Praising Cliff Lee's Curveball and Fixing His Changeup

Cliff Lee has not only been one of the bright spots of the 2013 Phillies, but he's one of only four players that can be trusted for a good start, at-bat, or appearance (Brown, Papelbon, Kendrick).

Among NL pitchers, Lee ranks 3rd in WAR and 2nd in IP with a 2.55 ERA, 1.01 WHIP, and .223 BA against. Those marks will certainly earn him an All-Star nod and depending on how the next month goes for the beleaguered Phillies - a trip to a contender.

The most amazing part of the incredibly passionate Phillies fan/Cliff Lee love affair is the lack of admiration for his curveball. In the 2011 and 2012 seasons, Lee's curve yielded a .151 AVG and .232 SLG. This season, it's been the decisive pitch in 31 at-bats, 17 of which are strikeouts. We go crazy for Hamels' changeup, but Lee's curve is just another pitch. What gives, Philly?

Lee's dominance can be partly explained by his Pitch Movement Chart. Considering 7.8% of his pitches are curveballs, we can easily say he has five pitches that hit three of the four quadrants.

Again, Lee's curveball has been spectacular, registering an 18.8% whiff rate (% of swings that don't make contact), .226 AVG, and a gorgeous arc on the PMC. When right-handed hitters have two strikes, they must be prepared for all five pitches - one of the reasons his curveball is so effective. While other pitchers rely on an 'out pitch' with two strikes, Lee can throw anything, leaving a defensive hitter almost unable to make quality contact (only his changeup allows a BA over .182 with two strikes to righties).

The most amazing aspect of Lee's superb 2013 performance is the lack of success from his changeup, a pitch he throws 14% of the time.

That changeup posts a .296 AVG and .426 SLG. While you can argue that he should throw it less, you can also argue that its mere existence creates more confusion for the hitter, making the other four pitches more effective. Lee hasn't thrown his curveball to a righty in any 3-ball count this season, relying on his two-seam (34), changeup (15), and cutter (7). Take away the changeup and a hitter knows they're getting 91mph tailing right or 87mph cutting left. Mix in 84mph with similar action to the two-seamer but with more drop and tail and we've got a hitter guessing - but not off-balance.

Lee's changeup averages 84.9mph. His four-seam and two-seam both average 91.3mph and his cutter averages 87.5mph. He throws five pitches, but four of them are within 6.4mph of each other. If Lee's changeup were to go from 84.9 to 82.9 with the same arm action and movement (likely more movement with the slower speed), he would likely see better results.

The only Phillies with a smaller velocity gap between fastball and changeup are Antonio Bastardo (he never throws it), Roy Halladay (was hurt/sucked), John Lannan, and Chad Durbin (he really sucked).

Cole Hamels, who boasts one of the best changeups in baseball, throws a 92.3mph fastball and 84.7mph changeup (7.6mph difference). But it's Lee, whose changeup is far from one of the best, who gets more horizontal movement and nearly as much vertical drop as Hamels. The only chartable difference is the velocity gap.

Fooling a hitter only to have him still hit the pitch hard because it wasn't slow enough - even though its entire purpose is to be slow enough - is quite frustrating.

(Cliff Lee Heat Map 2013 - Changeups to RHH - Catcher's POV - Fangraphs.com)

First, Lee rarely throws his changeup to lefties (3% of pitches, 17% to righties). So, I'm only going to analyze right-handed hitters.

Second, location is an issue. Even though he hardly throws it on the inner-half or upper-half, where change-ups go to die, righties are still hitting .300 off it.

Only 28.7% of his changeups to righties end up being a ball. That's absurdly low. And poor strategy. And it needs to drastically change.

Nibble with the change. Keep it on the outer or lower black or off the plate. We all know how much he loves to throw strikes, but it's too his own detriment with this particular pitch. Only his cutter sees more of the plate than his change. If your worst pitch sees more of the plate than almost any other, get it off the plate.

That part of the heat map where it's yellow - just above the bottom of the zone and just inside the right edge - needs to be on the black lines. It's not asking too much. It's a necessity for an 84mph pitch in Major League Baseball.

Also, 177 of his 192 changeups have come with 0, 1, or 2 balls to the batter. He can throw that pitch off the plate without it being a walk - but he still throws it for a strike.

Would nibbling and thus throwing more balls throw off his entire game? Or would it make him even more effective by hiding his worst pitch and making it better? I'd sure like to see it play out.

I did a chart for Lee's zone usage and statistics. This doesn't have much relevance to his off-speed pitches, but I did it and thought you may find it useful anyway - so here you go.