The Baltimore Orioles: Some Further Anylisis

I wrote about the Baltimore Orioles on Monday. As it turns out, people have strong feelings about the Orioles. Most of the comments I got fell into one of these categories:

– The writer is an ass-clown, the Orioles are winning with hustle and poise and grit and spunk, and you “pencil neck baseball anylists” can take your WAR and shove it somewhere dark and humid.

– The writer is dead-on, the Orioles are winning with a dwindling stash of magic pixie dust, and you O’s fans better enjoy it now before your team CRASHES AND BURNS SUCKERS!!!1

– The writer is engaging in something called ‘satire,’ and evidently it worked.

For the record, I really like the Orioles. Clearly they are experiencing some luck, and they are also doing a lot of things well. I think they are an interesting outlier and I hope they continue to, um, outlie. It’s fun and interesting when our best knowledge turns out to be not perfect. Right?

So how far out are they lying, really? Well, here is what I did, and I can get away with it because, see, I’m on Notgraphs. What I did was, I went back ten years and looked at the difference between every team’s WAR and its actual wins. The MLB average figure for this difference — the replacement-level record, you might say, for a given year — is always between 42 and 44 wins. (Which kind of speaks well for WAR, right?) So for the average team, you take 42-44 wins, depending on the yearly average, and tack on its total WAR, and you get something very close to its actual record. But teams very often overperform or underperform that number by a little bit, thanks to managing and pixie dust and whatever else. Every year, a couple of teams deviate from that number by a good bit, winding up with 5-10 wins more or less than what they “deserve” based on WAR. (I’ll call that number WTSNHW, or Wins They Should Not Have Won, in keeping humorously with my last post.*) And once in a while, a team really overachieves and ends up with a WTSNHW of 10+. Here is that list, for the last decade:

YEAR | TEAM | WTSNHW
2008 | Angels | 16.4
2012 | Orioles | 16.3
2003 | Athletics | 16.2
2005 | Yankees | 15.2
2007 | Diamondbacks | 14.1
2004 | Yankees | 12.7
2006 | Athletics | 12.6
2007 | Mariners | 12.4
2005 | White Sox | 11.6
2003 | Royals | 11.6
2002 | Braves | 10.8
2004 | Reds | 10.6
2002 | Athletics | 10.1

There are some well-known overachievers here, notably the Moneyball A’s of 2002 and ’03. There are some more surprising entries, too. There are bad teams on here, and good teams. Most of them had good to excellent bullpens, and probably closed out a lot of one-run wins, O’s-style. Nearly all of them, interestingly, had stronger pitching than hitting. (An exception being the ’04 Reds, who had horrible pitching, mediocre-at-best hitting, and somehow managed to win 76 games anyway.) Some of them may have been unfairly punished by WAR for poor fielding, like the ’05 Yanks, who posted an astonishing -140 fielding runs thanks to UZR hating their entire outfield and middle infield (but won 95 games). I’m guessing that many of them, like the O’s, had relatively erratic starting rotations, capable of giving up a ton of runs one day and throwing a gem the next.

All of this tells us something about WAR, and if I weren’t writing on Notgraphs, I might actually put some effort into figuring out what that something is. Anyway, the main point is that, with a month left in the season, the 2012 Orioles are already very close to being the biggest overachievers in recent memory. It’s not totally unprecedented. It’s not totally inexplicable. It’s just interesting and cool. At least, I think it’s pretty cool. I think I’ll be casting a vote for Showalter for MoY, and also for President of the United States. And, as a small gesture toward O’s fans, I think I’ll make a photomosaic. Look, tiny sexy Jim Palmers! Please don’t hate me.

* You’d get similar results (though far from identical!) by looking at run differentials. But since this has become a referendum of sorts on WAR, I figured I’d go ahead and ride that horse.




Print This Post



Sort by:   newest | oldest | most voted
Josh M
Guest
Josh M
3 years 8 months ago

Jim Palmer approves

Illinois glass M. Michael Sheets
Guest
Illinois glass M. Michael Sheets
3 years 8 months ago

I thought the comments in the first post would have made you understand by now that you southerners just can’t understand the Orioles. Try solving Mississippi’s problems, like New Orleans, before trying to tell me how to feel about the Orioles.

GiveEmTheBird
Guest
GiveEmTheBird
3 years 8 months ago

Well said.

samuelraphael
Member
3 years 8 months ago

Baltimore should have to vacate the wins they should not have won.

GUI
Guest
GUI
3 years 8 months ago

NL Central division teams really earn their keep… only a paltry 10.6 wins to vacate

Boss Sauce
Guest
Boss Sauce
3 years 8 months ago

The Orioles’ WTSNHW since 2004:

2004:-5.3
2005:-2.8
2006:-2.2
2007:-3.7
2008:-1.4
2009:-1.3
2010:3.5
2011:3.1
2012:15.6

Now I’m no fancy mathematician, but that’s a trend. R-squared is .69, which means we can explain 69% of the WTSNHW using the year. By my calculations, the Orioles’ WTSNHW will be a robust 31.9 by 2025. By about 2070, the Orioles will need only field a replacement level team to win every game. Simple statistics, folks.

John Thacker
Guest
John Thacker
3 years 8 months ago

Clearly Buck Showalter just Knows How To Win.

MikeS
Guest
MikeS
3 years 8 months ago

What you have suggested is something that epidemiologists and others that use statistics swear by – you learn the most by looking at the outliers. You learn about them, about the non-outliers and you learn about how to refine your model.

Greg W
Guest
3 years 8 months ago

But, really, where can I get free software to make a photomosaic?

rokirovka
Guest
rokirovka
3 years 8 months ago

I think you hit the nail on the head when you point to the Orioles’ “erratic” starting rotation, “capable of giving up a ton of runs one day and throwing a gem the next.”
Indeed, the Orioles have had more “outlier” pitching performances than any other team in the AL. Define an outlier as a game where a team gives up either 2 or fewer runs, or 6 or more runs.
For most teams, the total number of such games is remarkably consistent throughout the league: To date in 2012, AL teams have an average of 84.3 such games, with a very small standard deviation of only 3.6. Almost all AL teams have had between 81 and 87 such games.
The Twins have had 80, barely outside the standard deviation. The Indians have had 90, a standard deviation and a half above the average.
But the Orioles have had a whopping 93 such games, *two and a half* standard deviations above the AL average! Their 48 games giving up 6+ runs have made their overall stats look very poor, but their 45 games giving up 0-2 runs have boosted their actual W-L record quite a lot.

Phil Castle
Guest
Phil Castle
3 years 8 months ago

If you use mWAR, you’ll find that wins align perfectly … Of course.

wpDiscuz