One of the first pieces I wrote for Baseball Prospectus outlined my strategy for identifying closers, and the article was so well received that I repeated it with a couple of tweaks last season. After I read the comments from one user who utilized the strategy and won his league’s saves category, it was a given that the piece would come back again in 2013. Some of the pitcher notes I made in the 2012 closer matrix included:
- Andrew Bailey: “Hasn’t thrown 50 innings since 2009, buyer beware.”
- Heath Bell: “K% dropped 10 points, Marlins should have given money to Stanton.”
- Jose Valverde: “K% has dropped six straight seasons. That won’t end.”
- Joe Nathan: “Solid second half masked by awful first half.”
- Jason Motte: “BB% down three straight seasons.”
- Sergio Romo: “BB% improved three straight seasons and K% is unworldly.”
The premise of the strategy is simple: closers are born to fail. Year after year, a noticeable percentage of ninth-inning men get hurt or lose their jobs. As we learned last month, relievers who went undrafted in the 15-team Mixed ToutWars auction accounted for 227 saves last year, though saves were rather scarce in single-league formats, as only six free-agent relievers reached double digits. That was an anomaly, and you should not let one year with an unexpectedly low failure rate deter you from exercising caution. After all, you should have been shaken to your fantasy core when you witnessed Mariano Rivera go down in a crumpled heap on the warning track in Kansas City while shagging fly balls. Rivera, even in his 40s, had been as durable as they come.
The experts vary on how to treat closers, and the format in which you play can dictate your opinion as well. If you play in a head-to-head league or use a no-trade format like NFBC, you cannot punt a category, and even if you play standard roto, saves are still one-eighth or one-tenth of your scoring. Several years ago, I won an AL-only league while compiling only 13 saves on the season, but that required nailing nearly every other category and besting my closest competitor by a single run to win the league outright. Relying on that sort of strategy annually is imprudent, and that’s why I came up with my own way to embrace my fear of closers.
This plan involves staring into a matrix—a matrix of what I consider to be the essential skills for evaluating closers.
- Ground-ball percentage: Ground balls are the toughest balls to hit out of the park. The 2012 league average for relievers was 45 percent
- Home runs per nine innings: A home run is the quickest path to a blown save. The 2012 league average for relievers was 0.9.
- Average: Batting average is a variable stat, but the less projected balls in play, the better. The 2012 league average for relievers was .239.
- Strikeout rate: Punchouts keep hitters off base and prevent bad defenses from turning routine plays into errors that extend innings. The 2012 league average for relievers was 22 percent.
- Walk rate: Free passes are recipes for disaster in high-leverage situations. The 2012 league average for relievers was nine percent.
- Splits: Pitchers with drastic splits tend to become specialists, while those who handle like- and opposite-handed batters equally well are usually entrusted with save opportunities.
The table below is a report from the Player Forecast Manager of all relievers that are projected to earn at least one save in 2013. In the past, color-coding has been used to show which pitchers are above and below league averages, but I failed to account for our color-blind subscribers. This year, in addition to color-coding and enlarged text, the matrix contains bolded print and italicized print; the former represents something that a given reliever does significantly better than the league average, while italicized print indicates a well-below-average skill. The last column contains a brief comment about each player. All but the splits stats are 2013 projected statistics from the PFM. If you would prefer a downloadable spreadsheet over the embedded worksheet below, you can view that file here.