In last night’s Yankees-Angels game, Fox debuted a new device which apparently tracks the speed of a pitch from when it leaves the pitcher’s hand to when it gets to the batter. They claimed that Burnett’s 93 mph fast balls, when leaving his hand, slowed down to about 86 mph by the time they reached the batter. McCarver went on and on about this without really questioning the result.
How can a ball slow down that quickly over such short distance? If I drive my car 93 mph, and then take my foot off the gas, I’m certain that it doesn’t slow down by 7 mph in 60 feet. A baseball is pretty aerodynamic. My conclusion: they’re doing something wrong. Convince me otherwise.
The car has a lot more potential energy built up due to its mass.
The baseball is a lot SMALLER than your car. To draw a reliable comparison between the baseball and your car, you have to figure out how large a baseball is, compare that to 90 feet, then figure out how large your car is and figure out how many feet it would need to travel to get the same distance relative to size as the baseball did.
In your premise, the car is starting out already moving. The baseball is starting from zero. I don’t know if that makes a difference or not, but it’s something to think about.
Not gonna defend it exactly because I’m not smart enough, but I do know that a lighter object will carry less energy down range so comparing a baseball to a car is not a good comparison.
I obviously didn’t think through the car analogy. Please ignore that part. That wasn’t my proof, and I thought of it this morning. But I immediately questioned the numbers last night. I still think that a baseball can’t possibly slow down that much that quickly.
Given that that is only 7.5% of the initial velocity, I wouldn’t be that surprised. Baseballs aren’t really that aerodynamic, which is why they curve as much as they do when you throw them the right way. I’ll let someone with more understanding of the proper equations do the calculations on the subject.
Which is not to say that FOX has it completely right. I’d have to say that any assertion about velocity of the ball that has that much precision is probably not supportable without indicating a pretty generous (a couple mph either way) margin of error… :eek:
In the paintball world there is a lot of talk of switching from .68 caliber balls to .50. As such there has been a lot of math and numbers and simulations floating around.
A .68 caliber ball weighing 3 grams leaves the barrel at 300 fps or 204 miles per hour.
At 60 feet the ball is down to 150 fps or 102 miles per hour.
A baseball is much heavier then a paintball and will carry much more energy at 60 feet but it seems believable to me.
If I were designing a radar (or whatever) to be able to measure that, I’d have it collect the baseballs position and/or speed at several times in its approach, then run a fit through those points. You can then get the ball’s velocity at the pitcher’s hand and at the plate from the fit.
You’d know from the physics of the situation what form to expect the velocity WRT time to take, and before using it in your broadcasts, you’d do lots of measurements to verify that form empirically. You might only need two or three measurements to get the fit. Any more than the minimum would just help to reduce the noise in your measurements.
7 mph of slowdown? Don’t really know, but it doesn’t seem outrageous.
I can believe the slowing down. What IS BS is the attempt to show pitch location by a single point on the graphic. Often times the pitch will cut the corner and end up well outside, yet if you look at the graphic the pitch appears to be well outside.
There has to be a scale of drop off. Using the results can you compute when the ball comes to zero? Is this about the right distance the ball should move when thrown? IANAM.
I always wondered about this as well. Technically, a strike means that at least a portion of the ball must cross vertically over home plate (between the knees and armpits, although I think the latter has been revised). But is home plate a 2-dimensional plane or a 3-dimensional box? Home plate is really 5-sided, but I’m not sure if the front triangle is considered.
It seems that my assumption that the ball speed can’t possibly decline that much may be flawed. It still seems impossible to me, but I have no real reason. I wonder if a tall pitcher with a 12 to 6 arm angle has less of a speed reduction than a short side-armed thrower.
Those numbers are very believable. A sample situation from The Physics of Baseball by Robert K. Adair contains similar numbers from a section on wind resistance.
Yeah, I’d just take a high speed camera and measure the speed of the ball at every frame, compensate for any triangulation errors and test it against the radar. You could probably throw out any curvature of the baseball path as being fairly minimal, but if you were a real stickler, you could also position a camera above the pitcher and batter.
Let’s see. Cecil says a curveball can curve up to 18" side to side (Do curveballs really curve? - The Straight Dope)
There will be small variation, but the sideways curve is going to be fairly equal over the flight of the ball. We’ll say the ball travels 55 feet in flight (since the pitcher’s hand is well forward of the rubber when he lets go). And the plate is 17 inches front to back.
So the ball should curve, while passing over the plate, a max of 18 inches/55 feet * 17inches = 0.4 inches. That’s not really enough movement to ‘cut the corner’ enough to clearly be a strike, yet be ‘well outside’ at any other point before passing the plate. And that’s going from the front to the pointy rear, not the square part of the plate, which is half the size.
Now, if you figure the catcher is three feet farther back, that gives you another inch and half or so of movement, which does start reaching barely-noticeable-difference level. But if it’s a strike (side to side) over any part of the plate, it’s not going to be way outside at any other part.
To clarify the “single point” discussion - the strike zone is defined at the front of the plate. This what Pitch f/x and the other various pitch-tracking tools use when “locating” the pitch. They generally also measure the amount of break, both downward and left-right.
Developments in this area are rapidly advancing, and the science of analyzing pitcher performance based on this data is, in my opinion, going to be the next major step in baseball statistical analysis.
And if it’s on the front or back doesn’t make much difference - if you can throw a ball from the mound that is going to swing back around and clip off the end of the point and not touch any other part of the plate, you’re some sort of freak of nature.