Monday, March 18, 2013

Have the Twins been lucky in the Gardenhire Era?

"Luck is the great stabilizer in baseball"--- Tris Speaker, Hall of Fame Centerfielder

Baseball is a game of numbers. The statistical side of the game continues to grow as the Sabermetric world adds new information. There are stats for what seems like everything that a person can think of in relation to baseball.

One area of baseball is hard to quantify and it can be the area that makes the game so exciting to follow. There is an element of luck involved and it can be the stabilizing effect on the game. Luck can be involved in everything from the location of a pitch, the way a ball squeaks through the infield, or even how a team aligns their defense. 

All of these small elements of luck can add up to wins and losses at the end of the year. Every run scored and every run allowed has an impact on the outcome of games. The luck factor adds something to baseball that allows fans for every team to have something to cheer about. 

In the book Baseball Between the Numbers, the Baseball Prospectus Team has an entire chapter devoted to managers and the moves they make during the course of a game and a season. Luck can play a factor in some of those moves. The manager has some control over the outcome of the game but how much luck is involved?

One of the ways to look at how lucky a team was during a season is to compare their actual record to their Pythagorean record. For those not familiar with Pythagorean winning percentage, here is the definition from Baseball Reference
"Pythagorean winning percentage is an estimate of a team's winning percentage given their runs scored and runs allowed. Developed by Bill James, it can tell you when teams were a bit lucky or unlucky. It is calculated by: 

(Runs Scored)^1.83
---------------------------------------------------------
 (Runs Scored)^1.83 +  (Runs Allowed)^1.83
The traditional formula uses an exponent of two, but this has proven to be a little more accurate."

To give an extreme example, look at last year's Baltimore Orioles. They used an excellent bullpen and a little luck in one run games to win 93 games and take home a wild card spot. Since the team only scored seven more runs than it allowed, their Pythagorean winning percentage had them winning 82 games.

Ron Gardenhire has been at the helm for the Twins since the 2002 season. There have been some high and low points during his 11 years as manager. But the questions remains, how lucky have the Twins been during his tenure?

Below is a comparison between the Twins actual win-loss record and their estimated Pythagorean win-loss record since Gardenhire took over as manager.
As you can see, in recent years, the Twins haven't strayed very far from their expected total. During the last eight seasons, the team has been within a couple games of their estimated mark. This shows that the formula is usually close to predicting the actual record of a club.

The most recent outlier came in 2006. In that season, the Twins had to hunt down the Tigers in the last weekend of the season to win the Central. Minnesota needed those extra three wins or the team would have finished out of the playoffs. It would seem that luck favored the Twins in 2006.

In the early years of Gardenhire's managerial career, there seemed to be even more luck on the side of his teams. The 2004 squad won 92 games and finished five games higher than their expected mark. Minnesota won the division by nine games so those five extra wins made little difference in their title.

The 2003 squad has a little bit different story on their side. Much like the 2004 team, they finished five games higher than their Pythagorean W-L mark. The AL Central was much closer and the Twins only won the division by four contests. Chicago finished second and they were two wins under their estimated mark. A couple of bounces the other way and the White Sox might have been in the playoffs.

Gardenhire's first year as manager saw the team exceed it's Pythagorean winning percentage mark by the most amount of games. Those eight extra wins were great and it helped the team to 94 wins, the most the team had won since the 1991 season. The team won the division by 13.5 games so all of those extra wins might not have been the difference in the race.

It would seem that luck has been on the side of the Twins in more than one season of the course of the last decade. The 2003, 2006, and 2009 teams all needed their extra wins to raise another division championship banner. Take those titles away and Gardenhire's resume doesn't look nearly as good. 

Luck has been on the side of the Twins in more than one season and fans can only hope that the team's luck will return in the coming years. 

1 comment:

TT said...

There is no real evidence the Pythagorean winning percentage is anything more than a novelty.

Its essentially a confusion of the impact of mean (average) and medians. The two are certainly related, but they aren't the same.

Teams win games one at a time. That means it is the median for those games, not the mean, that is important. Like any other situation where the median and the mean sometimes diverge, the mean does not tell you what the winning percentage "should" be absent "luck".