Results tagged ‘ statistical analysis ’
One of the biggest shortcomings in evaluating players’ performance based on their performance statistics (i.e., the outcome of the batter—pitcher matchup) is the uneven quality of opponents that players face. Baseball analysts have made it a priority to adjust for the ballpark in which the players perform, recognizing the profound effect it can have on a player’s accomplishments, as measured by his performance stats. But far less energy has been devoted to adjusting for the quality of the competitors a batter or pitcher face over the course of a season. In the case of ballparks, perhaps the potential effects are more obvious as we watch a 320 foot pop up into the right field corner at Yankee Stadium land in the seats for a home run. We intuitively know (or the broadcasters will tell us) that in 29 other ballparks the right fielder would have camped under the ball to record a routine out. As a result, park factors have become a common element of baseball analysis and player evaluation. It’s far less obvious (and difficult to measure) that the MLB schedule dictates that some hitters routinely face difficult pitchers, while others face pitchers that are less accomplished. The quality of opponents is even more varied for starting pitchers. Take an unbalanced schedule and match up a starter with an opponent team every 5th day and you’re apt to get an even more skewed distribution of batter-pitcher matchups.
There are many ways to create a statistical adjustment in the form of an “opponent factor”—much like a park factor—some more complicated than others. The simplest albeit crude adjustment is to look at the OPS of the opponent a pitcher faces. For perspective, the MLB average OPS for this season is .726 (.718 in the NL and .734 in the AL, with the difference largely due to the designated hitter). The Rays’ David Price has faced opponent lineups this year that have hit left-handed starting pitchers to the tune of a .757 OPS. On the other extreme, the Cardinals’ Jaime Garcia (currently injured) has faced opponents with a .693 OPS. The reasons this gap is so large is the combination of the unbalanced schedule and the wide range of offensive performance across teams. For example, the Cardinals are batting a remarkable .861 against left-handed starting pitchers, while a team in their same division, the Cubs are hitting a meager .645—a difference of over 200 percentage points in OPS.
Using our simplistic measure of the quality of a starting pitcher’s opponents, it’s interesting to note the clustering of the “degree of difficulty” by Division. The AL East hits .751 against LH starters and .738 vs. RH starters. The comparable MLB-wide OPS stats are .729 and .724, respectively. So, it is not surprising to see 4 Oriole starting pitchers among the top 20 pitchers that have faced the toughest lineups this season. In fact, 9 of the top 20 are AL East pitchers. There are only three National Leaguers in the top 20—Mike Minor (2), Eric Bedard (14), and Paul Maholm (20). Minor faced the Yankees twice, with their gaudy .839 OPS against lefty starters.
On the other end of the spectrum, the bottom 20 list—the pitcher’s that faced the weakest lineups— is loaded with NL’ers, including every Rockies starter. Several notables are Madison Bumgarner and Gio Gonzalez. Out of 140 starting pitchers, they rank 129 and 134, respectively, in terms of the degree of difficulty their opponent lineups presented. The only American League starters in the bottom 20 are Scott Diamond (123), Ricky Romero (125).
I incorporate these adjustments into my Starting Pitcher Ranking system (SPR), which I will write more about in the coming weeks. Given the acceptance of ballpark effects and the overall level of sophistication of statistical analysis in today’s game, it’s time to formalize our adjustment of all players’ performance stats for the quality of the opposition.