This is a great piece of analysis, it was really interesting breaking down the phenomenon of a turning ball into its parts. I had a couple of questions though.
1. Why did you not include revs on the ball in the model? I get that in the fixed effects you werew mainly dealing with trajectories, but I would think that revs would be an effect we would want to separate from the ground. Is it because that would be subsumed under the bowler's identity?
2. You say in reference to other factors that affect turn: "Ball condition: the state of the ball (new or old). We do not have any data for this." Why not inlcude the over number or nth ball in the innings as a variable for this? I think that would be an okay way to measure age of the ball.
2. I could include the over number but that's sometimes an incorrect proxy as the new ball might not be taken at 80 overs and so on. Still, could be added
Ah sorry didn’t realise there’s no revs data. Also good point about new ball not always being taken after 80 overs, I completely forgot about that. Although a model restricted just to the first 80 overs of innings might be interesting, even just as a test.
This is a great piece of analysis, it was really interesting breaking down the phenomenon of a turning ball into its parts. I had a couple of questions though.
1. Why did you not include revs on the ball in the model? I get that in the fixed effects you werew mainly dealing with trajectories, but I would think that revs would be an effect we would want to separate from the ground. Is it because that would be subsumed under the bowler's identity?
2. You say in reference to other factors that affect turn: "Ball condition: the state of the ball (new or old). We do not have any data for this." Why not inlcude the over number or nth ball in the innings as a variable for this? I think that would be an okay way to measure age of the ball.
Thanks!
For 1. We don't have ball revs data at all
And
2. I could include the over number but that's sometimes an incorrect proxy as the new ball might not be taken at 80 overs and so on. Still, could be added
Ah sorry didn’t realise there’s no revs data. Also good point about new ball not always being taken after 80 overs, I completely forgot about that. Although a model restricted just to the first 80 overs of innings might be interesting, even just as a test.
It's too much!!
this ground wise data can be used to plan the batter response as per their strength and optimal shot selection