FanPost

Hoffman's latest blown save was much worse than you think

On Tuesday May 18th, 2010, Trevor Hoffman entered to pitch the bottom of the 9th inning against the Cincinnati Reds. Starting with a 2 run lead, the Brewers closer gave up a single, a two-RBI home run, a double, a walk, and another single to score the winning run for the Reds. It was arguably the most demoralizing blown save in recent Brewers memory.

However, you already knew that. Here's something you don't know: it was much worse than you think.

Using the Pitchf/x data from texasleaguers.com, I compared Hoffman's latest blown save to both his incredible 2009 season and his encouraging 2-save performance in early May. The data show that Hoffman's mechanics have taken a big step backward. On Tuesday against the Reds, both his fastball and changeup lacked any significant horizontal movement.

Hoffman's changeup was especially awful. It is common to say that a pitcher was "throwing batting practice" when they have a bad outing, but it usually is a figure of speech. On Tuesday, Hoffman's changeup was literally indistinguishable from a batting practice pitch thrown by coaches in their 60s. Don't give the Reds hitters too much credit for the blown save. Most people reading this could have hit Hoffman's changeup on Tuesday night. It was that bad.

Flashy Pf/x charts and graphs after the jump...

From a right-handed pitcher, both fastballs and changeups move in on right-handed batters. From the catcher's perspective, they move to the left. Hoffman's changeup moves more (about 3.5 inches) than his fastball (typically 1.5 - 2.5 inches). That difference in horizontal movement, plus the difference in vertical movement (6-7 inches) combined with the speed difference (10 mph), results in an unhittable combination. Both pitches look the same coming out of Hoffman's hand, forcing the hitter to guess where and when the ball will be in the hitting plane. In 2009, one out of every five of Hoffman's changeups caused batters to swing and miss (20.6% whiff).

Pitchf/x tracks pitch movement relative to a "reference pitch," which is a theoretical pitch thrown at the same speed in the same location with no spin-induced movement. This theoretical reference pitch has 0.0 inches of horizontal and vertical movement; in the real world, it would be a standard backyard hard toss. The location of each pitch is tracked when it crosses the plate, and its horizontal and vertical movement is recorded in inches relative to that theoretical reference from the catcher's perspective.

Horizontal movement is negative if the pitch's spin causes it to move left from the catcher's perspective. It is positive if the pitch moves to the right.

Vertical movement is positive if the pitch's backspin gives it aerodynamic lift, causing it to fall more slowly than a pitch without backspin. From the catcher's perspective, it appears to rise because the human brain expects thrown objects to sink due to gravity. Positive vertical movement indicates a "rising" pitch. Similarly, vertical movement is negative if the pitch's spin causes it to sink more quickly than a pitch without spin.

Movement is independent of location, which is very important to understand. In Pitchf/x, movement only measures the deviation from the theoretical reference pitch.

 

With all of that in mind, here's a blink comparison of three sets of data. The figure changes every 5 seconds.

  1. All of Trevor Hoffman's pitches in the 2009 season, charting spin movement by pitch type.
  2. Hoffman's pitches in two encouraging saves: 5/1/2010 at San Diego and 5/7/2010 at Arizona.
  3. Hoffman's pitches in the blown save at Cincinnati.

Image and video hosting by TinyPic

It is difficult to believe that the data from May 18th is from the same player as the other two sets. After my last fanpost, in which I showed that Hoffman's changeup was uncharacteristically inconsistent, I noticed that he was improving his mechanics with each subsequent outing. It looked like he was on the road to recovering his dominance of hitters. His two saves on 5/1 and 5/7 were encouraging; Hoffman's individual pitches were very similar to those thrown in 2009, at least in terms of movement and consistency. He was still having some location problems, but it seemed like his mechanics were much more sound.

He regressed a bit in a non-save situation on 5/12 against the Braves, giving up 3 runs and 2 walks while facing 7 batters in a loss. He looked sharp in a hold situation against the Phillies on 5/16, retiring 3 batters in order.

However, the road to recovery fell into a gaping chasm on 5/18 against the Reds. From the moment Hoffman took the mound, his fastball and changeup had no significant horizontal movement. From the hitter's and catcher's perspective, those pitches were straight as an arrow. The Reds hitters simply waited for Hoffman's 73 mph unmoving changeup. On Tuesday, that slow pitch was literally indistinguishable from a weak batting practice pitch.

 

Want another view of how bad Hoffman's pitches were on Tuesday? Here's the view from the top.

Pitchf/x tracks the location of each pitch from the release point to 5 feet beyond the plate. The top pitch virtual plot shows the average path of that pitch type over the entire data set. If you could watch the game suspended from a crane above home plate, this is what you would see.

Looking at Hoffman's changeup from 2009, you can see the average 3.5 inch horizontal movement in on a right handed hitter (or left, from the catcher's perspective). You can also see that he paints the corner of the plate with it. Next is Hoffman's changeup from the two saves in early May, 2010. He's still catching too much of the plate, but the pitch movement is similar to 2009.

Image and video hosting by TinyPic Image and video hosting by TinyPicImage and video hosting by TinyPic

Finally, in stark contrast, the last plot is from the May 18 blown save. Not only was the changeup arrow-straight, but average location was right down the middle of the plate. At 73 mph, I think most people reading this could hit that pitch. Professional hitters will have solid contact every time, without a doubt.

 

Hoffman's fastball took a similar dive in quality, though it didn't get hit quite as badly. You can see that he had decent movement and great location with the fastball in 2009. The fastball had nice movement but was catching too much plate in the two saves in early May.

Image and video hosting by TinyPic Image and video hosting by TinyPicImage and video hosting by TinyPic

Finally, just as with the changeup, Hoffman threw his fastball straight as an arrow and right down the middle of the plate on Tuesday. At 85 mph, it wasn't hit as much as the changeup, but it's still close to batting practice speed.

 

If Hoffman is not hurt, I have no idea how a Hall of Fame pitcher suddenly loses the movement on two of his pitches for an entire inning. He must have realized what was happening, because he has a perfect view. I can understand location problems. I can understand mechanical problems leading to inconsistency. However, this is a complete mystery to me. Age doesn't account for this, unless his memory is failing so badly that he suddenly forgot how to throw his two main pitches. I don't think I have ever seen a meltdown of this proportion; Hoffman went from pitching master to pitching novice in just a few days.

If this cannot be corrected, Trevor Hoffman is not even valuable in a middle relief role. If he bounces back from this outing and has a successful season, it will be an incredible comeback. Hoffman is in a very bad place right now, as an athlete. I'm rooting for him, but I am not terribly hopeful.

And if anyone puts Hoffman in a save situation right now, they should be fired. Without a solid record of good performance to prove that May 18th was just a fluke, he can't be trusted.