Cole Hamels provides one of the most unique and challenging pitching profiles of just about anyone in the NL.  Although his most praised pitch is his devastating change up, which moves all over the place and drops when it is at its best, he also features a strong fastball with varying velocity in the low-90s.

These facts, and his approach to pitching, render Hamels rather difficult to classify: is he a junkball pitcher because his change up is his best pitch? This idea is planted in my head from the Neyer / James Guide to Pitchers, in which arguments are provided for the classification of a junkball pitcher. They suggest that usually when you call a pitcher a junkballer, you're saying his change up is his best pitch. I'm not sure that I completely agree with that, simply because I also believe that certain junkballers can (and do) have good fastballs. They simply don't have fast fastballs. 

To this point, a pitcher like Chris Young for the Padres, Hamels, Scott Olsen of the Marlins, and Ted Lilly of the Cubs provide some example of pitchers were relatively slow fastballs that nevertheless utilize their fastball in effective ways. Some offset the velocity with a deceptive delivery (like Young, who stands incredibly tall on the mound, rendering his fastball rather sneaky), whereas others offset their velocity with a very effective change of speed or breaking ball. 

Glancing at his player card, compiled by Josh Kalk utilizing pitch f/x data (an excellent resource), we can see that Hamels' fastball averages just under 91 MPH in velocity for the 2008 seasonIn terms of movement, it rides high into left-handers, and slightly away from right-handers (movement in this case measures Hamels' pitches against a spinless ball at the same velocity, which means that a pitch that "rises" simply employs the type of spin that keeps it from dropping as a spinless ball would).

Thus, from the player card, we learn that Hamels' fastball is a cross-seam fastball that rises (12.09" compared to a spinless ball) in against a left-hander (2.68" horizontal movement in on a lefty, compared to a spinless ball). It is clear from these readings that Hamel is not throwing a sinker or a cutter as his main pitch, like a Brandon Webb or a Jamie Moyer, and if we compare his fastball to some  hard-throwers we can confirm that its movement patterns are indeed those of a cross-seam, rising fastball:

Hamels' fastball (90.92 MPH): rides up 12.09" and in 2.68" on lefties

Moyers' sinker (82.06 MPH): rides up 6.13" and in 9.01" on lefties

Webb's sinker (88.61 MPH): rides up 1.28" and in 7.22" on righties

Ben Sheets' fastball (93.54): rides up 10.76" and in 4.17" on righties

Carlos Zambrano's power sinker (91.84): rides up 4.17" and in 9.57" on righties

Reading these charts, we can begin to distinguish some characteristics of fastballs. What is commonly a cross-seam, riding, or rising fastball, usually the pitcher's fastest pitch, will "rise" according to pitch f/x numbers, like Sheets' fastball and Hamels' fastball listed here. By comparison, we can see that various sinking fastballs do not "rise" as high as the fastballs of Sheets and Hamels, and they also have much more extreme horizontal movement (in or out, usually in on a same-handed batter).

Using these characteristics, I'd like to point out three traits off the top of my head:

-Sinking fastballs are thrown by many pitchers at almost any velocity, and must therefore be discerned by their horizontal movement, and their low vertical movement

-Power fastballs and junkball fastballs, when both thrown as cross-seam, rising or riding pitches, have uniquely similar movement measurements

-Power fastballs and junkball cross-seam fastballs must therefore be determined by velocity, not movement.

This is where things get interesting for a pitcher like Hamels, who throws a rather obvious cross-seam fastball, and does not extensively throw a sinker. His average velocity is a full 2.50 MPH slower than someone like Sheets, and if we show other hard fastball throwers, Hamels' average fastball looks even slower:

C.C. Sabathia's fastball (94.87 MPH): rides up 8.48" and in 5.67" on lefties

Radhames Liz's fastball (94.62 MPH): rides up 9.28" and in 4.94" on righties

Bobby Jenks' fastball (95.15 MPH): rides up 9.64" and in 2.61" on righties

These examples display, further, the trend of a cross-seam, rising fastball, rather than a sinker with lower and more horizontal movement.

Thus, in order to classify Hamels, we cannot conclude from the movement of his fastball that he is a sinkerballer, so we must look at his velocity. This is where a cut-off is difficult. It will probably be rather arbitrary, as I do not have a source right in front of me that says "the average fastball in the MLB is 'x' MPH." Further issues with that statement might be that it fails to distinguish systematically between pitchers that are throwing cutters, sinkers, and riding fastballs with different frequencies. 

For instance, if all sinkerballers like Brian Shouse, Dave Bush, Jeff Suppan, Jamie Moyer, Mitch Stetter, and Chad Bradford are included in the group of "Pitchers who throw a fastball," the average fastball velocity will be lower because there are a rather strong number of relievers and pitchers in general who throw sinkers anywhere from 80-to-89 in the MLB. 

 However, if we simply discriminate against those sinker- and cutter- pitchers, and only calculate cross-seam fastballs, or fastballs that exhibit the characteristic of a rising fastball, we might end up with an "average" fastball that does not reflect the actual average topspeed velocity of most MLBers.

And so, I think the case of classifying Cole Hamels reveals at least two very interesting points about a fastball:

(1) There are so many different types of fastballs and fastball movements thrown by and exhibited by MLB pitchers that lumping all "fast" pitches together under the rubric "fastball" paints in broad strokes over very nuanced distinctions.

(2) Top-speed velocity does not always need to equal a "fastball." Most people probably equate a fastball in their imagination as a strong, cross-seam, rising fastball, exhibited by someone like Nolan Ryan or C.C. Sabathia or Ben Sheets or Bobby Jenks. But I'm reluctant to simply call each player's topspeed pitch a "fastball" because it could easily miss the fact that such a pitcher does not actually throw a fastball.

For instance, if you watch Brian Shouse frequently, and get a chance to see him up close, it's rather clear that he is not throwing a fastball. And I don't simply mean that his topspeed velocity cracks 82 on a good day. What I mean, rather, is that he throws so many variations of a sinking pitch and a slider that the movement and characteristics of those pitches are not indicative of a "fastball" in any way.

So, maybe "Fastball" by itself could simply mean "top speed velocity." And then we'd have to augment that with other words, like a sinking fastball, a riding fastball, a cross-seam fastball, a 2-seam fastball, a cut fastball, or split-fingered fastball, etc.

But in any term, it is clear that Cole Hamels throws a fastball. A good 'ol riding, rising fastball, that he used aggrssively inside against right-handers yesterday against the Brewers, and that he augments with that excellent change-up.

If we guess that the average MLB fastball, or perhaps topspeed velocity, is somewhere around 90 or 91 (I believe I've heard that thrown about somewhere), does that mean Cole Hamels is a junkballer? A pitcher whose fastball averages out at league average? Of course, anyone that looks at Hamels' results would find it absurd to suggest that Hamels' fastball is "average"....even if the velocity is such.

So, maybe I should simply agree with Rob Neyer's and Bill James'  pitching guide, and call a junkball pitcher a pitcher whose best pitch is a change up. But I'm not entirely certain that that captures what Cole Hamels accomplishes on the mound, or the bliss with which I desire to claim his fastball for the junkballers' camp....