Re: [TML] Currency Phil Pugliese (12 Jul 2016 19:03 UTC)
Re: [TML] Currency Bruce Johnson (12 Jul 2016 19:37 UTC)
Re: [TML] Currency Michael Houghton (12 Jul 2016 21:17 UTC)
Re: [TML] Currency Peter Vernon (12 Jul 2016 21:27 UTC)
Re: [TML] Currency David Shaw (12 Jul 2016 20:52 UTC)
Re: [TML] Currency Abu Dhabi (12 Jul 2016 20:56 UTC)
Re: [TML] Currency Bruce Johnson (12 Jul 2016 21:45 UTC)
Re: [TML] Currency Tim (13 Jul 2016 01:32 UTC)

Re: [TML] Currency Tim 13 Jul 2016 01:32 UTC

On Tue, Jul 12, 2016 at 09:52:26PM +0100, David Shaw wrote:
> OK, but how many of them are good enough to fool a US citizen? They might
> fool me because, as a UK citizen who has never visited the USA, I have
> never seen a real $100 bill except in quick flashes on the TV but would
> they be accepted in a US store?

My expectation would be that a lot of them would be accepted in many
stores, but not in the banks to which they regularly deliver their
takings.  If there's just one or two every now and then, they're
likely to be just written off and reported in the regular statistics.

If there's an increase in the rate, then things start happening.  Law
enforcement gets notified, it probably gets into the news one way or
another, probably including specific flaws to watch for in the notes.
For extremely good counterfeits, they may only be detected by
collisions with the serial numbers of real notes, and that's likely to
trigger a much larger response.

In reality, counterfeits are not that good, because real notes are
already expensive to make even with the economies of scale of a
national mint.  The counterfeiter has the additional expense of having
to match someone else's manufacturing quirks.  E.g. all notes with
serial numbers starting with Q5 might have had the front and back
plates misaligned by 0.02-0.04 degrees clockwise and offset by
0.08-0.12 mm in a particular range of directions -- well within the
specified tolerances, but easily recognizable if you know what to look
for and near impossible to arrange deliberately.  I don't know if
these techniques are used in real currency, but it seems likely.  The
mints know what they produce, and variations can definitely be
detected.

Even up to Traveller technology and a fair way beyond, it seems likely
that manufacturing tolerances will lag behind sensor technology.
Unavoidable variations in the production process can be detected and
serve as "fingerprints" for batches of notes, or even for individual
notes.  If a bunch more notes than usual arrive in banks not matching
their registered fingerprints, then it's likely that many of them are
fakes.

This gap probably only ends with manufacturing processes where every
atom can be placed in exactly the desired configuration -- and cheaply.
Traveller is nowhere near that level, so it seems likely that
counterfeit notes will always be at least statistically detectable.

- Tim