Stats & Info: Matthew Carruth

FanGraphs: Angels are getting lucky

June, 14, 2010
Despite the generally accepted wisdom, a team’s win-loss record is not always the best measurement for how well it has been performing during a season, especially early on. Statisticians prefer to do whatever they can to increase the sample sizes of their measurements, and while each game yields just one win and loss, it involves roughly 75 plate appearances and hundreds of pitches. Therefore, a team’s record is more prone to fluctuation than its overall hitting and pitching stats are. Evaluating teams based on the more numerous plate appearances provides a more sound measure of a team’s performance to date.

One such method of evaluation along those lines is BaseRuns, which is a formula used to predict how many runs scored and allowed a team should incur based on the number of hits, walks, home runs, stolen bases and total bases. Those predicted run totals can then be put into another well-tested equation, called Pythagorean Record, to produce how may wins and losses a team should have based on those more stable predictors.

We can compare that predicted record to a team’s actual record to find out which teams have been especially lucky or unlucky. Three teams stick out from these results as being especially lucky, Pittsburgh being one. It probably is surprising to hear Pittsburgh regarded as lucky, given its 23-40 record, but consider that the Pirates' run differential is minus-140 runs, by far the worst in baseball. The Pirates should hold MLB’s worst overall record, but instead, they sit six games ahead of the Orioles. The Astros have similar benefits, having MLB’s third worst run differential but a record about six games better than expected. Trumping all teams, however, the Los Angeles Angels sit as baseball’s luckiest team by this measure.

It is not atypical to find the Angels considered a “lucky” team by analysts. Quite often, their difference in actual wins over predicted wins is chalked up to savvy baserunning, a reliable bullpen and steady guidance from manager Mike Scioscia. Skeptics of these write-offs have extra reason to scoff this season, as the Angels have been successful on just 40 of 61 stolen base attempts (66 percent) and their bullpen has a 4.79 ERA, which is third worst in the AL.

Projected over a full 162-game season, the Angels are on pace to win a whopping 16 more games than BaseRuns indicates they deserve. As it stands now, they are 36-30 and own a .545 winning percentage, which would be good for about 88 wins. Yet they've scored exactly as many runs as they’ve allowed, and based on their overall profile, BaseRuns says the Angels would be lucky to even be .500 and that their record should be 29-37, which would give them 72 wins over a full season. Angels fans might be flying high right now with their team’s recent success, but they would do well to exercise cautious optimism for the rest of 2010.

Matthew Carruth is a writer for FanGraphs.

FanGraphs: Who's really carrying his team?

May, 12, 2010
Today on, Jerry Crasnick writes about players who are carrying their teams in a variety of ways. And while we can't put a number on any emotional or intangible lift a player gives his squad, we do have some cold, hard numbers that tell us which guys are really carrying their squads. To start, let's look at which players have the highest percentage of their team's wins above replacement.

Based on the numbers, no one is carrying his team quite like Choo, but that's a byproduct of being an excellent player on a bad team. The same can be said for McCutchen and Bourn. But I think the spirit of carrying a team implies something more. To make it relevant, you have to be able to carry a team that wins.

Therefore, let's give extra credit to Chase Utley and Roy Halladay of the Phillies. The pair has the most WAR of any batter (2.5) and pitcher (2.2) respectively this season and account for 23 percent of a first-place team's WAR.

Moving to the American League, Nelson Cruz and Justin Morneau are both are off to scorching starts for division-leading clubs. Morneau’s 2.2 wins above replacement is already more than half as many wins as he has in any season of his career, including his 2006 MVP season and accounts for 10 percent of the Twins' total WAR. Cruz’s 1.9 wins above replacement is possibly more impressive given that he did that in only 19 games before hitting the disabled list with a hamstring issue. He’s been a force in a lineup that has so far disappointed in offensive production. His production has accounted for 13 percent of the Rangers' total WAR.

On the other end of the spectrum are some players that teams were counting on to perform and have faltered so far. Chief among those would be Aramis Ramirez (-0.9 WAR) of the Cubs, who has been well below replacement level.

Howie Kendrick (-0.5 WAR) and Erick Aybar (-0.2 WAR) of the Angels have seen bigger than expected regressions from their highs in 2009. Both are under replacement level and are a big reason why the Angels have been so disappointing. Over in Boston, a team that Theo Epstein built on pitching and defense has seen good hitting and the expected solid defensive play, but has been completely let down by its vaunted pitching staff. Josh Beckett’s struggles are highly visible (0.5 WAR), but the entire staff has been less than dominating, and excluding Daniel Bard and Jonathan Papelbon, the bullpen has combined for -0.6 WAR. Defenders can only do so much when the ball is being lined all over the park.

Matthew Carruth is a writer for FanGraphs.

FanGraphs: Seattle's left field problem

April, 14, 2010
What do Jeffrey Leonard, Greg Briley, Kevin Mitchell, Mike Felder, Eric Anthony, Vince Coleman, Rich Amaral, Jose Cruz Jr., Glenallen Hill, Brian Hunter, Rickey Henderson, Al Martin, Mark McLemore, Randy Winn and Raul Ibanez have in common? From 1990 through 2004, each player in turn was the regular starter in left field for the Seattle Mariners. Not a single player repeated during that entire span. When Raul Ibanez occupied the primary starting role from 2004 through 2008, it marked the first time that the same person played left field regularly in consecutive years since Phil Bradley did in 1986 and 1987.

With Ibanez in Philadelphia, however, the Mariners left field carousel is back. Last year, the position was manned by Endy Chavez, Ryan Langerhans, Wladimir Balentien, Michael Saunders and Bill Hall. Saunders and Langerhans are in Triple-A Tacoma, while the other three are with other organizations.

For now, the de facto starter is Milton Bradley, but there are problems with that arrangement even without discussing Bradley's off-field baggage. He has had surgery on both knees in the past, and in 2009 he he missed games with minor pains in his left quad, groin, right calf, right hip, both hamstrings, left knee and right quad. Safeco Field holds a vast expanse of real estate to cover in left field, and Milton Bradley attempting to cover that much ground is bad for the team’s defense and bad for Bradley’s future health.

It would be a tougher decision if the Mariners were playing Bradley in left field in order to make room for a Vladimir Guerrero-type bat at DH, but they’re not. Most projection systems have the Mariners’ current DH platoon of Ken Griffey Jr. and Mike Sweeney at replacement level, meaning that they are not any better than what you’d expect to get from a player claimed on waivers. (Last year, Griffey was worth 0.3 wins above replacement, and Sweeney was 0.4.) There is no one blocking Bradley from moving to designated hitter. Not to mention the fact that Bradley had the best year of his career in 2008, when he posted a 4.6 WAR as a full-time DH for the Rangers.

With Bradley at DH, left field is opened up for a platoon of Eric Byrnes and either Ryan Langerhans or Michael Saunders; either pairing would provide solid defense coupled with enough hitting to allow Bradley to move to DH with no overall loss in left field. The Mariners are costing themselves about two projected wins with their current arrangement and, needing to make up ground in the AL West, cannot settle for another sub-par left fielder, even if that is the norm for the organization.

Matthew Carruth is a writer for FanGraphs.

FanGraphs: The Sports Guy and FIP

April, 6, 2010
I join many others in welcoming Bill Simmons to the statistical revolution in baseball. While statistics have always been an integral part of baseball, we have learned to tinker with them to more accurately reflect what happens on the field, and more importantly, what will happen. One of the more difficult puzzles to crack has been evaluating pitchers due to how entangled their performance is with that of their defenders. Accepting the move from ERA to Fielding Independent Pitching is probably the single biggest step one can take on the right path of separating the two, and Simmons has made that leap.

The Sports Guy made the case for FIP in his piece by referencing White Sox closer Bobby Jenks, who posted a decent 3.71 ERA last year, but whose secondary stats added up to a more mediocre 4.47 FIP. Given that Jenks’ ERA was significantly lower than his FIP, he concluded that Jenks wasn’t as good as his traditional numbers made him appear. While this is usually true, and the process he used to make his conclusion works most of the time, there is one more important number to check.

Home Runs allowed, an important input to the FIP formula, are not as skill-based as had been thought throughout history. Research has shown that pitchers have little control over how often a fly ball actually leaves the yard. In fact, if you want to predict how many home runs a pitcher will give up in 2010 you are better off looking at his 2009 ground ball ratio rather than his 2009 home run totals, the latter of which can occasionally include some good or bad luck due to wind, park, or random variation.

That is why xFIP exists, to correct those home run rates. It is simply the FIP formula with an expected home run rate based on fly
ball totals, rather than actual home run rate. Substituting for the average amount of fly balls that turn into home runs leads to more accurate future projections than just looking at FIP by itself. Sticking with Simmons’ example of Bobby Jenks, he certainly did look like garbage at times last season, in large part because 17 percent of his fly balls went for home runs. That’s nearly double his career rate and well above the league average, which is around 11 percent. While Jenks’ ERA of 3.71 didn’t match up well with his 4.47 FIP, his xFIP of 3.63 shows that Jenks had some bad luck on fly balls clearing the wall.

Bobby Jenks rediscovered his strikeout touch in 2009 (8.3 K/9) and kept his walks low (2.7 BB/9). If his home runs hadn’t ballooned, he would have been seen as one of baseball’s best closers. While Bill was right to look at FIP to see whether Jenks’ ERA reflected how he really pitched, making that last small step to using xFIP to project his future performance will give him, and you, an even greater advantage.

Matthew Carruth is an author of FanGraphs.