West Coast Flooded With Denver Coins, East Coast With Philly

Here in Tucson almost all coins encountered in change are "D" coins. Those that are "P" are 20 years old or so, though once in a great while a new one shows up.

"S" cents fomr 68-74 are found here on occasion.

dcunitedfan, I went through my quarter can. It represents about a year of circulation about 60 miles north of Denver. Out of 692 quarters, 530 are D (77%) and 162 are P (23%). All were accumulated within 5 miles of my house - grocery, liquor stores, coffee shops, etc.

I then split up the Ps by decade:
60s 25
70s 17
80s 30
90s 40
00s 50

I expected that the vast majority of coins would be D, since the mint is so close. However, the decade distribution of Ps seems to indicate that it doesn't take that long for coins to travel - there are more from recent decades. Oddly, zero P mints from the 2010s though. I was surprised by the number of 1960s still kicking around, mostly 65 and 66. I was also surprised by the number of bicentennial quarters still out there - 7 of them (3 D, 4 P). The latter are the two smaller separated piles in the first photo.

I was too lazy to split up the Ds by decade, but I figure it's probably a pretty even distribution correlated with mintage. I did notice as I separated them that a 1990s coin had about an even chance of being P, but that's non-scientific. The 1990s saw a large influx of people to Colorado (which has continued since). And of course tourist traffic is high as well - the local coffee shop has a shelf lined with coins from all over the world.

Separated by mint:


P mint by decade, 60s at bottom:

Us folks in Minnesota get more Canadian coins. How do you figure that?

dcunitedfan, 80525 is close enough. I'm not sure what "coin-level" data is going to tell you. Can you expand on what you think you might prove with such a detailed analysis?

Well, you've got my results at least. Zero P quarters in a sample of 692 for the past 7 mint years. I think your results will be all over the map (literally).

This may help illustrate why I'm looking for a list of the exact year/mint for each coin, along with where and when the coins were found. I've already seen in my own collection that D coins don't show up very often at all until a few years have passed since the mint date (eg for 2016, going back in time I start seeing them in greater numbers 5 years back, or 2011). kbbpll, you've seen similar results from the D side of the map. That leads me to the hypothesis that it takes maybe 5 years or so for coins show travel throughout the US in sufficient numbers to show up in greater than 5% of the results for their year.

To prove this out, I need real data that I can count and graph. Grouping coins by decade might be workable for this one specific hypothesis, but it probably won't be sufficient if I come up with other ideas to test. Also I'd like to do a series of maps showing the find patterns for same year, previous year, n - 2 years, n - 3, etc. That can't be done properly without coin level data.

Which is why I'm looking for an accurate census of whatever portion of coins that people feel like taking. Say you are a CRHer and get a box of nickels. I don't necessarily need a spreadsheet with all 2000 - if you only feel like doing a spreadsheet for a couple of rolls, that's still something. More is better of course from the standpoint of sampling accuracy, but the more can come from more people in the same area or from a single person doing more coins, it should make no real difference either way. And even small samples will be of great value to me if they come from different states, since I'm starting from no data at all in most of the US right now.

It's probably abstract sounding without some examples of the graphs I'm hoping to come up with, but right now I have so little geographic coverage that I can't begin to generate a map from real data. If you google "choropleth map" you'll see what I'm aiming to produce. I think it's going to be sort of a slow start, chicken-and-egg thing where the map I would produce now would have almost no geographic variation and so be pretty uninformative, but once I can get my hands on a few samples from across the country, it will start showing the differences I'm talking about, and then I hope as I can publish maps of greater detail, more people will be inspired to contribute some data, and the map will improve further.

Funny but I was just thinking earlier today that the same thing that the most likely "battleground" between P and D will be either the Mississippi bordering states, or perhaps somewhere in the midwestern states (Mark1959's observation started making me wonder if the Appalachians are the true divider)

The S pattern will be interesting too, but now that so much time has passed since they made business strikes, a much more distant echo.

Re sample sizes - got to start somewhere. The larger the better of course.