What is the best source for average weights and measurements?

So I really like ANS's ONLINE COINS OF THE ROMAN EMPIRE (OCRE) site. http://numismatics.org/ocre/

It's helpful to get a fix on average measurements, as the screenshot shows. But sometimes the site has only one or two exemplars. What is the best resource for average weights and measurements for a given Roman coin type?

 

Log in or Sign up to hide this ad.

Forvm member coin galleries, ACsearch, & CNG are what I use.

 

Coryssa https://www.coryssa.org/index.php

It actually does the averaging for you as you click through the links. You can see in this example the average of over 5,000 folles for Diocletian. You could get even more detailed results by applying filters to type or mint

 

Coryssa. Wow. I was unaware of this resource. Thanks!

 

Wow. Thanks. On first glance this seems to be a significant source of free info that I had never seen before.

 

BMCRE has a lot of info in its introductory material. Take this, for example, from BMCRE4:

 

Fascinating - I had not seen this either. A pity Medieval (Europe) seems to have been cut - but still - an amazing effort!

Beware of course of using the average weight incautiously. Many damaged coins are included in some series.

Rob T

 

It occurs to me that - in principle - an auto bar chart generator might be tagged to this set up. That would make it far more informative regarding weight matters.

Am rather a fan of bar charts myself............

Rob T

 

I really dislike statistics manipulated or presented in innocent ignorance that gives an average number for a reign. There is noting educational about an average weight for a coin of Commodus, Septimius Severus or Gallienus (to name a few) that does not take into account dates within a long reign or the use of various mints. While nothing is perfect, the idea of a bar chart by mint and date would be interesting. Dates might need to be yearly or more exacting in some cases. Can we explain changes in standards by a ruler who reigned for a few months but changed during that time?

I view statistics as like photography. They both can be used to reveal truth or promote error according to the intent and skill level of the practitioner.

 

If bar charts are too diificult to program, a standard error of the mean would be very informative.

 

Yes I much agree. Actually I would go further - following a guy called Gorard - who over a number of papers came to the conclusion that teaching complex statistical approaches was in many disciplines doing more harm than good.

Here is one of his early papers, (attacking even the use of standard error)

https://www.leeds.ac.uk/educol/documents/00003759.htm

I am afraid I would have concerns about this. Not primarily the theoretical concerns about the technique itself (as Gorard voices), rather the fact that data trawled mechanically from web sites will contain all sorts of damaged coins which have no relevance to the intended issue weight. Most times a bar chart will throw up a peak specific to the well preserved coins, with a tail of junkier stuff behind it. That would be much more meaningful.

Rob T