Donaldson’s Latest Train Wreck
I generally don’t like picking on such low-hanging fruit but let’s be honest. The Providence Journal’s Jim Donaldson is a first-rate horse’s ass whose disdain for the Red Sox and dimwitted analytical abilities have both been well documented over the years. His latest effort includes a back-of-the-envelope, position-by-position look at the Red Sox lineup changes and how the respective players can expect to fare in comparison to the player that held the position in 2005. It’s painful stuff, and a quintessential example of the shoddy work being trotted out by so many of the mainstream papers. I’ll take it bit by bit.
Instead of Kevin Millar at first base (.272, 9 HR, 50 RBI, .355 on-base percentage, in 134 games), we have Kevin Youkilis (.278, 1 HR, 9 RBI, .400 OBP, in 44 games) and J.T. Snow, recently of the San Francisco Giants, for whom he batted .275, with 4 HR and 40 RBI, in 117 games. His OBP was .343.
Offensively, that’s arguably a wash. And there’s no question, at least when Snow’s on the field, the position will be stronger defensively.
This is how the analysis goes throughout the whole piece. By virtue of parenthetically telling you the player’s batting average and on-base percentage, along with how many home runs, RBI’s and games played a player logged in the prior season alone, Donaldson feels that he is able to perform the sort of statistical analysis that can allow one to draw conclusions. Now, I am no mathematician but there are so many problems with this. First, it’s evident that Donaldson has no clue that players’ performances tend to vary greatly from year to year. For instance, J.T. Snow has a career OPS+ of 106, but has only come within 3 points of that figure once in any given season. So obviously, simply reeling off even relevant numbers for the previous season, much less RBI totals, would not prove to have a helluva lot of predictive value. Second, he clearly hasn’t the foggiest idea about park factors, and this will grow increasingly evident in other examples. And finally, it still ceases to amaze me that writers refuse to use more straightforward and accurate means of noting performance. I wouldn’t dare think Donaldson’s ilk would adopt any sort of Runs Created formula or OPS+ which requires not only a more advanced (and accurate) means of expressing performance quality but they also show, gasp, context. Even on a broadsheet like the ProJo, it’s high time we see at least the occasional BA/OBP/SLG. Anyway, for a reliable comparison, here is how Youks and Snow stack up compared to Millar.
Millar (2005): .272/.355/.399
Youkilis: .268/.400/.417 (ZIPS), .256/.374/.429 (PECOTA)
Snow: .267/.346/.378 (ZIPS), .267/.340/.418 (PECOTA
Donaldson goes on…
New second baseman Mark Loretta hit .280 in 105 games last season for San Diego, with 3 HR, 38 RBI, a .360 OBP, and 8 steals in 12 attempts in 105 games. Tony Graffanino, who took over from Mark Bellhorn last year in Boston, batted .309, with 7 HR, 38 RBI, and an OBP of .366. He had 7 steals in 9 tries. Give the offensive edge to Graffanino, but only slightly.
First of all, Graffanino hit .319 but had a lower on-base. It was .355, not .366. Second, I don’t think that Donaldson has any clue that Petco Park can suppress one’s offensive numbers or that Fenway can enhance them. To simply reel off numbers without mentioning where either player played his home games manifests such a fundamental lack of understanding of the game of baseball that it becomes almost unthinkable to me that he gets to professionally cover the sport.
Graffanino (2005): .319/.355/.457
Loretta: .320/.387/.441 (ZIPS), .292/.360/.423 (PECOTA)
Renteria batted .276 in 153 games, with 36 doubles, 8 homers, 70 RBI, and an OBP of .335. He had 9 steals in 13 attempts and, although he struck out 100 times, he walked 55 — a better ratio than Gonzalez’s 31 walks and 81 strikeouts. Gonzalez batted .264 last year, with 30 doubles, 5 homers, 45 RBI, and on OBP of .319 in 130 games. He attempted 8 steals and was caught 3 times.
Clearly, the Red Sox are hoping the runs Gonzalez can save with his glove will make up for his lack of production at the plate.
Another park factors story. Nothing new. Gonzalez was worse than Renteria last year but only slightly so. He looks a lot worse of course because he played home games in spacious Dolphins Stadium, something that is obviously lost on poor Jim.
Renteria (2005): .276/.335/.385
Gonzalez: .275/.326/.451 (ZIPS), .261/.311/.442 (PECOTA)
At third base, there’s no question Boston was better off with Bill Mueller, who hit .295 in 150 games, with 10 HR, 62 RBI, and a .369 OBP, than Mike Lowell, whose .236 average last year was a career-low. Lowell hit just 8 HR in 150 games for Florida, drove in 58 runs, and had an OBP of .298.
And everybody knows that if you put up successive all-star seasons and then your performance falls off the table, there’s no chance that the down year was simply an outlier.
Mueller (2005): .295/.369/.430
Lowell: .283/.356/.489 (ZIPS), .263/.330/.454 (PECOTA)
That brings us to center field, where Coco Crisp will be replacing the charismatic Johnny Damon, who’ll be batting leadoff for the Yankees this season.
Crisp hit .300 last year to Damon’s .316. Crisp had 42 doubles; Damon, 35. Crisp had 16 homers; Damon, 10. Crisp scored 86 runs and drove in 69. Damon scored 117 and had 75 RBI. Crisp played 145 games; Damon, 148. While Crisp stole 15 bases in 21 tries, Damon was 18-for-19.
In 2006, I’d rather have Damon, even though he can’t throw at all…
I am not sure how he arrived at that conclusion but whatever. Here’s how Coco compares based on actual, sophisticated mathematical projections.
Damon (2005): .316/.366/.439
Crisp: .321/.370/.466 (ZIPS) .299/.357/.467 (PECOTA)
So according to the most sophisticated projection systems that are publicly available, the Red Sox have not only not regressed at the positions Donaldson references, but they have actually improved at each. But hey if Donaldson wants to surmise that the Sox offense might take a hit, there’s an argument to be made. Projections are just that – projections – and there is room for human observation going into a given season. But Donaldson has this to say at the outset of the piece…this isn’t an observational column but instead an attempt at statistical analysis.
A statistical comparison — you know that the stat geeks reign supreme at Fenway these days — of the ’06 Sox to last year’s lineup, on a position-by-position basis, seems to indicate that this year’s team packs less offensive power than the ’05 bunch.
See, he called it a “statistical comparison,” not me, and even took a potshot in the process. And boy did he ever embarrass himself.
Donaldson isn’t smart…
One funny thing is that I’ve seen Crisp play maybe a dozen times or so in the last 12 months. He doesn’t have an arm either, Jimmy boy.
That’s fantastic. Nice work.
Sully, you didn’t put effort in quotes.
My biggest concern is that by titling the post as I did, I have done trainwrecks a great disservice.
Good stuff..the guy is not the brightest bulb in a dim room of variable visibility. What I just said I’m not even sure, but have a great weekend anyways.
Nice take on all the stats and the reporter. I found your blog the other day and I’m glad I did.