We have started that list. Both exposed-length groups contain more than one overall length. The person measuring used the front of the jaw as their reference.

Did everyone use that same reference face and the same jaw arrangement? A mixed reference could create groups before seating variation enters the picture.

Check the end against the stop too; a nice visible end does not guarantee the hidden one is sound.

Two people measured. One used the jaw face, the other the chuck face. That explains the large split. We have not merged the readings into a seating allowance.

Good. Repeat with one agreed reference.

And repeat the same blank, Elliot Chan, so a new list of different blanks does not leave the original seating question untouched.

Setter agreed the reference and repeated the identified blanks. Most reconcile with their overall lengths. One does not repeat; inspection found damage at its stop end and excluded it from the trial set.

Where is the damaged blank now? Keep its identity and reason available rather than let excluding it erase the incoming issue.

Retained separately by quality, with the damaged-end finding. The corrected sheet is with the loading designer. It explains these samples but is not approval for automatic loading.