MLB.com Columnist

Mike Petriello

Colon heads to free agency as historical outlier

With 190 innings in '17, he'd become just third pitcher to accomplish feat five times after 40

Colon heads to free agency as historical outlier

When teams evaluate players, they often do so with comparables. That is, what have other players with similar ages and skill sets done in the past? For example, if a team is looking to spend on Mark Trumbo this offseason, they'll certainly check out how previous right-handed hitters with big power and limited defensive value have done heading into their age-31 season, as Trumbo is.

It's a smart and time-tested strategy. It adds valuable information in just about every case. But what happens when there are no comparables? What about the rare case where a player is doing something that has rarely happened? Welcome to the fascinatingly unprecedented free agency of Bartolo Colon.

You know the Colon story by now, one would assume. He arrived in the bigs with Cleveland in 1997, and after nearly a decade of solid performances, he fell completely off the map after winning the 2005 American League Cy Young Award -- failing to pitch 100 innings in any of the next four years, and missing the '10 season entirely. Since Colon resurfaced with the Yankees in '11, he's reinvented himself as something of baseball's resident cartoon character, dominating highlight shows when he does things like behind-the-back flips and becoming the oldest player to hit his first home run.

Colon fires seven strong innings

Yet for all the entertainment value, it's easy to forget that Colon has been a pretty valuable pitcher, too. Only seven pitchers have thrown at least 190 innings in each of the past four seasons, and any list that includes Madison Bumgarner, Jon Lester and Max Scherzer is impressive. Of 175 qualified pitchers over that span, Colon's 3.59 ERA ranks 59th, slightly ahead of Dallas Keuchel and Justin Verlander. Plus, since a weak starting-pitching market counts Rich Hill, Jason Hammel and Jeremy Hellickson as the cream of the crop, Colon is inarguably one of the better options out there.

And yet, Colon will be 44 in May. He throws one of his three fastballs between 90-99 percent of the time, depending on the source. If he throws another 190 innings next season, he'll be one of only three pitchers to do it five times after the age of 40 -- and one (Phil Niekro) was a knuckleballer. So how do you value that?

That's why, when MLB.com recently looked at a list of Top 25 free agents, Colon didn't appear. It's not because he's not expected to be valuable in 2017. It's because that list was ordered based on projected Wins Above Replacement based on a player's age and performance over the past three seasons, weighted with the most recent season counting six times, two years ago the times, and three years ago one time. It's a way to apply baseball-reference.com's WAR to Tom Tango's Marcel projection system.

It's a good system that works well, except that Colon is such an outlier at his age that he breaks all historical precedent. Let's explain, shall we? For this exercise, the dataset included 26,543 pitcher seasons dating back to 1950. In order to find pitchers who were similar to Colon, we ran the following filters:

1. Find seasons that were worth between 2.9 WAR-3.9 WAR. (Colon, per baseball-reference.com, had a 3.4 WAR season for the Mets in 2016, making him the midpoint.) That allows us to start with only a base of pitchers who were coming off similarly above-average seasons.

2. Sort the two previous seasons to show pitchers who were within one-half WAR in either direction of Colon's previous two years, also making him the midpoint. This allows us to further distill down to pitchers on similar trend lines, though of course not weighted as heavily as the most recent season.

3. Sort by age, oldest first, since of course we're trying to find Colon comparables, and pitchers who may have had similar seasons between ages 24-26 don't really help us out here. We should note that no one had a similar path to Colon over 40 years of age.

So what are we left with? Here's the four oldest starting pitchers who satisfy our requirements. Why four? You'll see. It's an… interesting group of pitchers, and that's sort of the point. Colon stands alone.

Colon's impressive outing

Mike Morgan, 1998
That was his age-39 season, so while he was younger than Colon is now, he also made his debut in '78 at a shocking 18 years old, so Morgan had been around just as long. Though a 4.18 ERA for the Twins and Cubs in '98 may not seem great, do remember how explosive the offensive environment was at the time, so he'd actually been 13 percentage points better than average, and with a 3.2 WAR. He'd pitch four more years, though only one as a starter, and while his '99 record with Texas was 13-10, that was far more about the Rangers' offense than his performance, given a 6.24 ERA.

R.A Dickey, 2010
We're already into knuckleballers, and Dickey won't be the last. In 2008-09, Dickey had been a spot starter for Seattle and Minnesota, and at 35, his '10 Mets debut was his first truly valuable Major League season, with 3.6 WAR and a 2.84 ERA. The next offseason, he signed a two-year deal with the Mets, won the '12 National League Cy Young Award, and he was traded to Toronto in the deal that brought Noah Syndergaard and Travis d'Arnaud to New York. Now 42, Dickey just gave the Blue Jays four solid seasons.

John Smoltz, 2003
Smoltz is a Hall of Famer, but his path there wasn't straight. After a decade as an Atlanta starter, he missed the 2000 season due to injury, then returned as one of baseball's better closers from '00-04. He then returned to the rotation in '05 and stayed there until he retired in '09, at age 42.

Charlie Hough, 1982
And here we are with another knuckleballer. Hough spent most of the 1970s as a Dodgers reliever, spending his first full year in a starting rotation with Texas in 1982 at 34. It was a very good year, worth 2.9 WAR, and he'd go on to start the first game in Marlins history and pitch until he was 46.

Everyone else was below 34 years old, and half of the four comparables we have were knuckleballers. It's a good way to show just how Colon breaks all of our models. There's no one like him, and there may never have been anyone like him. It's a good lesson that if a pitcher has pinpoint control, their velocity and arsenal aren't as vital as you'd think. (Only Clayton Kershaw found himself in fewer 3-0 counts last year, for one example to show how Colon gets by.)

Any team signing Colon this offseason will be taking a risk, because there's nothing to compare him to, and teams like data points. Colon's agent will surely point to the actual on-field production over the past few years. Some team will sign him, for one or two years, and for relatively few dollars. That team may end up with a bargain. They'll just have little to compare it to.

Mike Petriello is an analyst for MLB.com and the host of the Statcast podcast. This story was not subject to the approval of Major League Baseball or its clubs.