06 May 2004

Say It Ain't So, Joe...

Would it not amaze you, if you were a business analyst, to find out that Jack Welch or Bill Gates was staunchly and stubbornly convinced that the secret to success in the business world was having a lot of money in your bank account?

Wouldn't you be astonished if, as an aspiring actor, you read an interview with Tom Cruise or Kevin Costner in which he stated that the primary concern for an upcoming actor is to make sure he gets roles in movies that will pay him $20 million?

Wouldn't you just laugh out loud if one of the Joint Chiefs of Staff held a press conference explaining that the key to winning the wars in Iraq and Afghanistan is making sure we kill more of theirs than they do of ours?

Wouldn't you just stand there agape if you went to a Billy Graham Crusade and heard him say that the best thing for you to do with your life would be to make absolutely sure that God chooses you as one of The Elect?

How ridiculous would this be, for the so-called "experts in their field" to preach to you that the best measure of your value is something that is either completely out of your control, or a result of your circumstances, once you've already ascended to the pinnacle of your profession?

In a related story: Joe Morgan.

Well, as ridiculous and laughable as these scenarios sound, Joe Morgan manages to somehow get away with doing exactly that, pretty much all the time. Joe has written books on baseball, has a weekly chat on ESPN.com (which Mike Carminati performs the service of picking apart for us), writes columns for the website and he even propounds, from his nationally broadcast bully pulpit on ESPN Sunday Night Baseball, these silly ideas about what makes a good baseball player. Morgan, as ESPN is so apt to point out, was the NL MVP in 1975 and '76, so you'd imagine that he would know what makes a good baseball player, since he was a great one. Except he doesn't.

Joe Morgan will tell you, for instance, that ERA is an overrated statistic, and that the real measure of a pitcher's abilities is Wins & Losses. Never mind that there's an entire league in MLB in which the pitchers almost never get an opportunity to actually produce some of the offense that would be necessary for them to be credited with a win in any given game. Never mind that even in the league in which pitchers do "hit", they do so at the bottom of the lineup, where there is little or no pressure to contribute very much offensively. A pitcher may do the absolute best he can, shutting out a team for nine innings, and striking out every batter in the process, but allow an unearned run to score because he has the Marx Brothers playing infield defense behind him, and be credited with a loss. Conversely, a pitcher may allow five or six runs per game, but still get credit for the win, because his teammates become indwelled by the 1927 Yankees whenever he takes the mound.

Take, for example, Jeroime Robertson, the erstwhile Astros rookie who "won" 15 games in 2003, with a 5.10 ERA. How good was he? He was so good that the Astros bumped him from the rotation for 14-game loser Tim Redding and his 3.68 ERA, and then sent him to AAA. Then, just to make it very clear how valuable those 15 wins were, they traded Robertson to the Cleveland Indians for a AA-level non-prospect and the rights to a minor league outfielder they had already taken in the Rule 5 draft. Sounds to me like the Houston front office believed that the 5.10 ERA said a lot more about Robertson's abilities (and future) than those 15 wins did.

But I digress. This column is actually going to look at another of the Sabermetric Sins that Morgan commits regularly: Overvaluing Runs and RBI.

Morgan's latest column on ESPN.com deals primarily with two issues, the "realness" of the 2004 Texas Rangers and the relative merits of Runs and RBI as compared to such arcane notions ar on-base percentage. Let's take the second part first, shall we?

On-base percentage has always been an important stat, but RBI and runs scored are the truest tests of what a player does to help his team win. Once runners get on base, someone needs to drive them in. OBP by itself does not equal success. How often does a team get four walks in an inning to drive in a run? OBP is essential, but a good OBP alone does not guarantee a win or a successful season.

To Morgan's credit, and he does deserve some, he admits that OBP is essential. This is, in fact, the best thing I think I've ever heard or read from him about the subject. And he's not wrong about someone needing to drive in the runners on base in order for them to score, but he fails to see that the crucial point in building an offense is finding the people who will get on base in the first place. He still manages to cling to the old adage that scoring and driving in runs are the best measures of a hitter's quality, without recognizing that the "skill" of clutch hitting doesn't really exist.

Don't believe me? Take a look at the teams that have led the majors in batting average with runners in scoring position over the last five seasons, including this one:

BA RISP 2000 2001 2002 2003 2004
1 Rockies Indians ChiSox Royals Astros
2 Royals Seattle Angels Atlanta Detroit
3 ChiSox Rockies Yankees Seattle Angels
4 Giants Giants Rockies Toronto Atlanta
5 Indians Royals Astros Boston Twins
6 Toronto ChiSox Boston ChiSox Royals
7 Texas Astros Seattle Oakland Texas
8 Oakland Texas Cards Expos L.A.
9 Seattle Oakland Arizona Florida Reds
10 Yankees L.A. Texas Angels Florida

Five years, fifty positions, 23 different teams.

If you throw out Colorado, who until recently almost always led the majors in everything offense-wise due to their ballpark, you've only got three of the fifteen teams that manage to repeat their performances among the top 5: Atlanta(2003-04), Kansas City and the Tribe (both 2000-01). For that matter, Atlanta's standing on this list is in serious jeopardy, since they're clearly not as good an offense as they were last year and they've hardly got any of the same hitters they had last year (Sheffield, Lopez, and Castilla left in the off-season, and Chipper's been hurt this year), so there's really not much to compare anyway. In the top ten, it's a little better, with 18 of 38 possible non-Colorado teams repeating, or 47.4%. You can do better flipping a coin.

Well, you may say, "Those are teams, and teams turn over players a lot. Surely there must be players who can consistently hit in the clutch?" Well, by this same (admittedly limited) measure, there are, or I should say, there is.

Mike Sweeney.

	2000	2001	2002	2003

1 Helton Suzuki Ramirez Helton
2 Cirillo Alomar Sweeney Sweeney
3 Sweeney Conine Bonds Posednik
4 Delgado Biggio Tejada Sheffield
5 Thomas Beltran Bernie I. Rodriguez

The only players who re-appeared on the 2000-2003 top five lists of players's batting averages with runners in scoring position were Todd Helton (who plays in Coors Field, and therefore will almost always hit well in almost any situation) and Mike Sweeney. Don't get me wrong, Sweeney's a fine player, and he's hitting .367 with RISP this season, but he's not typically one of the first players mentioned when people start talking about drafting a fantasy team, you know?

The perfect example of this is the defending World Series champion Florida Marlins. Florida finished 15th last year in OBP (.333) among 30 MLB teams. But when the Marlins got runners on base, they were good at forcing the issue -- using the hit-and-run, stealing bases and exhibiting aggressive baserunning ... all of which help produce runs.

Unbeknownst to Joe Morgan (but knownst to us), the 2003 Florida Marlins are the perfect example...to illustrate my point, though, not his. Do you see what Morgan does here? It's subtle, but read what he says again and see if you can detect the flaw in his argument...

...back already? OK, here's what he does: He cites a team that is known to have had success in winning (the Marlins won the NL Wild Card and then beat three teams to win the 2003 World Series, of course) and mentions that they did so DESPITE the fact that they did not get on base any more frequently than the San Diego Padres (a .333 clip, to be precise). So far, this is true. He then mentions the things they did to supposedly augment this relative weakness and allows you (the reader) to assume that these tactics MUST have been successful, because after all, the Marlins won, right?

Well, yes, they did win, but not the way Joe's argument implies. It is true that they finished 15th in team OBP, but they also finished 17th in Runs scored. You'd imagine, wouldn't you, that if their small-ball techniques were actually helping them to score runs, to take advantage of the few times they had runners on base, that they might be ranked higher in runs scored than their rank in OBP, wouldn't you? Or at least the same? But the Marlins are actually ranked lower in run scoring than in OBP, which would tend to speak against stealing bases and moving runners over as useful approaches to scoring runs. In fact the 2003 Marlins won on the strength of their pitching (10th in MLB in ERA), and luck, as they won about 4 more regular season games than their Pythagorean projection would have suggested they'd win, based on the runs they scored and allowed. Not small-ball.

As I see it, a good hitter either scores runs or drives in runs -- and a great hitter does both. Great hitters will account for about 200 runs per season (a combination of runs scored and RBI). When you look at baseball history, that's the benchmark for great hitters like Hank Aaron, Babe Ruth and Willie Mays (as well as today's great hitters, like Barry Bonds). The most important stat in baseball is the combination of runs scored and RBI.
(bold added)

Is that true? Do managers and general managers and scouts really use the addition of runs and RBI as a measure of how successful a player is or will be? Seems to me I'd have heard something about that if it were actually happening.

Maybe it's just me, but wouldn't you say that these guys drove in/scored 100 runs because they were great hitters, rather than (as Joe implies) that they were great hitters because they drove in/scored 100 runs? I'm not a Minor League scout, but somehow I seriously doubt that there's anything on a standard scouting report discussing how many runs a player scores or drives in. Scouts talk about a hitter's swing, his mechanics, his physique, patience, power, batting eye, etc. They don't mention runs scored or batted in because those are situational stats, a result of opportunity more than talent, and not something you can project. The greatest hitters in history "accounted" for 200 runs because their managers knew they were great and therefore batted them amongst the other good hitters on the team, thereby maximizing their opportunities to score and/or drive in runs. Barry Bonds (who, by the way, must have had an "off" 2003 by Morgan's standards, with only 90 RBI) would not be a 100-Run, 100-RBI guy if you batted him 8th or 9th in that Giants' lineup.

Bill James did a study (in the New Bill James Historical Baseball Abstract, p.785) in which he illustrates the silliness of walking a batter (Babe Ruth circa 1921, in this case) every time up. He placed a slightly modified Ruth in the cleanup spot within a lineup of really terrible hitters, and then ran separate simulations to show what would happen if

A) Ruth was walked every time he came up and
2) Ruth was allowed to hit within this terrible lineup.

Well, the A) lineup outscored the 2) lineup significantly, with about 10% more runs per season on average and winning about ten more games per season.

And the #5 hitter behind Ruth in this illustration? Gino Cimoli. An outfielder with a career line of .265/.315/.383, who never hit more than ten homers or drove in more than 72 RBI in a season in reality, suddenly becomes great by Joe Morgan's standards, with a .253 average and 9 homers, but 151 RBI. Can you say "victim of circumstance"? That's great. I knew you could.

I view baseball as an individual game within a team concept. It's individual because, whether you're hitting or pitching, you're the one standing there. But everything is done in the context of teamwork and team play (putting the team first). RBI and runs are the ultimate measure of a player's contribution to a team, and they're also dependent on teammates. Home runs, though, are the ultimate measure of a power hitter's individual accomplishment.
(bold added)

Again, Morgan does acknowlege that RBI and Runs depend upon one's teammates, but he doesn't seem to give enough weight to the concept. Look at the way he just throws this line in amongst a paragraph describing the role of home runs in the game and how the team/individual aspects of the game interact. Does that make it seem like Joe gets the notion that OBP is the most important component to scoring runs? I don't think so.

Look, Joe, if you had a published e-mail address, I wouldn't have to keep fisking you like this, but you don't, so let me lay this out for you:

1) To win a game, you need to score more runs than the other team.
2) To score runs, you need to have men on base, and then drive them in somehow.

I think we both agree on these points. Here's where it gets a little tricky:

3a) It is difficult, if not completely impossible, to predict which players will hit best when it most matters, and it is all but impossible to make sure your best hitters are at the plate when you've got runners on base

3b) It is much easier to predict/project which hitters will be able to get on-base consistently from year-to-year, as you can see here:

OBP	2000	2001	2002	2003	2004

1 Indians Seattle Yankees Boston Houston
2 Rockies Rockies Seattle Yankees Orioles
3 Giants Indians Arizona Cards Indians
4 Seattle Houston Boston Toronto Rockies
5 Houston Oakland Giants Atlanta Texas
6 Oakland Texas Angels Rockies Twins
7 Cards Giants Oakland Seattle Yankees
8 ChiSox Arizona Philly Philly ChiSox
9 Yankees Cards Texas Twins Boston
10 Angels Twins Cards Giants Reds

The same forty opportunities to repeat here as we had with the last table, but this time only 17 different teams appear, not 23. And even if you don't include the Rockies, the ratio of repeaters is over 68%. (At the end of the 2004 season, don't be surprised if that ratio is higher, as I have serious doubts about whether the two Ohio teams will still be among these ranks come September. Well, maybe the Reds will.)

there4) It logically follows that it is easier to build a good offense by acquiring guys who will get on base a lot, so that when they do hit, there will be runners to drive in. Not necessarily that it's the only way to do this, but it's easier.

So there you have it, Joe. I imagine that you've hear dthe argument before, but it seems that either you're unwilling or unable to consider it. If you had an email address, I could write to you directly, instead of having to bring this up in public all the time, but you don't, so everybody gets to read my rants instead.

Thanks for giving me something to gripe about.

Stumble Upon Toolbar