Bowling speeds - cricket

Manxman_in_exile

Probably not the right board, but as I got a very satisfactory response last month to my thread as to how camera tracking worked on space launches, I thought I'd ask here.


How do they measure bowling speeds at cricket?


I assume they use some sort of radar or laser speed gun, but I don't see how it can be accurate. To give an accurate speed the ball would have (I assume) to be travelling directly at the measuring device, but this can never happen. (Unless the delivery is an illegal beamer - and even then it's going to be following some sort of parabolic path).


The ball is always going to be travelling slightly obliquely to the measuring device and so is always travelling further than the device "thinks" (more so the more short pitched the ball is as opposed to a Yorker, say). In motoring speeding cases I believe this is referred to as "slippage" or the cosine effect.


Just to add: (1) I'm well aware of earlier attempts to measure the speed of cricket deliveries. The University of Western Australia, about 40 years ago, used high speed photography from the side with the bowlers bowling against a grid layout so they could accurately measure how far a ball travelled in a specified time. (2) Amongst attempts by many others, The Sunday Times (again about 40 years ago), tried to use a time lapse calculation using slow-mo captured from TV coverage. Obviously this, if successful, would give an average speed rather than an instantaneous one, and was also subject to considerable error.


I've also tried googling it but nothing I've found (so far) seems to address this.


I've also listened to discussion on Test Match Special but (despite the fact I love the programme) they are hopeless on anything touching upon technical issues. (I remember a particularly highly regarded member of the commentary team posing the question, after a four had been scored: "I wonder when a ball hit to the boundary is travelling at its quickest? Does it speed up, do you think?" )


So, does anybody know?


(PS - just struck me, maybe they use HawkEye as opposed to a speed gun?)

Heedtheadvice

You presume that radar only measures in one direction?
There are three axis radars and doppler radars that with a bit of calculation can measure velocity (not just speed) very accurately.


Not saying that is used but results would (and probably do) fill in the boring bits between the short bursts of action much like throwing the ball to fellow team members to keep them awake ! :rotfl::)

EssexExile

I would imagine that with cameras all around the ground it would be easy to know exactly where the ball is at any moment, and therefore easy to calculate its speed.

RumRat

Easy when you know how.....
https://www.quora.com/In-cricket-matches-how-do-they-measure-the-speed-of-the-delivery

Manxman_in_exile

Heedtheadvice wrote: »

You presume that radar only measures in one direction?
There are three axis radars and doppler radars that with a bit of calculation can measure velocity (not just speed) very accurately.


Not saying that is used but results would (and probably do) fill in the boring bits between the short bursts of action much like throwing the ball to fellow team members to keep them awake ! :rotfl::)

Thanks - I wasn't aware of that. I doubt that's what they use in cricket, though.

Manxman_in_exile

EssexExile wrote: »

I would imagine that with cameras all around the ground it would be easy to know exactly where the ball is at any moment, and therefore easy to calculate its speed.

Yes. That's why I mentioned HawkEye at the end of my OP. But I don't think they use HawkEye to measure delivery speed - otherwise I don't think they'd continually be questioning the accuracy of the "speed gun" on TMS. (I could email TMS but I doubt they'd understand the question!).


Having said that, there has been quite a lot of questioning re the "predictive" accuracy of HawkEye in cricket decisions and as to how accurate it is in line calls in tennis. (eg https://web.archive.org/web/20111003222351/http://www2.geog.ucl.ac.uk/~mdisney/teaching/1006/papers/collins_hawkeye.pdf)


However, HawkEye is generally accepted by everyone in cricket and tennis, although when watching Wimbledon I'm quite sceptical about the alleged accuracy of some of the very marginal line calls.

Manxman_in_exile

RumRat wrote: »

Easy when you know how.....
https://www.quora.com/In-cricket-matches-how-do-they-measure-the-speed-of-the-delivery

My laptop doesn't seem to like that link - but if it's one of the ones I'm thinking of it's not very useful. I'll try again later.


EDIT: It is one I've read before and I don't think it's very reliable (although that impression may partly be due to English not being the author's first language).


I don't think HawkEye is used, as the link claims - but I may be wrong.


PS - it is interesting what mis-apprehensions sports fans often hold. I was listening to a discussion about VAR the other day and goal line technology was raised by a contributor. They thought the football had a chip inside and had to be corrected by a match referee who said: "Er...no, there isn't."

Manxman_in_exile

Heedtheadvice wrote: »

You presume that radar only measures in one direction?
There are three axis radars and doppler radars that with a bit of calculation can measure velocity (not just speed) very accurately.

So when various motoring forums (like Pepipoo and here) talk about the cosine effect and slippage when a vehicle is travelling obliquely to the measuring device (thus underestimating the speed) does this effect not exist, or is it the case that the police don't have sufficiently sophisticated equipment to overcome it?


I can't believe what they use in cricket is sophisticated!

Heedtheadvice

Oh those effects certainly do exist!


For a radar (or similar single axis measurement systems) they measure speed along that axis.
Crossing at 90degrees to that axis means there is no change along the axis therefore no speed is measured along the axis.


Progressively as the object changes from 90degrees to 0 degrees the speed along the axis changes from zero to 100%. It follows the cosine law. Cosine of the angle is 1 at zero degrees and 0 at 90degrees (but is not a linear change with angle and small angular differences have very small measurent errors: even 25degrees off axis only results in a 10% error !!) (180 degrees is also 1 and 270 degrees also zero).


For a measurement only along a single axis then both the other two axis angles come into play with cosine laws

Manxman_in_exile

Heedtheadvice wrote: »

...


For a radar (or similar single axis measurement systems) they measure speed along that axis.
Crossing at 90degrees to that axis means there is no change along the axis therefore no speed is measured along the axis.


Progressively as the object changes from 90degrees to 0 degrees the speed along the axis changes from zero to 100%...

I suppose that is one of the points I was questioning in the OP. Because the ball is always going to be travelling at an angle > 0 degrees to the axis of measurement, the measuring device is always going to underestimate its true speed.