the other combos of runners and outs.  The interesting part is to pull out the data for the Nats and see how they compared!  I'll give you a hint: "The Nats leave 'em loaded! (With a twist.)" The numbers don't lie (after the jump).

Here's the Nats' 2009 run expectancy:

Something interesting jumps out right away: look at the average runs scored with no outs and no one on (that is, the start of the inning).  The Nats scored 0.47 runs per inning in 2009, which is essentially league average.  Are there any particular oddities?  Well, the last time I did this (around the AS break last season), the Nats had trouble scoring with 2 outs, at least compared to league average.  Looks like that pattern continues: the Nats score fewer runs than league average in most of the 2-out baserunning combos, and about the same in the rest.  How much faith can we put in these stats, though?  Here's a table of how many times the Nats were in each base/out combo during 2009 (out of a total of 6,273 PA).

To compare to league average, here's the percentage of Nats' PA for each combo (in other  words, the run expectancy table says the Nats scored 0.48 runs, on average, after a PA with no out and no one on... they had 1,512 PA with no one out and no one on in 2009, which was 24.1% of their PA).

And here's the NL average:

We can add up the numbers in a few interesting ways:  73% of Nats PA were with no runners or a man on first, versus 72.8%  NL average.  Oddly, the Nats had 19.1% of their PA with a runner on first, versus 17.9% for league average. The Nats had a runner in scoring position 27.0% of the time (27.2% NL average).  Most damning, the Nats loaded the bases for 3.1% of their PA versus 2.8% for league average.  Given that the Nats' bases-loaded run expectancy is about 5% below league average, I think the extra bases-loaded PA came from loading the bases and failing to score, rather than from loading the bases more often...

Update: A walk ain't a granny, but it still counts

PhDBrian notes in the comments that the Nats seemed to walk quite a bit last year when the bases were loaded.  I took a look at the Federal Baseball stat-o-matic (otherwise know as baseballreference.com) and found the following:

The Nats walked 22 times in 197 PA with the bases loaded (11.2% of the time).  League-wide, NL teams walked 227 times in 2759 PA with the bases loaded (only 8.3% of the time).  The secret to success?  Walk 'em in!  Note that the Big Walky had five of those RBIs the (very) Easy Way.  Of course, the Nats still scored fewer runs than NL average this way--a walk only brings in one, and any hit other than an infield single is likely to score two.