I have analysed over 4200 Price coin types of Alexander The Great from the Pella database and made some graphs. This is what I've found :)

[AncientCoinnoisseur]

So, this was a quite old project, but interesting nonetheless! My method of scraping the information was not very efficient, since I am not that good with programming, but without further ado, let's go! Note: the Pella database is far from perfect, so many measures are probably the result of mistakes in the data itself, so this project is simply: "this is what you will find in the database, errors included".

MAX VS. MIN DIAMETERS:

Here the diameters of the different denominations are to scale. The black bars help visualise the difference between the max and min diameter of each denomination!

 

COMPARING THE DIAMETER DISTRIBUTIONS OF THE VARIOUS DENOMINATIONS:

Here I put the diameters in decreasing order of their average and tried to show both a box plot and a scatter plot to see their distribution for each denomination:

 

COMPARING THE WEIGHT DISTRIBUTIONS OF THE VARIOUS DENOMINATIONS:

Here I did the same thing but with the weights. Since the dekadrachms tend to skew things a lot, the left graphs include them, while the right ones do not (click to zoom the picture if it's difficult to read):

 

2 HEAVIEST, 2 LIGHTEST, 2 LARGEST AND 2 SMALLEST FOR EACH DENOMINATION:

In these 4 pictures I took each denomination and put, from left to right, top to bottom: the two heaviest, the 2 lightest, the 2 with the biggest diameter and the 2 with the smallest diameter. Note: the diameters are not to scale in these 4 pictures, for readability purposes!

 

DIAMETER VS. WEIGHT FOR EACH DENOMINATION:

It is quite interesting to see how each denomination puts itself into its own space in a 2d scatter plot! There are some outliers, probably caused by rounding errors or other reasons. It would be nice to have clean data!

 

DIAMETER VS. WEIGHT FOR THE MOST COMMON DENOMINATIONS:

Here I put the 6 denominations with most entries, each with their own diameter vs. weight plot! 

 

SOME FINAL STATISTICS:

Here I put some additional info you might find interesting!

 

I hope you liked it!

Ancient Coinnoisseur

 

[Restitutor]

This is so freaking cool, thank you for sharing it! The size distributions are just wild, especially the Drachm. I am not well versed in Greek coinage but curious if there are any known reasons/educated guesses for why the sizes of so many coins in the same category would so radically differ? Makes me wonder too if the size/weight bands tighten depending on coin mint, or if we would see the same wide bands even at the same mint.

[AncientCoinnoisseur]

@Restitutor thank you so much! Glad you liked it! I am sure the database has several mistakes that should be corrected, but some differences are there nonetheless. I am curious too, would be fun to repeat the study separating coins based on the mint!

[Limes]

I can only echo the words of @Restitutor. Amazing job, and fantasticly displayed! Im also not really into this coinage, but every coincollector should to applaud your effort I think!  

[AncientJoe]

Impressive work and a great way to visualize the data!

[AncientCoinnoisseur]

@Limes thank you!

@AncientJoe Thanks! I’m a big fan of data visualisation, so I always try to present the information in a way that makes it enjoyable (I do the same with my infographics on coins!).

 

If you have any ideas about another project like this I’m all ears 🙂 I love statistics!

[lim]

It would be interesting to analyze the deviation from the norm of weight and diameter, only of lifetime coins of Alexander the Great. For example, coins with the lion of the city of Babylon. And so go through all the mints.

Edited November 27, 2024 by lim

[wuntbedruv]

Interesting stuff.

Is it possible that some of the massive variances could be caused by data entry error at the Pella end? E.g could some of the 'drachms' at the lower end of the diameter and weight scales actually just be misattributed hemidrachms (perhaps incorrect drop-down option selection) or alternatively wrongly entering the weights/diameters?

A good dataset is regularly data-cleaned!

[I_v_a_n]

3 hours ago, wuntbedruv said:

Is it possible that some of the massive variances could be caused by data entry error at the Pella end?

Yes, for sure. There are a lot of obvious mistakes taken from Pella database. Unfortunately, it must be manually verified to become a reliable.... a huge volume of work.....

[Brennos]

Excellent work!
I think it would be very useful to the people who manage the database so that they can correct the errors that seem to be quite numerous: it's impossible for a tetradrachm to weigh 0.66 g or 1.97 g.

Edited November 27, 2024 by Brennos