Modeling Aerial Pursuit Curves

[crontab]

I originally tried to send this to Oleg personally, but his mailbox is full.

Is there a particular reference you would recommend regarding the mathematics and algorithms involved in modeling aerial pursuit curves? I'm interested in possibly applying these to the area of baseball fielding.

I imagine it's something like multi-state, e.g. "point and fly", "anticipate", "anticipation with crude adjustment increments". I hope this makes sense, I do not know this area at all.

-Chris Long (San Diego Padres)

[AndyJWest]

A bit of a long shot, but the general algorithmic solution to such problems (not involving actually modelling the underlying physics) involves a PID controller: http://en.wikipedia.org/wiki/PID_controller
Whether this will get you anywhere, I've no idea, but from your description, you may be trying to do something similar.

[Splitter]

Wow, interesting concept if I am thinking about it the right way. Sabermetrics are fascinating.

Are your results something that we will be able to read or will it be kept internal to the organization?

Splitter

[Splitter]

http://www.hsu.edu/uploadedFiles/24/lloyd.pdf

But you probably have that already. Maybe something in the bibliography or source section helps?

Not an equation, but a neat article I ran into a while ago (old): http://www.jcs-group.com/military/wa...f1/gunner.html

Splitter

[crontab]

This is related to the new Field/FX data, and it's unclear how to best evaluate it, but pursuit curves seem to me to be a logical area to investigate. The job facing a fighter pilot, of course, is much more complex than that facing a center fielder, but there are similarities.

[major_setback]

Originally Posted by crontab

I originally tried to send this to Oleg personally, but his mailbox is full.

Is there a particular reference you would recommend regarding the mathematics and algorithms involved in modeling aerial pursuit curves? I'm interested in possibly applying these to the area of baseball fielding.

I imagine it's something like multi-state, e.g. "point and fly", "anticipate", "anticipation with crude adjustment increments". I hope this makes sense, I do not know this area at all.

-Chris Long (San Diego Padres)

No wonder his mailbox gets full!!
Dear dear dear

[crontab]

Originally Posted by major_setback

No wonder his mailbox gets full!!
Dear dear dear

I'm a fan of Sturmovik, too, of course! Wonderful simulator.

-Chris

[Igo kyu]

I think the origingal poster means pursuit curves, particularly lead pursuit, lag pursuit and pure pursuit. I don't know about the maths, but theory is going to be simpler than real life anyway, so the maths probably doesn't tell you everything.

[Splitter]

Major,

What the OP does is try to quantify athletic performance (baseball) with measurable data. He has influence over decisions that range into the hundreds of millions of dollars each year. Clubs are trying to analyze players every day so they make the right draft and trade decisions.

Baseball has always been about statistics in its' way, but only fairly recently have clubs and fans started to dig deeper into stats and even come up with many new stats. What he does is exceptionally complicated and I understand exactly 2% of the mathematical equations that I get to see on the subject .

What he does is part of how decisions in baseball get made. And since I had actually heard of him before (not common for front office types to us fans), he is apparently pretty good at it lol. The organization he works for is probably worth half a billion dollars.

Pilots and baseball players actually have a number of similarities in their make up and what they do, so his request makes sense in its' way.

Splitter