By the Numbers: Strand and deliver
You can find more from Michael Salfino at Comcast SportsNet
Let's go back to the hurlers this week and apply some statistical forensics.
There's ample proof that simple models generally result in better predictions than expert judgment. Just looking at guys and telling you who I think are going to do better or worse risks being too strongly influenced by recent performance and the conventional wisdom regarding the current pool of players.
The best way to avoid biases and random guessing is to come up with a stat – this week it's strand rate – and see who the major outliers are relative to league-wide performance in that category. We expect regression from those doing unusually well and improvement from those doing unusually poorly. The key assumption, though, is continued opportunity. For example, our second "unluckiest" pitcher in strand rate, Andy Sonnanstine(notes), has already been shipped to the minors by the Rays.
I recognize that elite performers measured by conventional statistics like ERA will be outliers on the plus side and all of the really bad performers will be on the other end of the spectrum. Identifying outliers is just the first step. To best isolate good and bad luck we must also consider the player's career averages. Age and injury also have to be dismissed as being major factors. These work both ways, as players can get better due to entering their physical prime or may finally be healthy after a spate of injuries.
With the help of our friends at HardballTimes.com, let's look at qualifying starters who are stranding far more baserunners than the current MLB average of 72 percent. These are the guys who we suspect are lucky, in order of highest strand rate: Matt Cain(notes) (86.7 percent), Kevin Millwood(notes) (85.9), Danny Haren (84.5), Yovani Gallardo(notes) (83.7), Jered Weaver(notes) (82.8), Cliff Lee(notes) (80.9) and Ted Lilly(notes) (80.6).
Faithful readers will note that I just last week wrote that I'm sick of bashing Cain for reasons like strand rate, concluding that perhaps he's an outlier because he can dial it up when guys are on base. In other words, the more stuff you have, the luckier you can be. I refuse to bet that Cain will hold up in light of that strand rate, but I'm tired of betting against him, too.
The unlucky are those qualifying starters below a 65 percent strand rate. In order, they are Carl Pavano(notes) (60.4), Sonnanstine (62.3), Joel Pineiro(notes) (63), Jorge De La Rosa(notes) (63.2), Scott Baker(notes) (63.8), Derek Lowe(notes) (64.1), Brian Bannister(notes) (64.5) and Josh Geer(notes) (64.7).
You get wilder numbers with non-qualifiers. The biggest strand-rate outliers on the plus side here are Fernando Nieve(notes) (89.4 percent), Tommy Hanson(notes) (87.6), Ricky Romero(notes) (85.4), J.A. Happ(notes) (85.2) and Erik Bedard(notes) (83.6).
On the other side, i.e., the guys who should get better due to a poor strand rate, are Chien-Ming Wang(notes) (53.1 percent), Ricky Nolasco(notes) (57.1), Scott Kazmir(notes) (62.5), Ervin Santana(notes) (63.2) and Andrew Miller(notes) (64.2). Of course, Wang, Kazmir and Santana all have documented injury issues and it's not unreasonable to worry about Nolasco.
Now let's go to that second level of analysis and see if we can isolate the best and worst values from these groups.
BUY
Ricky Nolasco, Marlins: Yes, I know he's hot and his owners know it, but he's not even owned in about 10 percent of Yahoo! PLUS leagues. June stats: three wins, 1.91 ERA, 0.97 ratio, 33 Ks.
Andrew Miller, Marlins: He's not owned in about 80 percent of Yahoo! PLUS leagues and has 31 Ks vs. just 12 walks this month. Plus, he's allowed only three homers all year. The ERA will continue to sink.
Scott Baker, Twins: He's a "sabermetric darling" (a term my friend Steve Moyer of Baseball Info Solutions insists he coined). So you might not be able to get him at a discount. He's here because the over/under for his rest-of-season ERA is 3.70.
Jorge De La Rosa, Rockies: Available as a free agent almost everywhere, De La Rosa averages more than a K per inning and has a K/BB ratio of better than 2-to-1. Automatically claim guys like that. Yes, Colorado – but he pitched better there than on the road in 2008.
HOLD
Ervin Santana, Angels: Be very worried about the elbow, but watch him this weekend and note his velocity. If healthy, he's a huge upside play.
Cliff Lee, Indians: Don't get fooled twice. He proved last year, too, that he's more good than lucky.
SELL
Kevin Millwood, Rangers: Cash him in if you've been lucky enough to own him. Don't give him away, though, as finding a market for Millwood is not easy.
Tommy Hanson, Braves: I do respect the talent. But it's not just strand rate saying "sell." Hanson's walked 17 guys in 29 innings and isn't miserly with the homers (three jacks allowed).
Michael Salfino's work has appeared in USA Today's Sports Weekly, RotoWire, dozens of newspapers nationwide and most recently throughout Comcast SportsNet, including SNY.tv, for which he also analyzes the Mets and Yankees. He's been writing "Baseball by the Numbers" weekly since 2005.
