Throwing hard but not well: what makes velocity useful?

Baseball is, obviously, a different game than it was in the 1990s. The end of widespread chemical enhancement has ushered in a new dead(ish)-ball era of lowered offense and pitching dominance, but the change has as much to do with the pitchers as the hitters. New and better training regimens have made the 2010s into the decade of velocity. Starters and relievers alike throw harder than ever before. In just the 10 years of the PITCHf/x era, average fastball velocity has increased more than a mile per hour.

It's generally accepted that velocity helps a pitcher out. Our eyes confirm this—the best relievers of the last five years are probably Dellin Betances, Aroldis Chapman, and Craig Kimbrel, with career average fastball velocities of 96.7, 98.8, and 96.7 respectively—and research confirms it. It's also generally accepted that velocity alone isn't enough. From Baseball Savant, here are the 20 starting pitchers with the highest percentage of pitches thrown at least 95 MPH (minimum 250 pitches):

There are some very good pitchers on this list. Noah Syndergaard owns the highest percentage and is in the middle of a breakout season featuring dazzling velocity. Jose Fernandez is outstandingly fun and also just outstanding. Danny Salazar and Gerrit Cole have been quite solid, and I've heard that Stephen Strasburg kid is all right.

There are also some real stinkers. Joe Kelly, Wily Peralta, Luis Severino, and Andrew Cashner have all been pretty bad. While there are reasons to be optimistic about all of them, they almost certainly have been given more opportunities to demonstrate their potential than their softer-tossing peers and haven't made good on those opportunities yet. That same feeling of potential can make them very, very frustrating as well. Joe Kelly has great stuff; why is he so gosh dang bad?

Clearly, velocity alone isn't enough. What does it need to be paired with for a starter to find success? Of the above 20 pitchers, eight have an ERA-, FIP-, and DRA- better than average, whom we'll call the good group, and six have been worse than average in all three categories, whom we'll call the not-as-good group.

A common feeling about pitchers with high velocity is that, if they don't pan out as starters, they can always be pushed to the bullpen and converted into elite relievers. While that might be true for individuals in these groups, as a whole there's nothing that suggests one is better suited to relieving than the other (besides their overall quality). The platoon split for the bad group is only slightly worse than the good group—.071 wOBA versus .066—and given their higher overall wOBAs, it's a negligible difference. The bad group has just as many pitches used at least five percent of the time as the good group, 4.2 versus 4.1, and a pretty similar velocity gap between their fastest and slowest pitches, 12.8 MPH versus 11.3 MPH. It's not as simple as the good group having something in their arsenal that the other group doesn't.

It's generally believed hard pitches are more effective when thrown high in the zone, but per Baseball Savant, 36.7 percent of the not-as-good group's fastballs are high, versus 32.7 percent of the good group. That said, it's not throwing high that makes a hard fastball effective; it's throwing it precisely to a very small portion just above the top of the strike zone. Miss low by a few inches on a high fastball, and the pitch that was supposed to induce a whiff or a pop up instead gets crushed. Control and command are tricky to measure, but the good group does have a higher zone rate than the bad, 48.4 percent to 45.7 percent. They also have a lower walk rate, 7.9 percent to 9.6 percent, but since that's one of the components of FIP and one of the way these groups were selected, I'm hesitant to put too much stock in it.

It might also have less to do with the control of these pitchers and more their confidence in avoiding crushing contact when their control isn't quite good enough. The good group has a substantially lower contact rate than the bad group, 74.8 percent versus 82.7 percent, which seems like it should mean they have a greater margin of error if they do miss their spot. Again, it's part of how they were sorted, but the good group has a much, much higher strikeout rate than the bad, 28.2 percent versus 16.4 percent. If I had to guess at the main difference between the two, it would be the quality of their stuff, as measured both by control and by movement or deception.

Velocity isn't the only thing that makes a pitch effective, but it is the most easily measured. I suspect that means that when a pitcher can fire up the radar gun but can't control his pitches, avoid hard contact, or generate whiffs, he gets a longer leash than someone who has one or more of those skills but can't throw as hard. Velocity comes in a nice, neat package, complete with a number for every pitch, which makes it harder to dismiss a guy who has it, even if he's got nothing else. In other words, get used to frustration, because the hard throwers who can't do anything else probably aren't going away anytime soon.

. . .

Henry Druschel is a promising but disappointing Contributing Editor at Beyond the Box Score. You can follow him on Twitter at @henrydruschel.