<$BlogRSDUrl$>

Talking Baseball

Your weekday baseball fix. Some days.



Posted by Jon on Tuesday, March 16, 2004

A Full Plate

Two of the last four posts have referred to Dusty Baker’s now infamous comments regarding walks and OBP. A synopsis? He doesn’t care to preach either walking or OBP to his hitters! Unbeknownst to him, the dynasty of the last decade used both to their advantage, and Baker’s San Francisco Treats were perennially among the leaders in both categories. (Incidentally, I wonder how the Giants would have fared without Mr. Barry "Season Averages of 131 Walks and a .433 OBP" Bonds. But that’s a story for another post.)

Mike and I received some feedback regarding Baker’s success in spite of his hitters’ futility in drawing walks and reaching base. Cubs fan John Honkala, offered this interesting tidbit:

While I am big proponent of walks, I don't think that they are the end all be all a lot of sabermetricians et al. make them out to be. Here's how the World Series teams from the last three years have ranked:

Florida 19
NY 1

Anaheim 26
San Francisco 5

AZ 5
NY 15

Certainly, walks mean something. A top five in all three Series. But, more notably, the past two years' champions have performed worse than more than (approx.) 2/3 of the rest the league when it comes to walks. I bet I can find another statistic that is similar to this, where top performers show up in the Series while poor performers do also.
John makes some great points here. Statistics are necessary in understanding baseball, but with so many available stats, they can easily become misleading. Walks, which were severely undervalued in baseball before the sabermetric revolution, have now become overvalued. The problem, I believe, is a confusion between walks and on-base percentage.

It should be noted that Mike and I discussed walks only in reaction to Baker’s comments. Dusty first lumped the two statistics together. Mike and I merely responded.

Which is the offensive statistic that correlates most with run production? Not walks, but OBP. Many fans read Moneyball, developing a sense that the best way to boost a team’s OBP is to develop players with a tendency to take walks. There was an untapped market of guys who consistently walked, which Billy Beane promptly took advantage of to increase the A’s OBP. As a result, we have come to value the walk almost as much as we value on-base percentage, which is a gross overvaluation. When we look at the World Series opponents’ OBPs, we see a different story – one of getting on base, regardless of the number of walks seen:

..............MLB BB Rank...........MLB OBP Rank
Florida.............19....................13 (tied)
New York (AL).......1.....................2

Anaheim.............26....................6
San Francisco.......5.....................5

Arizona.............5.....................8
New York (AL).......15....................13 (tied)

Average WS Team.....11.8..................7.8
In the last three years, the average team appearing in the World Series was about the 12th ranked team in terms of accumulating walks. In terms of OBP, though, the contending teams are closer to the leaders of the pack, averaging about eighth in OBP across Major League Baseball.

For Anaheim’s 2002 World Series Championship team, the real disparity between walks and OBP is apparent. Almost last in walks, they come in with the sixth-best OBP during that season. I believe what we’re really valuing here is on-base percentage, and not necessarily walks. Generally, walks lead to a higher OBP, which certainly produces more runs. But the focus of any team shouldn’t necessarily be their walks, but their OBP, however it is enhanced. Walks are just one way to increase a team’s OBP.

2003..........MLB Walk Rank...........MLB OBP Rank
Chicago (NL)........21....................25
It should be no surprise that last season’s Cubs team relied on pitching to reach the playoffs, with a dominating 1404 strikeouts over the season (the most ever by a pitching staff) and the sixth-fewest runs allowed per game in baseball. With such dominating starters, Dusty managed to win despite his hitters’ woeful production. Getting on base more than only five teams in baseball did not help Baker win ballgames. With a higher OBP, his team would certainly have scored more runs and won more games. Walking would have helped.

Now Making an Appearance…

Looking at annual team stats prompted me to examine how statistics can be skewed for hitters on exceptionally good- or poor-hitting teams. So I took the 2003 team hitting statistics, trying to account for enhanced or stifled numbers due to total team plate appearances. There’s no doubt that better-hitting teams end up with more plate appearances, but do these PA’s pad the stats? Differences in the number of times a team reaches the plate depend on many factors – team OBP, ability to extend innings, number of extra innings played, and amount of times a team is not leading at home in the bottom of the ninth. Regardless, there may be statistical differences.

On average, a 2003 team accumulated 6087.5 total at-bats and walks. The Red Sox surpassed this average to a notable extent, with a major league leading 6389 total plate appearances. The Mets, at the bottom of the list, received only 5830. Following is a list of every team’s 2003 total plate appearances.

Team, Plate Appearances
Boston 6389
Yankees 6289
St. Louis 6252
Atlanta 6215
Toronto 6207
Philadelphia 6194
Minnesota 6167
Texas 6152
Seattle 6147
Houston 6140
Colorado 6137
Pittsburgh 6110
Arizona 6101
Baltimore 6096
San Diego 6096
Arizona 6095
Milwaukee 6095
Tampa Bay 6074
Oakland 6053
San Francisco 6049
Kansas City 6044
Cleveland 6038
Cincinatti 6033
Chicago (NL) 6011
Chicago (AL) 6006
Florida 6005
Anaheim 5963
Montreal 5959
Detroit 5909
Los Angeles 5865
New York (NL) 5830

AL 6109.6
NL 6068.3
MLB 6087.6
We understand that the DH makes pitching in the AL tougher, but how does it skew hitting stats? In the AL, each spot in the lineup receives more plate appearances, and thus it becomes more valuable to send a good hitter to the plate. As one would expect, AL teams to recorded one extra plate appearance in every four games.

Plate appearance leader Boston accumulated more than 300 over the average Major League team. Compared to the average, Boston hitters enjoyed almost two extra plate appearances per game, which evens out to about 34 extra plate appearances for each spot in the lineup, one through nine. Expectedly, the Mets averaged almost 29 fewer plate appearances per nine spots in the lineup when compared to the major league average, or more than 257 lost plate appearances for their hitters. The difference increases when comparing these two teams directly.

The Red Sox enjoyed 559 more plate appearances than the Metropolitans over the entire season, which whittles down to a per game difference of 3.5 more plate appearances. This means that each spot in Boston’s lineup was given about 62 more appearances at the plate. They certainly made the most of their time in the batter’s box, but their extra plate appearances may skew their final evaluations as hitters.

The extra plate appearances hitters on elite scoring teams pile up, padding stats for some players. It is no coincidence that Boston led the majors in total hits, doubles, walks, runs scored, and at-bats, and that they came in second in home runs, and fifth in triples. They created the opportunities for additional production that could not have occurred on the Mets because of a lack of plate appearances. What I’m trying to say is that the Sox’ did not lead the league in plate appearances per hit (Toronto’s was higher). Their ability to extend innings allowed each Boston hitter more opportunities to pad their own stats (when compared to the rest of the league), and the team’s overall offensive stats. So the next time you’re comparing Boston’s offense to that of the New York Mets, remember that the same Boston hitters, with the number of Mets’ plate appearances, would have had a number of total hits reminiscent of the Kansas City Royals, not the record-breaking powerhouse offense we know them as. The Sox created these extra opportunities through their steller hitting, but the extra plate appearances make it more difficult to compare their players and the whole team to the rest of the league, and especially to bottom-of-the-barrel offenses.

To further illustrate my point, I calculated the plate appearances per home run for the nineteen players who hit the most home runs in 2003. They are ranked in order of plate appearances per home run.

...............HR...............PA/HR
J. Lopez.......43...............11.4
B. Bonds.......45...............12.0
J. Edmonds.....39...............13.4
S. Sosa........40...............14.5
J. Thome.......47...............14.7
A. Rodriguez...47...............14.8
F. Thomas......42...............15.4
R. Sexson......45...............15.6
A. Pujols......43...............15.6
C. Delgato.....42...............16.2
Ja. Giambi.....41...............16.2
G. Sheffield...39...............17.0
R. Palmiero....38...............17.0
J. Bagwell.....39...............17.8
M. Ramirez.....37...............18.0
A. Jones.......36...............18.0
P. Wilson......36...............18.2
A. Soriano.....38...............19.0
B. Boone.......35...............19.7
A few interesting things to point out. Of those who swatted at least 35 dingers, the top three home run producers were actually Javy Lopez, with an amazing 11.4 PA/HR. Bonds we expect to be there. But to me, Edmonds was a surprise.

Some statistical web sites offer at-bat per home run ratios. But it is important to do the same for total plate appearances. Depending on a player’s walk rate, the number of at-bats he receives can vary. Soriano doesn’t walk, so his AB/HR ratio differs from his PA/HR ratio by about two at-bats/plate appearances. Feel free to use either stat, but I prefer total plate appearances to an arbitrary stat that disregards walks.


### So what do you think? We want to know. | | E-mail us ###