Google Custom Search

You may have noticed the Google Custom Search thing to the right. You may have wondered if it was different than any other kind of search, Google or otherwise. Here's the scoop:

If you're anything like me you'll be reading an article and come across something that you want to know more about, either to fact check or for mere curiosity's sake. Let's say you're halfway through one of my posts and suddenly you want to look up information for, I don't know, Ron Karkovice. Pop 'er in the custom search and it'll open up a new window (so you won't lose your place in the original post) full of legit search results. "Ron Karkovice" yields results from Baseball-Reference.com and Wikipedia and all those good relevant sites.

But say you want information on pitcher Matt Smith. Holy cow, there must be a zillion different Matt Smiths populating the internet! Put him in the search anyway. Pow... Wikipedia, MLB.com, The Baseball Cube, etc. The search has been tweaked to yield baseball-pertinent results.

Type in "pants" and you'll get BASEBALL pants. Smooth, eh?

Play around with it, use it as a research tool, do whatever.

That's Mean: Fun With Regressions

I just noticed something, though this has probably been common knowledge for months. At press time (my living room doubles as a high-volume press room for daily periodicals, don't you know) the Texas Rangers have the best run support in the AL, averaging 5.63 runs per game. But they have the worst pitching/defense, allowing 5.91 runs per game and committing 100 errors so far (leading the league in both categories). They're currently four games above .500 and trailing the Los Angeles Angels by 11.5 games, good for second in the AL West. Bill James' Pythagorean expectation formula, which describes how "lucky" a team is, puts the Rangers six games under .500 and in theoretical third place. They've gotten some good breaks, but if we're to believe in "regression to the mean" then the Rangers will slide a bit before the season ends.

As I type this the Rangers are tied with the Yankees 5-5 in the 9th inning. I predict the Rangers will end up losing to the Yanks by a score of 5.63 to 5.91.

Interestingly, despite leading the Majors in errors, Texas also lead the Majors in turned double plays. The only explanation for this I can come up with is that the team's pitchers must be putting lots of men on base and lots of balls in play (many of which happen to end up as tailor-made double play grounders).

What does all this mean? I don't know, just weird statistical anomalies I noticed.

In fact I was looking up information on how "unlucky" Blue Jays' pitcher Roy Halladay has been this season. During tonight's game against the Athletics (see the post date) things were going his way as he allowed only one run while the Jays put up six. But in Halladay's previous start he limited the Rays to only three runs, while the Jays supported him with... zero. In fact, seven times this season Halladay has allowed four runs or less and came away with a loss or no-decision.

Here's a telling stat. The Blue Jays' bats produce 4.28 runs per game (discounting tonight's game, which would actually bump it up to 4.33). When Halladay starts the team only gives him 3.79 runs per nine innings (again, discounting tonights game which, if my math is correct, would bump his support up to 4.01 runs per nine innings).

With an ERA now at 2.77 he's going to get a lot of wins, but he's also going to get a lot of undeserved losses, and his 13-8 won-loss record reflects that. I don't know how to Pythagorize a pitcher's record, but Halladay has the third-best ERA in the AL, second-best WHIP, second-best K-to-BB ratio, second-most strikeouts, tied for most shut-outs, and has the most complete games (by far). Somebody appearing at the top of those rankings should have more than 13 wins to his credit by this point. He definitely shouldn't have eight losses. But I suppose that's what happens when you pitch for a team fighting with Baltimore for last place.

And, as far as Cy Young Awards go, this may come back to hurt Halladay. Wins and losses, as we've just proved, are meaningless when it comes to a pitcher's talent. Obviously a good pitcher is more apt to win games, but it ultimately depends on his team's offense. Many voting members of the BWAA still think that a won-loss record is a good reflection of a pitcher's abilities. If Halladay ends up with a respectable 18-12 record, he's going to lose out to Cliff Lee who'll probably go 20-4. (It should be noted, however, that in many of Halladay's "second-best" stats Lee is the one in front of him, and if Lee does win the Cy Young Award then he probably does deserve it).

Pythagorasly speaking, Toronto has been "unlucky," especially while Roy Halladay has been on the mound. Expect the Blue Jays to perform better the remainder of the season (more a law-of-averages type thing, less a Cito Gaston type thing), and expect Halladay, if he continues to do his part, to have fewer losses and no-decisions.

Texas won, by the way, beating New York 9-5. So much for regressions.