Author Archive

Beating the Shift (& Physics)

After reading some questions in Dave’s chat today (May 7), and in response to never-ending questioning from un-informed commentators across baseball, I wanted to provide what I think is a very simple explanation for why groundballs are so often pulled. Here goes:

In terms of the direction a ball travels after hitting a bat, there are three factors:

  1. Vertical contact point on the barrel – below the sweetspot of the barrel is a groundball, above the sweetspot is more of a flyball
  2. Horizontal plane at contact – is the bat pointed directly perpendicular to the angle of the pitch and batter’s body, is the batter’s swing out in front, or is he behind?
  3. Vertical plane at contact – is the bat plane directly parallel to the ground, pointing slightly up towards the sky, or pointing slightly down towards the ground?

To start, let’s fix two of the three factors above and leave the third factor as the variable. Let’s assume that since we’re talking about groundballs, factor 1 will be set so that the contact point is slightly below the sweetspot of the barrel, resulting in a groundball. Let’s also set factor 2 and assume the batter squares up the pitch, so that in the horizontal plane his bat makes contact with the ball when it is directly perpendicular to his body. And now let’s focus on factor 3, the vertical plane.

In practice, the vertical plane is already set. Batters (almost*) always make contact with their hands above the head of the bat, so that it is vertically pointing downwards. So given that we’ve set factor 1 to be below the sweetspot on the bat and factor 2 to be square horizontally, we know that the groundball WILL be pulled. Why? Let’s look at it at the extreme to illustrate the point.

Visualize the vertical plane being taken to its extreme, so that the bat is pointing directly downwards. Now, since we’ve set the other two factors in place, we know that the ball will make contact with the bat slightly towards the batters side (which would be below the sweetspot were the bat at a more normal vertical angle). Which way will the ball deflect? Always towards the pull side, and it may actually hit the batter. Now clearly, this is an exaggerated example, because a batter wouldn’t make contact with a ball while his bat is completely vertical. However, the same physics apply when the bat is pointed only slightly downwards vertically, just to a lesser degree. So the point that we’ve established is that there is a structural, physics-based reason that groundballs tend to be pulled, because the bat is (almost*) always pointed slightly downwards at contact.

With that established, let’s take a step back to factor 2. This is the factor which most heavily influences whether or not the groundball will actually be pulled. If the batter is late or behind on the pitch, than in the horizontal plane (factor 2) the bat will be pointed more backwards toward the catcher than if it were exactly perpendicular. In this case, the ball would be pushed the opposite way. On the opposite end of that spectrum, if the batter is early on the pitch and his bat is pointing more forwards towards the pitcher than backwards towards the catcher, the groundball will be pulled more extremely.

So how early or late the batter is on the pitch is clearly an important determinant in whether the groundball will be pulled, pushed, or hit right back at the mound. But remember what we’ve already established – there is a natural tendency for groundballs to be pulled because of factor 3. So if a batter is to take a groundball the opposite way, he must not only be late on the pitch (factor 2), but he must be late enough to overcome the natural pulling force caused by factor 3.  And the more his bat is pointed downwards, the larger that natural pulling force is. The lower and more inside the pitch is, the more the bat will likely be pointed downwards, meaning that the batter must be that much later on the pitch to push it the opposite way.

I’m sure a lot of you will read this and say “duh”, because it is all pretty intuitive and this is FanGraphs after all. But for those of you who have just learned something, any time you ask why a batter can’t just push his groundball away from the shift, remember that he’s having to overcome the natural laws of physics, and, per Dave at 12:58 this afternoon, “it’s not easy to do.”

* All analyses exclude the superhuman skills of Evan Gattis


Appreciating Mike Trout

I apologize up front for beating a dead horse with a stick, but Mike Trout is incredible.

As of July 11, he’s sporting the following line:

  • 320/399/560 164 wRC+, with 21 SB (87.5% success) for good measure

Last year, Mike Trout’s amazingness was well documented, especially on this site. His 2012 (should be MVP) season line:

  • 326/399/564 166wRC+, with 49 SB (89.1% success)

Notice anything about those two lines? They’re basically identical.

At first glance, that’s not particularly interesting. He’s really really good, as we all knew. But what makes it interesting is that he’s actually shown significant signs of improvement in seemingly getting to the same place as last year. He’s walking slightly more (11.1% rate vs. 10.5%), but more importantly, he’s cut down on his K% by over 5%, from a slightly worse than league average 21.8% to a better than average 16.7%. Hence, despite his BABIP dropping by a meaningful 26 points from .383 to .357, he has maintained the exact same AVG and OBP.

Basically, he’s replaced some BABIP luck from last year with actual improvement. His BABIP is still well above league average (currently ranking #15 among qualified), but given his unique combination of speed, power and nearly 23% line drive rate (league average 20.9%), I’m inclined to believe a .350 BABIP is a reasonable true talent level.

I’ve focused on his BABIP and K%, so let’s dig a little deeper into those two rates. In terms of BABIP, his LD/FB/GB rates are essentially the same as last year. Directionally, he’s also hitting about the same percentage of balls in play towards the left, center and right as last year. This could serve as evidence that the decline in BABIP has been nothing more than luck, and that there is no change in the controllable inputs. In terms of his improved K%, what jumps out is that his zone contact rate has improved by 5% so far this year from last year, contributing to a 2% improvement in overall contact rate. He’s seeing 4% fewer fastballs and 1-2% increases in offspeed stuff (sliders, curveballs and changeups). Additionally, he has seen 3% fewer pitches in the zone but been swinging overall at an identical rate. That data can probably be taken multiple ways, but I’d read it that he’s making better contact, despite swinging the same amount at an overall blend of seemingly tougher pitches to hit.

It seems clear that he’s showing improvement, which is to be expected for a 21 year old in his second full major league season. And simple aging curves foretell that there’s much more improvement to come. Using Tango aging curves (1919-1999 data) to get a sense of what Trout’s profile might look like at his peak, the signs are again very encouraging. I’ll use age 27 for a peak year (arbitrarily):

Where a 1.00 is peak for the category

  • Age 21: BB: 0.66, K: 1.32, HR: 0.68 and SB: 0.87
  • Age 27: BB: 0.88, K: 1.01, HR: 0.95 and SB: 0.88

I won’t actually project his numbers forward using these rates, as this is meant to be purely representative and I don’t care to get into debates about calculating correctly, but basically:

  • His walk rate should improve
  • His K rate should decline
  • His HR rate / power should increase
  • And his SB rate / speed should still be more or less the same

Mike Trout is already incredible, so maybe it’s not fair to compare him to the average player’s aging profile. And maybe it’s just not in our best interests to – I’m not sure my mind can handle the concept of a player as amazing as Trout getting that much better.