Tony Lawrence Coins and Stuff: Why is MD (Machine doubling) worthless?

A recent thread at Collectors Universe asked "Is Machine Doubling ever worth a premium?"

Most of the answers were a resounding "No!" but there was at least one person who asked

whatever happened to "buy what you like"?

That's a fair question, I think, but let's look at why variety and error collectors are so disdainful of machine doubling. One response was

No premium applies to this type of error as it is strictly caused by the machine itself and not through true human error as it were.

But that's not entirely true. As explained at the "Mechanical Doubling" article at the Wexler Double Die site:

The loose coining press parts result from the continuous pounding of the presses as they strike coin after coin, and from wear on the coining press parts. The coining press operators have reported that when they see this doubling on the coins, all they have to do is grab their tools and tighten up all of the loose parts in the coining press. As soon as those loose parts are tightened, the doubling no longer occurs.

I'd say loose parts are human error - or neglect, perhaps, but that's close enough for me.

More likely the disrespect comes from how common this type of error is. Modern presses are very fast and strike very hard: keeping things tight is probably very difficult. That thought raised another question in my mind, though: is MD more common in modern coins than in older?

I did not get a good answer from the experts, but I would guess that it is. More speed and force would seem to be a good reason for parts to shake lose more often. Not that it never happens on older coins: this picture of an 1873 MD nickel is rather dramatic:

But that picture reminded me of a Two Cent piece that is in my collection. It too has a doubled date:

Doesn't that look like Machine Doubling to you? It certainly does to me, but apparently it's a repunched date (see the "1871/1871 2C VP-002" link under References). I asked how I could tell the difference if I were not aware of the VP-002 and was told

If you go look at the photo of the 1871 2 center at the [VP-002] link posted above, the notch at the lower right of the second 1 is very evident. If this coin were machine doubled, that notch would be filled in.

I don't see the "notch". I know what he's talking about because Wexler says the same thing at the link I already mentioned.

genuine doubled dies are characterized by a splitting of the serifs on letters with serifs, or a “notching” of the corners of the letters which are doubled. This splitting of the serifs or notching of the letter corners will not be found on coins with mechanical doubling.

I did some Googling on split serifs and found "Determining Die Doubling from Other Forms of Doubling" by J.T. Stanton, but that also failed to clear my confusion.

However, the the person who posted the picture of the 1873 nickel explained:

Think of machine doubling as a drawer sliding out from the device of the coin. The drawer must be full-width, because the slide is the same along the entire edge of the device. The notch shows us that this is not machine doubling, but (in the case of a 2-center) an RPD.

THAT helps a lot!

I also found "Collecting my Thoughts: Understanding Machine Doubling Damage" by Alan Herbert, which mentioned something that had bothered me about the "twisting" die theory:

From the very first, attempts were made to claim that doubling was caused by the die twisting at the moment of impact. A little thought will discount that theory, because you are dealing with a die which is meshing with the forming design on the coin under 25 or more tons per square inch of pressure. The amount of force needed to rotate the die under those conditions would run into the thousands of tons. If such a force were available, it would shove the entire design out of position. MDD always affects only part of the design, so this cannot be the cause. To prove my point, hold your hands together, with your fingers interlaced. Now, try and move one hand sideways without moving the other.

But wouldn't the same argument apply to "bouncing dies"? Well, it would unless the die had tipped slightly so that only one part hit the coin again. But Alan Herbert confused me more:

The strike ends with the "final impact of the die pair." Die bounce or chatter involves only one die. While you can find MDD on both sides of a coin, the cause is different on each side as there are at least three forms of MDD, each with a different cause.

He unfortunately neglected to explain the three causes. Fortunately I was able to find a page that does describe them - it's in the references. However, the "slide doubling" described there seems just as unlikely as "twist doubling", for the same reasons.

I remain somewhat confused by this whole subject.

References:

Is Machine Doubling ever worth a premium? It sure does look really cool!!!

Wexler's Double Die Site "Mechanical Doubling"

1871/1871 2C VP-002

Determining Die Doubling from Other Forms of Doubling

Collecting my Thoughts: Understanding Machine Doubling Damage

Three basic forms of machine doubling