Why we need computers calling balls and strikes

[paintmered]

Originally Posted by kaldaniels

Interesting plan paintme.

Would would you say to someone like Doug who (paraphrasing) says that pitch f/x is very very accurate already. I don't even know where to begin to question the issues (if there even are any) with pitch f/x.

I don't know what pitch f/x uses or if it's real-time. It may be completely viable already.

[Roy Tucker]

Great analysis, paint.

So how do you do the strike zone? I think that's the hard part.

[dougdirt]

Originally Posted by RedFanAlways1966

Is Pitch FX 100% accurate? It has been argued here, but does it change for every hitter based on his height and his stance (Rose being an extreme with his squat)? It is 2013 and I understand that people will use technology as they deem fit.

Is it 100% accurate, no. Is it 98% accurate? Probably so, especially once you have corrected data. Yes, the strikezone height changes batter to batter and even from at bat to at bat with each batter.

No MLB umpire is perfect, but most are damn good. That each man has a bit of a varying strikezone is OK by me and most teams (just be consistent).

This shows that it isn't consistent. The harder you throw, the smaller your strikezone is.

I want baseball types ruling baseball games. Not a Silicon Valley type that has never played the game nor understands the importance of the game and it's history. Wendelstedt can call the game I attend while Gates can be responsible for the software I utilize to study the stats from that game.

I want the game I love to be decided by the players on the field. Not the guy wearing the umpire hat.

[dougdirt]

Originally Posted by paintmered

I don't know what pitch f/x uses or if it's real-time. It may be completely viable already.

Pitch F/X uses three cameras, using standard 30 frames per second video, places at different points in the ballpark to triangulate the ball, batters and strikezone.

Google can help you out with more information if you are truly interested in reading up on it all.

[paintmered]

Originally Posted by dougdirt

Pitch F/X uses three cameras, using standard 30 frames per second video, places at different points in the ballpark to triangulate the ball, batters and strikezone.

Google can help you out with more information if you are truly interested in reading up on it all.

In that case, Pitch F/X will have a higher percentage of pitches measured incorrectly. That number may only be 1-2% more than something more accurate, but it will be higher. For it's current application, it doesn't matter if it's "not good enough". The inaccuracy applies equally to all umpires, and its output should be viewed only as a relative comparison between different umpire's ball/strike accuracy. I doubt it would be suitable to replace the role of an umpire. I'd at least need a bit more convincing.

For what it's worth, the Pitch F/X setup is much less complex and probably costs 20% of what I suggested above. It's diminishing returns, and it all depends on how accurate you need the measurement to be. An inch will cost you millions...

[Brutus]

I hate to beat a dead horse, but the computerized strike zones are dependent on humans calibrating the zone before each at-bat. There is some definite wiggle room for being off by an inch or two with each player, and worse is that because it's only calibrated before the plate appearance, a slight shift in batting stance can throw off the zone even further.

[kaldaniels]

So lets assume that 25% of Chapman's 98+ taken strikes are balls.

Is it too much to ask to see some photographic prove?

Lets take it a step further. If you are a fireballer getting jerked around why not make a huge stink in the press? 25 percent error rate is beyond comprehension.

[kaldaniels]

Originally Posted by dougdirt

Pitch F/X uses three cameras, using standard 30 frames per second video, places at different points in the ballpark to triangulate the ball, batters and strikezone.

Google can help you out with more information if you are truly interested in reading up on it all.

How far does a normal fastball travel in 1/30 of a second?

[dougdirt]

Originally Posted by kaldaniels

So lets assume that 25% of Chapman's 98+ taken strikes are balls.

Is it too much to ask to see some photographic prove?

Lets take it a step further. If you are a fireballer getting jerked around why not make a huge stink in the press? 25 percent error rate is beyond comprehension.

It is possible to do if you wanted to spend the money for MLB.tv (for archived game footage) and wanted to cross check PFX data. The problem is, you are talking about thousands and thousands of pitches here. That isn't something many people are going to be able to undertake in terms of going back and loading up the video, capturing it, editing it and comparing it against pitcher XYZ and his slower fastball. That doesn't even come into play with the fact that camera angles in every ballpark are different, so you aren't going to be able to do much with the video unless you are solely looking at it in one ballpark.

But it is pretty simple, if you accept that the Pitch F/X system is incredibly accurate (and MLB, MLB Teams and every television station that covers the games all do), then what the author is saying is true.

[paintmered]

Originally Posted by kaldaniels

How far does a normal fastball travel in 1/30 of a second?

A 95 mph pitch will travel 4.6 feet in 1/30 of a second.

[paintmered]

Originally Posted by kaldaniels

So lets assume that 25% of Chapman's 98+ taken strikes are balls.

Is it too much to ask to see some photographic prove?

Lets take it a step further. If you are a fireballer getting jerked around why not make a huge stink in the press? 25 percent error rate is beyond comprehension.

For a 20 pitch outing, it means he gets one fewer call than the 92-94 mph group.

[dougdirt]

Originally Posted by kaldaniels

How far does a normal fastball travel in 1/30 of a second?

Usually there are 20 frames (so 60 images in total between the three cameras) between release and the front of home plate for an average pitch.

[kaldaniels]

Originally Posted by paintmered

A 95 mph pitch will travel 4.6 feet in 1/30 of a second.

Ok, so isn't it a problem that that is much greater than the depth of home plate?

[dougdirt]

Originally Posted by kaldaniels

Ok, so isn't it a problem that that is much greater than the depth of home plate?

No, because we can use math/science to determine exactly how the ball moves based on gravity, velocity and spin of the baseball, all of which are measured by the system. Deceleration of the ball is accounted for, as is the spin and break of the baseball.

[kaldaniels]

Originally Posted by paintmered

For a 20 pitch outing, it means he gets one fewer call than the 92-94 mph group.

But couldn't just Chapman (or someone more articulate/media friendly (no disrespect intended to Aroldis)) hold a presser showing 100 pitches in the strike zone and 25 of them called balls? That would really turn the tide against umps I'd say...granting that the info is correct.