After about two weeks’ work, I’ve got the first SALO 3 estimates.
- Only shot quantity ability is estimated so far.
- WAR values with MARKOV are not yet computed.
The estimates are much higher quality than any previous ones.
I can believe the estimates enough to start asking what they mean.
Now? I can confidently say I don’t believe this model I’ve crafted.
Discussion to come
Expect forthcoming posts about:
- Estimation obstacles and why they’ve taken till now to surmount
Possible multiple posterior modes and implications for samplingUPDATE: nope!
- The genuinely weird MARTIN-QUINN projections
Expect all these to contain a lot more questions than answers!