Welcome to projection season. This is the time for magazines, books and, of course, websites to make educated (and sometimes not-so-educated) guesses about the future of the burgeoning baseball season. Many projection systems use comparable players from baseball history go guide their projections. In other words, if you want to know how Mike Napoli will do this season, look back into the dusty volumes of baseball history and find a group of players who have posted similar careers to Mike Napoli. Look at what they did, weigh that data with Napoli's career trends and you've got yourself a solid guess as to how his 2013 season will turn out.

But what happens when the subject of your projection is unique? What happens if you go back into those dusty volumes but can't find anyone comparable? This is the conundrum facing baseball analysts when it comes to Bryce Harper. The Nationals outfielder hit 22 homers while posting an .819 OPS last season. That's quite good --19 percent above average to be exact -- but hardly special and certainly not unique. What makes Harper unique is his performance combined with his age.

How young is Bryce Harper? If he were a comedian he'd be doing Tuesday open mics at the Giggle Hut off Route 4 in Hackensack. If Bryce Harper were a house he'd be a stack of wood piled up in a muddy clearing. As a baseball player he should be just starting out, either by breaking into a college lineup or, if he was drafted out of high school, enduring long bus rides and $5 per diems in the deep minors. But Bryce Harper is not a college baseball player, and while he's young enough to know the kinked neck one gets from sleeping on a Greyhound bus, he's not in the deep minors either. He's a 19-year-old major leaguer with enough recent magazine covers to make you think he's either a movie star or a deceased South American dictator.

Last season Harper came up in late April and played the remainder of the season. Playing essentially a full season at 19 isn't unique, but it is unusual. Ken Griffey Jr. played 127 games as a 19-year-old in 1989. Mickey Mantle played 96 games at 19. Ty Cobb played 41 games at age 18 with 98 more the next season. If you include Harper, 34 players since 1904 have received 200 or more plate appearances at age 19. Of those players, Harper's OPS ranked third. So, if you use OPS as the final arbiter (which it isn't -- but for our purposes it'll do), Harper's 2012 season was the third best season by a 19-year-old in the last 110 years. That may not be exactly unique, but it's getting there.

The exciting thing about Harper isn't just that he's already very good -- if he kept putting up .850 OPS's for the next 15 seasons he'd have himself an excellent career -- but that he's this good at an age where the normal aging curve for hitters tells us he has lots of improvement yet to make. Imagine if Harper followed that typical aging pattern and kept improving each year until his late 20s. Considering his starting position and already above average skills, winning multiple MVP awards wouldn't be out of the question. It should not be a surprise that multiple MVP awards are precisely what many expect of him.

But "multiple MVP awards" isn't a very specific projection. How do we get a detailed sense of a player's future when so few players have followed the same path?

There are essentially two different ways to project out a young player's career. The first is to use a projection system, like Baseball Prospectus's PECOTA, which, according to Baseball Prospectus, "projects player performance based on comparison with historical player-seasons." It gets more complicated than that, but that's the main gist. PECOTA sees Bryce Harper's production falling off slightly in 2013. That's a surprising projection given Harper is a year older, very young and talented, and presumably should be getting better. So why does PECOTA hate Bryce Harper?

It doesn't, of course. When projecting your average player's age-27 season, PECOTA has literally thousands of possible player seasons to pick from. In other words, it can get very specific. With Harper it has far fewer options. One of the comparable players for Harper is Alex Rodriguez, who played his first full season with the Mariners when he was 20. He was amazing, hitting 36 homers with a slash line of .358/.414/.631. The following season he dropped off to a still good but nowhere near as good 23 homers with a slash line of .300/.350/.496.

But Alex Rodriguez isn't really all that comparable to Harper, right? Rodriguez came up early like Harper and possessed prodigious power like Harper does, but that's pretty much where the similarities end. Rodriguez possesses a different body type and plays (and played) a different position than Harper does. He actually came up to the majors a season earlier than Harper, and he hits from the other side of the plate. This illustrates the problem with projecting a player with Harper's specific skill set at so young an age. Where projection systems can usually be very precise, with Harper they can't; the data just doesn't exist. Therefore projection systems can't be as certain, and the range of possible outcomes is much greater than it normally would be.

"The error bars for a player like Harper are huge," says Dan Szymborski, a writer at ESPN and creator of the ZiPS projection system featured at FanGraphs. "And since there are so few players of Harper's abilities at such a young age, we probably don't even really know just how huge the error bars are."

To further confuse the issue, recall the 34 players since 1904 that received 200 or more plate appearances at age 19. By OPS, Harper ranked third on that list. Here are the age-19 seasons of the two above Harper and the two below.

 

Name

Age

Batting Average

Home Runs

OPS

OPS+

1.

Mel Ott

19

.322

18

.921

139

2.

Tony Conigliaro

19

.290

24

.883

137

4.

Mickey Mantle

19

.267

13

.792

117

5.

Cesar Cedeno

19

.310

7

.790

114

Now here is how those same players did in their age-20 seasons.

 

Name

Age

Batting Average

Home Runs

OPS

OPS+

1.

Mel Ott

20

.328

42

1.084

165

2.

Tony Conigliaro

20

.269

32

.850

133

4.

Mickey Mantle

20

.311

23

.924

162

5.

Cesar Cedeno

20

.264

10

.690

97

Two players, Mantle and Ott, took big steps forward. Conigliaro hit more homers but his overall production fell off, and Cedeno dropped to a below-average offensive player. Playing major league baseball at age 19 is a tremendous accomplishment, but even that isn't a gold ticket to immediate superstardom.

Even with the admittedly nebulous nature involved in projecting Harper, projection systems can help show us a range of possibilities. But while projection systems can cut through human biases, they can also miss specifics that a good scout will catch and take into account*.

Keith Law is a senior baseball analyst for ESPN in charge of scouting. He said he thinks Harper "will be an impact player in the middle of a lineup for a very long time, the kind of player who hits 40 homers in a few seasons, wins an MVP award or two, and at worst ends up garnering some Hall of Fame discussion."

*This is why smart teams gather and take into account both perspectives when developing players and building a major league roster.

The reason why, according to Law, is Harper's "unusual combination of overall athleticism and baseball-specific skills." It's that package that makes Harper unique, or at minimum, extremely difficult to project. For his part, Law feels Harper is one of a kind. "Has there been a player like him specifically? No, I tend to think most high-end players like Harper or Mike Trout are unique. Stars tend to have very specific skill sets that we haven't seen before."         

People learn by experience. But when something comes along that we haven't encountered before, it's hard to know what to make of it. Harper represents those uncharted waters. 

So how do we project Bryce Harper's 2013 season if he is unique, or something very close to that? The answer is we look at the numbers and use them as a guide. We listen to the scouts and try to learn what they think about the player and why they think it. We take that package of information, apply our own biases, and there you have it. But in the end, the only way to project a player as young, as supremely talented and as good as Harper already is involves accepting the fact that, while we can take educated guesses and apply past lessons, when it comes down to it, we just don't know. Hey, that's why we watch the games.