Ten Things I Didn’t Know Last Weekby Dave Studeman
September 01, 2005
Hurricane Katrina is the third-most intense hurricane to hit the US.
New Orleans, Biloxi and other areas of the South have been absolutely crushed by Hurricane Katrina. New Orleans, certainly the most unique city in America, may never be the same. Please consider supporting one of the many charitable organizations that are providing relief to Katrina's victims.
Now onto baseball-related revelations from last week:
This year's Wild Card race is certainly wild, maybe the wildest ever.
I actually like the Wild Card race format. I think it prolongs the excitement of the baseball season for more fans, and that's a good thing. I think it's a lot better for the game than either interleague play or the designated hitter, for example.
Just look at this year's race. As of Wednesday, five teams were within two games of the NL Wild Card (four of them from one division, the NL East) and three teams were within one game of the AL Wild Card. Is this the most competitive Wild Card race ever?
To answer the question, take a look at this table of the Wild Card standings as of August 31 for each of the past 11 years, including all teams within five games of the Wild Card at the time:
American League 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 0 0 0 0 0 0 0 0 0 0 0 0 0 -2 -1 -2.5 -1.5 -2.5 -0.5 0 -2 -5 -1.5 -5 -4 -1 -1 -3.5 -2 -4.5 -2 -3.5 -2 -5 -5 National League 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 0 0 0 0 0 0 0 0 0 0 0 -0.5 -1.5 0 -3 -5 -1.5 -1 0 0 0 -1 -3.5 -2 -2 -1.5 -0.5 -0.5 -2 -5 -2.5 -2 -3 -0.5 -3.5 -4 -3 -3 -1.5 -4 -3 -3.5As I said, there are eight teams within fewer than two games of the Wild Card slot. That's amazing; only the 1995 race, when 10 teams were within two games of the lead, can compete. In 1995, the Yankees (two games out as of 8/31) and the Rockies (in the lead on 8/31) eventually won the Wild Card races; I had almost forgotten that Colorado ever played in the postseason.
On the other hand, the 1995 season was 18 games short, due to the delayed start of spring training. Races are more likely to be competitive in a shortened season, so maybe 2005 does deserve the title, after all. You decide.
The other interesting thing about this table is the "curve" of the data, from tight races in the beginning to more barren years from 1997 to 2002 and more competitive races in recent years. At the same time, Wild Card teams have done very well in the postseason lately, winning the World Series each of the last three years. If that means anything...
There's a rumor that the Rangers are stealing signs.
I'm pretty sure this is only sour grapes, but Mark Buerhle accused the Texas Rangers of cheating after he lost a 7-5 game to them Monday. No, he wasn't referring to steroids. Buerhle claimed that the Rangers were stealing signs and relaying them to batters via lights in the center field office building. Shades of the 1951 New York Giants.
The Rangers do have 11 more victories at home than on the road (which isn't as big a difference as that of Tampa Bay's record, since they have 19 more victories at home), so maybe Buerhle is onto something. I did a bit of research to see how they're hitting at home vs. on the road.
Ameriquest Field in Arlington has been known as a hitter's park for a long time, but things are a little bit different this year. From 2002 to 2004, the runs scored per game in Arlington were 24% higher than average. But this year, the park is having only a relatively low 9% impact on runs scored.
The curious thing is that the impact is down for only the visiting batters. In fact, Ranger pitchers have actually yielded fewer runs at home (351) than on the road (354) while Ranger batters have scored 27% more runs at home (394 vs. 309). This doesn't support Buerhle's claim. In fact, it supports the opposite claim: Ranger pitchers are having their signs stolen in every other ballpark!
Seriously, it's not uncommon to see such a big difference between batting and pitching park factors in one year. Statistically, just about anything can happen in a given year.
What's up in Cincinnati?
Speaking of park factors, well, it's like I said. Anything can happen in a given year. Case in point: runs per game in the Great American Ballpark in Cincinnati were below average in 2003 and 2004 (actually, 8% lower than average) but this year, runs are scoring almost 20% more often in the GAB.
Here's a graph of each park's park factor over the three previous years, compared to this year's. Above the line means that scoring is up in the ballpark, below the line means scoring is down. I've only included relevant years for each ballpark (so the Nationals have no history for comparison).
As you can see, the biggest difference is in Cincinnati, but there have also been some big changes for the Rangers (as mentioned), the Yankees and Braves, among others. I don't know what this all means, but I didn't even know about it last week.
Hopefully, Felix Hernandez is no Dwight Gooden.
I got to see Felix Hernandez pitch against the White Sox the other night, and he is everything that Aaron and David Cameron claim he is. Even though rookie Brian Anderson (or, as my daughter calls him, that "white pasty guy") hit two home runs off of him (only to be sent to the minors a couple of days later), King Felix's stuff looked awesome.
It's ironic that Hernandez made his major league debut at the same time the last great teenage pitching phenom, Dwight Gooden, was arrested for driving under the influence (and subsequently fleeing the site). As you can see in this picture (compared to a picture taken earlier this year), Gooden has obviously had a very, very difficult time lately.
Perhaps the greatest thrill of my time as a Mets fan was watching Gooden come up to the majors in the early 1980s. I had never seen a 19-year-old ballplayer like him, and I haven't since (Hernandez notwithstanding). He had an effortless, fluid delivery, a fastball with "giddyup" and a curveball that was virtually unhittable. (The Mets' announcers called it "Lord Charles" instead of just "Uncle Charley.") The tradition of hanging a "K" for every strikeout started with Gooden (creatively copied with K’s that look like Felix the Cat for Hernandez). He also had that great smile and a demeanor that could seemingly handle New York. How wrong I was.
I hope King Felix can avoid the same fate, and I wish Dwight Gooden all the best as he attempts to get his life together.
Maybe it's time for a different approach.
Nate Silver wrote a neat article for Baseball Prospectus last week, in which he calculated the value great base runners add over more pedestrian leadoff batters. For his analysis, he took the following steps:
- He calculated the number of times a batter reached base (TRB), which is the numerator of OBP (you know, hits plus walks plus hit-by-pitch).
- He then calculated the value of a stolen base and extra base taken on a base hit—as well as the negative value of being caught stealing—relative to the value of reaching base. For example, a stolen base is worth 36% of a TRB.
- He then multiplied each base-stealing event times its relative value and added it to the total TRB for each batter.
- Finally, he divided this new TRB total by each batter's plate appearances to calculate a new on-base percentage (OBP), which he called speed-adjusted OBP (SOB, a great acronym).
As all aficionados of A’s manager Billy Beane and former Orioles manager Earl Weaver know, OBP doesn't just measure the percent of times a runner reaches base. It also measures the percent of time a batter didn't make an out. And adding base running-related TRBs to the equation misses the latter.
Let me explain with an example. First, I need to reformulate OBP. I'm going to change it from TRB/PA (or, times reached base divided by plate appearances) to a ratio of TRB/Outs (or, TRB/(PA- TRB)). A batter with a .350 OBP will have a ratio of .538 TRBs per out made, and a batter with a .370 OBP will have a ratio of .587 TRBs per out made—a difference of .05 points. So a difference of .02 OBP points equals a difference of .05 TRB/Outs. With me so far?
Now let's go back to Silver’s article. According to his analysis, Womack's great base running added 11 TRBs to his total, which increased his 2004 OBP from .349 to an SOB of .369. Holy Acronym, Batman! Womack suddenly looks like a decent leadoff man.
But when you look at the ratio of TRBs to outs, Womack's only rose from .537 to .576 (.04 points instead of the .05 I mentioned before). So Silver’s analysis overstates the impact of base running by about 25%.
Now, most of you are sitting there calling me a dweeb, geek or some other spot-on label. I'm not trying to pick on Silver or prove that I'm more anal-retentive than the average baseball fan. I think there's a bigger story going on here. Analytic sabermetrics is kind of at a crossroads right now, in which many of us are trying to answer questions that may not be best answered by our old bag of tricks.
For many years, there have been two fundamental approaches to run creation estimates: the Bill James/Runs Created camp, and the Pete Palmer/Linear Weights camp. The difference between these two approaches was articulated well by Tangotiger in his outstanding article How Runs are Really Created (be sure to read all three articles). Silver used the Linear Weights approach to calculate his base running TRBs and then applied them in a Jamesian fashion by manipulating OBP.
He did this for a very good reason: most sabermetric baseball fans understand the importance and language of OBP and on-base plus slugging (OPS). In fact, Silver was simply responding to a reader's request when he performed his calculations. If baseball analysts want to be heard and understood, we need to use consistent and accessible language.
But, for detailed and nuanced calculations like this one, OBP probably wasn't the best metric to use. A linear weights approach would have been more appropriate, albeit harder to communicate. That's what I mean by a crossroads: communicate well to a wide audience, or follow the best analytic approach?
We run into the same issue here at THT. That's why we like to use Gross Production Average (GPA) instead of OPS, because it's a bit of a crossover stat between the two camps. But that doesn't resolve anything. This dynamic will only intensify as sabermetrics becomes more accepted yet attempts to resolve more complex problems at the same time.
Baseball has much in common with the Supreme Court.
Well, I just about used up my allotment of words with that commentary. So let me throw in a few shorties. First of all, try playing this game about baseball and some Supreme Court justices. It's fun.
And while I'm busy linking, check out the Baseball Esoterica blog. I don't know if he can keep up the pace, but this blog has a lot of promise. And you might also enjoy reading The Reacquisition Theory, a very long article that documents how many player reacquisitions have failed.
Steve Finley has been benched.
The Angels announced that Steve Finley will not start for the rest of the season, and there is much rejoicing in Angel blogland. I hope, for their sake, it's true. Many people questioned the Finley deal in the offseason (particularly the wisdom of signing a 40-year-old to a two-year deal), but the real issue here is that it took the Angels too long to admit their mistake.
That J.T. Snow is one heck of a first baseman.
Someone recently commented to me that you can judge a first baseman by his infield's throwing errors and I thought "hey, we can do that!" So here's a list of throwing errors and fielding errors by each team's infield (not including first basemen).
Team TE FE % SF 7 25 22% ATL 12 24 33% STL 13 22 37% FLA 14 21 40% HOU 14 21 40% LAN 14 39 26% PHI 14 21 40% BAL 15 22 41% TEX 15 25 38% WAS 15 21 42% CHA 18 27 40% LAA 18 16 53% NYA 18 19 49% OAK 18 19 49% SD 18 22 45% NYN 19 24 44% CHN 19 28 40% CLE 19 23 45% PIT 21 24 47% MIN 21 28 43% TB 22 34 39% TOR 22 24 48% DET 22 26 46% ARI 23 20 53% CIN 23 20 53% SEA 24 20 55% MIL 24 30 44% COL 26 25 51% BOS 26 21 55% KC 28 30 48% Total 562 721 44%Forty-four percent of all infield errors are throwing errors, but there is a greater variance among teams in throwing errors than fielding errors. The Giants only have seven throwing errors, while the Royals have four times as many. Conversely, the team with the most infield fielding errors has about twice as many (39) as the team with the fewest (16). Funny thing is, both of those teams play in Southern California.
J.T. Snow, by the way, has played a little over half of San Francisco's games at first base. Lance Niekro has seen some significant action there too.
More Sudoku tips.
I mentioned my addiction to Sudoku over a month ago, and it seems that the rest of the world is catching on. The New York Times reviewed the history of Sudoku (registration required) and Bryan Donovan sent me a link to the best online version of Sudoku I've seen. Now, to top things off, you can play Hamster Sudoku.
Which makes me wonder if maybe Sudoku has jumped the shark.
References and Resources
Many thanks to Alex Belth for the Wild Card race idea and to Frank Vaccaro of SABR for his guidance with the wild-card standings. The great table was his inspiration. Retrosheet, the source of everything baseball, provided the stats. Also, thanks to The Cheat for mentioning the infield throwing errors idea. Too bad Konerko didn't do better.
Dave was called a "national treasure" by Rob Neyer. Seriously. Comments about this article can be sent to him through the miracle of e-mail.