The Great Coin Weight Experiment

[MunkyMan95]

I conducted my own little experiment with some of my PCGS slabbed Half Eagles.

 

Hypothesis: The more worn a coin is, the less it will weigh (versus its MS counterpart).

 

The Experiment: First, I gathered 3 different PCGS-slabbed Half Eagles. I then proceeded to put them on my digital scale (accurate to 0.1g) and weighed them. I used one VF35 Half Eagle, one AU55 Half Eagle, and a FR02 Half Eagle. The results were surprising.

 

Results:

AU55: 39.8g

VF35: 39.4g

FR02: 38.8g

 

Now, a note on the FR02 coin. This coin was a Classic Head Half Eagle, rather than a Liberty Head half eagle like the other two coins were. There is 0.001g of a difference in weight between the two (in MS condition), but there is also a 0.9mm difference in diamater, accounting for less plastic weight in the PCGS slab. I am also willing to accept that there will be inconsistencies within the PCGS slabs, errors in the weighing by the scale, yadda yadda yadda. Still, I would say that the coin actually has lost about half a gram of weight since it was in Mint State condition.

 

I would also say that from MS to VF, this particular half eagle lost 0.3g or so (again, inconsistencies in the plastic); and since AU was our control, I cannot speculate on the weight of that particular coin.

 

Can anyone prove or disprove my hypothesis? Has anyone else conducted a similar experiment to this one? What were the results?

 

I await your feedback!

[pendragon1998]

Your hypotheses are:

 

H0 - The mass of a gold eagle remains constant, regardless of wear.

(null hypothesis, what you're trying to disprove)

 

Ha - There is a reduction of mass with increasing wear.

(alternative hypothesis, an explanation which cannot be true if the null hyp. is true).

 

 

 

Thoughts on your "experiment":

 

1.

I've got some questions regarding the original mass of an "uncirculated" coin. Was there an acceptable tolerance for the coins as they were minted? How strictly was the tolerance actually enforced.

 

2.

I would be most comfortable if you controlled for date and mint of origin. Is a 1840 coin from Philly really equivalent to one from 1900 San Fran.?

 

3.

I don't like that you weighed slabbed coins. There may be a variation in the amount of plastic from slab to slab.

 

4.

I really don't like the small sample size. If I were doing this, and I had access to enough coins, I'd want at least 30 coins per grade so that I could compare mean masses for each given grade. There may be individual variation that is misleading.

 

5.

I'm not sure how you can determine a 0.001 g difference in mass between CHHE and LHHEs. Is that based on the scale you mentioned earlier, which according to you is "accurate to 0.1 g"? I don't know anything about your scale, but in my field work, we've encountered (fairly expensive) scales which are anything but accurate.

 

6.

Once you collect enough data, you'll want to perform a statistical test to see if there's really a significant difference in your means. Only then will your conclusion have any real merit

 

 

Don't take these criticisms too personally. I think that you have a laudable research idea. You just need to (1) refine your experimental approach, and (2) make certain that your results are statistically significant. Kudos on taking some research initiative.

 

Oh, by the way, you can never prove a hypothesis - you can only either disprove it or fail to reject it.

 

I might add that you might be on to something with your hypothesis that mass is being lost with wear. I'm only trying to show you that you can't draw the conclusion that it is without refining your project.

[bsshog40]

The only testing I've done is with a scale I bought off ebay. It seems to work pretty good but there was always at least 1-2 tenths of a gram difference.

[carson city]

At 800oz. gold the loss of one gram equal to about $27.00. Something to consider when buying junk gold or silver !

[MunkyMan95]

Michael,

 

This was not meant to be a super-controlled lab test with coins weighing completely the same down to the microgram. This was just a simple little experiment that took me all of 10 minutes to complete, and was at best a rough estimate of the weights of the coins.

 

1) The US Mint, as far as I'm aware, was very precise with their mesurements of gold planchets.

2) I did the best with what I had with me.

3) I mentioned that.

4) Again, I did the best with what I had with me.

5) I used the RedBook. According to them, the weight of a CHHE is 8.359g. The LHHE is 8.36g. 0.001g of a difference.

6) Again, not a super-scienteific controlled experiment.

 

Carson City, that was what I was trying to represent with this experiment. I'm glad someone caught on to it.

 

Bobby, SHADOW! (just kidding )

[pendragon1998]

This little diagram I whipped up illustrates the difference between accuracy and precision. When selecting a scale, both are important.

 

(The center crosshair indicates the true measurement)

[pendragon1998]

Michael,

 

...

 

6) Again, not a super-scienteific controlled experiment.

 

Carson City, that was what I was trying to represent with this experiment. I'm glad someone caught on to it.

 

That's fine - I wasn't trying to crack on you at all. I was just offering some thoughts on how the experiment could be refined to get some really conclusive results.

 

I think that you're roughly correct, though.

[MunkyMan95]

Michael,

 

...

 

6) Again, not a super-scienteific controlled experiment.

 

Carson City, that was what I was trying to represent with this experiment. I'm glad someone caught on to it.

 

That's fine - I wasn't trying to crack on you at all. I was just offering some thoughts on how the experiment could be refined to get some really conclusive results.

 

I think that you're roughly correct, though.

 

Michael, thanks for the input. I would LOVE to take this experiment to the large scale and get maybe 100 coins in varing grades. I want one of those scales that the US Mint used on the 'Secrets of the US Mint' show on the Travel Channel. You know, the one that was so sensitive that ordinary dust would throw it off? Roughly, I think that my results make sense, and one should always weigh the items they are considering purchasing at bullion value.

[Revenant]

This little diagram I whipped up illustrates the difference between accuracy and precision. When selecting a scale, both are important.

 

(The center crosshair indicates the true measurement)

 

That little diagram is in every freshman and most sophomore/junior engineering textbooks.

[pendragon1998]

Roughly, I think that my results make sense, and one should always weigh the items they are considering purchasing at bullion value.

 

Good grief - don't let stinkycheese hear that one!

[pendragon1998]

That little diagram is in every freshman and most sophomore/junior engineering textbooks.

 

Didn't intend to imply that I invented it, if that's what you're getting at. It's pretty common.

[rbrown4]

Roughly, I think that my results make sense, and one should always weigh the items they are considering purchasing at bullion value.

 

Good grief - don't let stinkycheese hear that one!

 

[dooly]

Guess it's boils down to how much glue they used

[MikeInFL]

Why not just grab a bunch of, say, copper Lincoln Memorial cents in various grades out of circulation, grade them, then weigh them. Provided you are accurate in your grading and measurements, you should see a trend in their weights if your theory is correct. Not only will this provide for more data to support your hypothesis -- it wouldn't cost much either.

 

That said, the one thing I would be sure an investigate further is the weight variance in the planchets -- as this could overcome any wear-induced weight loss you might measure.

 

Have fun...Mike

[Arizonadesertrat]

Anyone interested in silver dollars, particularly Morgans, should pick up a copy of John Highfill's book "The Comprehensive U. S. SIlver Dollar Encyclopedia." In it, several dealers state discuss their experiences with mint bags of silver dollars. There are also comments to the effect that, after picking up dozens of bags of dollars and qualitatively noting (i.e., estimating without actually weighing) weight differences, people could accurately guess which ones had circulated coins and which ones had MS coins.

[stinkycheese1]

Roughly, I think that my results make sense, and one should always weigh the items they are considering purchasing at bullion value.

 

Good grief - don't let stinkycheese hear that one!

 

 

Hey hey now, stop picking on me!

[pendragon1998]

Roughly, I think that my results make sense, and one should always weigh the items they are considering purchasing at bullion value.

 

Good grief - don't let stinkycheese hear that one!

 

 

Hey hey now, stop picking on me!

 

[Conder101]

Weimar White did a similar experiment years ago using mercury dimes in an attempt to see if grade could be determined by weight. He determined that it couldn't be done because of the tolerance allowance.

 

Now the tolerance on gold is much tighter than on the silver coinage so it makes it more plausible. The tolerance on a Liberty head half eagle from 1837 - 1849 was +/-.016 grams, 1849 to 1873 was +/- .032 grams, and 1874 to 1929 was +/- .016 grams. These tolerances are well outside the sensitivity of your scale so it can't tell the difference between a heavy or a light coin.

 

We also have no figures on the weight tolerance range of the PCGS shells and I believe that range will be so great as to swamp any weight change in the coin due to wear. It could easily account for a significant percentage of the "weight loss" of the low grade half eagle.

 

I think this could be an interesting study, but I think you would have to have a much larger sample size and you would need to remove the variable of the slab shell weight. So you need to be weighing raw coins. If you can't use raw coins then you could try and establish a reference weight and tolerance range for the slab shells by weighing a large number of cracked out shells from half eagle slabs.