By the Numbers: Selective memory

Michael Salfino
Yahoo! Sports

You can find more from Michael Salfino at Comcast SportsNet Washington.

When Carl Pavano(notes) was recently traded from the Indians to the Twins, we read and heard in many places that he was much better than his actual ERA if you subtracted his worst starts.

That seems like tautology to me. You're a lot better without your worst starts and a lot worse without your best ones. In Pavano's case, as noted by Baseball HQ's Ron Shandler in USA Today, Pavano without his five worst starts (as of August 4th) didn't have a 5.37 ERA, but instead a 3.28 mark.

I touted Pavano back on May 5 in this column because of his good K/BB ratio at the time despite an ERA on that day of 7.46. Since then, his ERA is 4.53, worse than I hoped, but still generally supportive of the recommendation.

So is Pavano a 5.09 pitcher (his current ERA), a 4.53 one (his ERA since May 1) or a 3.28 stud (the one minus his five worst starts when he was traded)? I can buy assessing true ability without a few worst starts, but five is too many. And I believe that certain pitchers with more marginal stuff (like Pavano) have less margin for error and are thus more capable of being severely pounded.

But to test it better, I looked at the ERA trailers among qualifying starters and determined what their ERA would be if we removed not five but just their three worst starts. Most starters now have at least 20 starts, so this at least gives you a better sense of what you can expect 85 percent of the time. Just don't try to figure out when those 15 percent rockings are coming because no one knows.

Here are the overall worst ERAs among starters who qualify (one inning for each game their team has played): Todd Wellemeyer(notes) (5.67 ERA), Francisco Liriano(notes) (5.63), Jamie Moyer(notes) (5.61), Joe Saunders(notes) (5.33), Mike Hampton(notes) (5.30) and Jeremy Guthrie(notes) (5.28). Wellemeyer and Moyer have recently been removed from their team's rotation and Saunders is hurt. And Liriano has been threatened with removal.

Here are their ERAs without their three worst starts: Wellemeyer (4.94), Liriano (4.65), Moyer (4.42), Hampton (4.21) and Guthrie (4.57). They're better, but still bad. And I'm not comfortable removing more than 15 percent of starts.

But remember, our Pavano stats were without his five worst starts. What would his current ERA be without only his worst three? It's 3.68 and it's reasonable to expect that Pavano the vast majority of the time if you are a Twins rooter. Just hold on to your hat the other days (or maybe use that hat to cover your eyes).

Let's scrub the worst three starts from some other pitchers and see if we can use the results to make some recommendations. "Sell" in this piece usually means trade, but this time it means "reserve" or "cut" (in mixed fantasy formats).


Cole Hamels(notes), Phillies: His ERA without his worst three starts drops from 4.77 to 3.75. He's allowed nine homers since July 1 and I think the chance of avoiding another disaster is bad. But if the price is low enough and if you're desperate enough for his undeniable upside, you can pay a steeply discounted rate. Hamels hasn't been able to string together more than three good starts all year. So most of his owners will be happy to sell.

Rickey Nolasco, Marlins: We said earlier the cat is out of the bag and down the block. Now, it's clear across town. You'll never get him, as his owners sure know he's turned it around. But I note he'll be much better than his overall ERA (4.86) going forward. Expect the adjusted ERA (with the three biggest bombings removed): 3.53.


Mike Pelfrey(notes), Mets: He throws hard and looks effortless when he's on. The overall ERA is 4.88, but it drops to 3.56 without the three worst starts. He's not a buy, though, because he's not going to help you enough in Ks or WHIP to be worth the rockings risk.

Scott Baker(notes), Twins: I thought he'd be helped more. His ERA goes down from 4.85 to 4.02 without the three worst starts, as there were a few more that just missed that cut. The ratio will be good, but the ERA is inflated by his fly-ball, gopher-ball ways (average of one blast allowed per start).


Chad Gaudin(notes), Yankees: He's moving to a new park and better team. Will he miss Petco? He had a 6.30 ERA and 1.95 WHIP at home for the Padres. Take out the three worst starts and he drops from 5.13 to 4.23. I'm selling though because he struggles with his control (4.8 BB/9) and when you do that at Yankee Stadium this year, you run a great risk of yielding three-run homers.