1/16/26 Report - Decision Strategies for Metal Detecting and Treasure Hunting: Research, Theory and Application.

 

Written by the Treasureguide for the exclusive use of the Treasure Beaches Report.

Fort Pierce Beach This Morning from Fort Pierce Jetty Beach Cam.

Two days ago, I talked a bit about signal detection theory as it applies to metal detecting target ID and discrimination.  Today I'll begin a discussion of decision strategies for detectorists.  

Different detectorists use different strategies.  Detectorists usually select strategies instinctively or by  feel rather than using a detailed formal process.  Detecting strategies usually aren't defined in specific or detailed terms.  Each decision strategy is based upon assumptions, and each strategy reflects characteristics of the detectorist.  

My thinking on this subject is rooted in academic theory and research, but I'll try to make it as simple and understandable as I can, while at the same time reflecting some of the academic authority behind it.  

Some treasure hunters will pursue the largest possible treasures.  These treasure hunters might pick a huge cache or sunken treasure ship that hasn't been found and spare no expense, time or effort in the search for that treasure.  You might think of the search for the Atocha.as an example.  That takes a certain kind of person.  Some people might call those people dreamers.  The probability of success on a search like that might be very small, but that does not deter those people who are driven to go after the maximum possible outcome. 

On the other hand, there are people who are more risk averse and try to minimize the possibility of worst things that could happen.  They are more concerned about possible negative outcomes. They might not want to invest a lot in the search or take chances that might leave them broke in the end.  They will seek a more modest but attainable treasure. 

One important point here is that treasure hunters or detectorists are different, and those differences are reflected in their decision making and behavior.  In the terms of the academic research, these differences might be referred to as operator characteristics.  If you want to check out the academic body of knowledge on the subject you can use the terms Minimax and Maximin, which are the names two commonly employed strategies that relate to the examples I gave above.   ***

In academic terms, Minimax (sometimes Minmax, MM or saddle point) refers to a decision rule used in artificial intelligence, decision theory, combinatorial game theory, statistics, and philosophy for minimizing the possible loss for a worst case (maximum loss).

A decision strategy that I recommend for detectorists is maximizing the expected value.  I'm going to give you the math on a simplified example just to illustrate the basics, but you don't really have to do the math to use the decision strategy.

I'll start with a very simple example.  Suppose there is a coin flip game in which you are rewarded five dollars for every heads flip and lose three dollars for every tails. Below are the steps for applying the strategy, which are also the basic steps you'd use for real-life applications.

1. Identify all possible outcomes of interest.  

2. Determine the probability of each outcome.
3. Multiply the value of each outcome by its probability.
4. Add all the products to get the expected value.

In the coin flip game where you win $5 if the coin lands on heads and lose $3 if it lands on tails, the expected value is calculated as follows: Expected Value = (5 x 0.5) + (-3 x 0.5) = $1.

If you attempt to use the strategy of maximizing expected outcomes in metal detecting, you do the same thing.  You identify the possible outcomes and the probability of each outcome and the value of each outcome.

I'm not saying to actually do the math but follow the logic. What are the outcomes. 

When I began detecting I actually kept a record of all my coin finds, including the total number and type or value of coin.  If I remember correctly, the average value of modern U.S. coins finds was about 7.5 cents.  That, of course, was for the beaches I was detecting.  Things could be different today and could be different for different beaches. 

Lets make the situation more complex and more realistic.  Not only might you find coins, but you might be hunting jewelry and dealing with junk.

I talked about trade-offs yesterday.  You can look at dug junk as having a small negative value.  On the other hand, good jewelry will have a high and possibly real high positive value.

If you've detected a site several times, you have a better idea of the types of objects that might be found there and the values.  As you detect you might be constantly taking note of what learn about the probability of different types of finds.  And that can definitely affect your decisions.

You may have learned that site X has tons of beer bottle cans, smashed beer cans, and other junk, and it has some coins, and perhaps an occasional cheap gold band.  You might figure that there is a high probability of finding some amount of coins, and a probability of finding a cheap gold band, and virtually no chance of finding a very expensive ring or a Rolex.

Compare that with another site, maybe Y where you are likely to find very few coins and very few gold items, but the gold you do find has a high value.  If you use the formula, you might find that site Y, even though it produces few finds, produces a higher average value.  There might be more days when you completely strike out at that beach, but on those days when you hit anything, you do well.  

As I've pointed out in previous posts, finding one Rolex is better than finding hundreds or even thousands of coins.  

If you look at the probabilities and your best estimate of values, you might be surprised and you will have good information for making good decisions.

I hope you can see how the logic of the average expected value formula can be used for site selection.  That kind of logic can also be used to help you make decisions about target ID and discrimination.

I've also talked a lot in the past about keeping good find records.  They can assist you in estimating the likelihood (or probability) of different kinds of finds for different sites.

---

Surf Chart from SurfGuru.com.

Nothing special going on with the surf or tides.

Good hunting,

Treasureguide@comcast.net