Written by the TreasureGuide for the exclusive use of the Treasure Beaches Report.
I just noticed something interesting in the Sedwick blog. Discussing the top selling coin from their May auction, they pointed out something that I never suspected. Here it is.
Strong bids for rare world and shipwreck coins surpassed $4.07 million in Daniel Frank Sedwick’s May 7, 8, & 10 Treasure Auction 29. This is a new record for the auction firm and an indicator of a robust market for collectible coins and currency.
The top selling coin in the sale was the single finest Mexico City-struck cob 8 reales Royal dated 1730 that realized $102,000 on a pre-sale estimate of $35,000 and up. A numismatic rarity, the coin has an overdate 1730/28/5 plus the king’s name and ordinal reworked with PHILIPPVS V over LVDOVICS I. It was graded by NGC as AU 58 which is rare among all Spanish colonial Royals as almost all known examples were holed and some even gilded shortly after minting...
That is something I hadn't heard before. Imagine cleaning a nice reale you just found and discovering that it is gilded. That would really throw me. I'd start wondering if it was a fake.
That is found in the May 26 post in the Sedwickcoins blog. Here is the link.
It is also interesting that the top selling coin was this amazing Royal rather than a gold coin.
---
I'm going to try to keep this as simple as I can, and it isn't easy. It is very useful, but there is a lot of math and technical language which has developed in the academic literature that reports the experimental findings.
In metal detecting you have two layers of signal processing. The metal detector processes the signals sit gets from the ground and attempts to screen out signals and other ground return to provide information about the nature of the target to the operator. The operator (detectorist) processes the information the detector provides to him and makes decisions about whether he wants to dig various targets or not.
Even though I'm going to characterize the decision as being binary (dig or no dig), the information that informs that decision is complex. It can consist of a variety of types of input and considerations, but includes the operator's expectations about what might be there and what strategy he wants to use as well as all the information he receives from the metal detector.
I previously talked about the situation as if the signal is either present or not (binary) when actually the signal can be complex and consist of various types of information. Not only might you have an audio signal, and that signal will have various characteristics that will include loudness, duration, and tone or pitch, but there can also be other information the detector provides, such as conductivity numbers or other target ID output.
Our brains automatically summarize a lot of information in making a decision without us being aware of each bit of input that goes into the analysis. Nonetheless, being aware of the factors involved with our own decision making can be very useful. Self-awareness and objective analysis helps us improve.
I've talked a lot about the two types of errors you can make when using discrimination. You can have false positives (digging junk) and false negatives (missing good targets). And I've been emphasizing the problem of not being aware of when you miss something good. What would you do differently if you learned that you barely missed the type of target you were always dreaming that you would find? For most people that would provide the motivation to change something.
Here are a few things you can easily do to help you learn what you might be missing. When in the field, just check yourself sometimes. Detect areas or targets that you aren't sure about or might even be negative about. Don't use discrimination unless you need to. You can always turn it on later if you learn from your initial results that you need it. That way you'll gain information and be more informed when you make the decision to discriminate. And maybe you won't have to discriminate so much. If you find, for example, that there is one or two kinds of predominant junk, you can reject or discriminate those out instead of discriminating more broadly.
Another thing is to do your homework. If you don't want to miss expensive watches, small gold rings, half-reales, or whatever, test both your detector and your techniques with those types of objects at home or in a test garden or at the beach. The more you test your detector and yourself, the better off you will be.
As you increase your discrimination, the probability increases that you will miss something good. The There are definitely times for discrimination, but it should be a well-thought out strategic decision unless you are simply out for a nice day at the beach and a little exercise.
Various sources of noise and limitations of the sensory and nervous system make perfect performance impossible. Here is a chart that shows hit rate (correctly identifying a good signal) and the false alarm rate (digging junk). If you ignore much of the unexplained detail and my sloppy notes, you'll notice perfect performance is near impossible and that as the hit rate increases, the false alarm rate decreases and vice versa.
 |
| From The Experimental Psychology of Sensory Behavior by John Corso (1967) |
My recent outings on the Indian River Ridge site and what I called the old house site illustrate this very well. On the Indian River Ridge site, I have virtually unlimited access and have decided to remove nearly every metal target that I can detect. On the old house site, by contrast, I had a very limited time but from past experience on the site, even though it had been years, knew there were likely some good targets on the site. I therefore chose to discriminate on the old house site. My decision rule was to only dig very promising signals, and I came away with an old coin and lead seal despite all the junk on the site and having a very short time to detect. My decision rule and the amount of discrimination I used was chosen based upon the circumstances, including the time available and my estimate of the likelihood of certain types of good targets. I considered the outing a success even though I'm sure I missed some good targets because of the increased amount of discrimination. That conclusion is supported by the fact that I previously detected the same site nine years earlier under similar circumstances and using a similar decision rule and missed the targets that I picked up on the recent outing.