Re: Freight car Distribution


Aidrian Bridgeman-Sutton <smokeandsteam@...>
 

The question should be: Would the Gilbert and Nelson conclusions hold up under the weight of additional data, or not?<<
Well, until someone finds analyses and presents that information, any
answer is going to be purely speculative and probably not very
helpful.

I think Tim Gilbert would have been among the first to agree that this
model was not the final word, but that it was the best model he could
find that agreed with the data theat he had available No one else has
yet dealt with the subject in the same rigorous way.

A bigger sample might tell us some things that support the G-N
theories and some that do not - we are after all dealing with
statistics, which means that there will be plenty of real life
variations from the theoretical model. Until further records are
unearthed and the analysis is done, the model we have is likely to
remain the best available for some time.

However any larger sample needs to be found before it can be analysed.
If anyone can add meanigful data - that is to say conductors' books
from other places and other railroads, then that would certainly help
refine the model.

Is it worth reworking the numbers hased on one additional wheel report
from one train? Probaly not, but if anyone can add a meaningful amount
of new data the same methods can be applied and new conculsians drawn
or old ones strengthened

Regards

Marge Ynaverra
Personal freight car statistician to Mr Bridgeman-Sutton

Join main@RealSTMFC.groups.io to automatically receive all group messages.