Danish West Indies : Frequenzy?

I have just entered my first coin from the DWI and saw the frequencies below, admittedly I don't understand them at all.

Do 53 to 54% of all numista users really own this coin? I know the rarrity index is not to be interpreted either, but it's around 28%, so half of numista's members can not have the coin.

Who can explain the frequency in a better way, than it's described in Numista, please?

Thanks in advance for enlightening me.

The explanation is at the bottom of your picture. If 100 members have this coin then 53 members have the 1859 coin. 
 

if 200 members have this coin then 106 have the 1859 coin. 
 

The NRI is an indicator of how many members have this coin. 

Of all the people who (claim to) have that coin, the % indicates how many of these have that specific variety.

It doesn't add up to 100% if users (claim to) have more than one of the varieties.

I don't see the need to make it even more clearer. I mean, who really thinks 53% of Numista members have that coin?

rsirian1

The explanation is at the bottom of your picture. If 100 members have this coin then 53 members have the 1859 coin. 
 

if 200 members have this coin then 106 have the 1859 coin. 
 

The NRI is an indicator of how many members have this coin. 

Enfin I understand it, but it's explained extremely badly in numista: 

Frequencies show the percentage of numista users who own each year or variety among all the users who own this coin.

The bold percentage is for me the percentage of all numista user! So when 53% is shown it doesn't really mean that at all.

I suggest to use Rsirian1's explanation, since it's clear and precise, the first line is enough:

If 100 members have this coin then 53 members of those have the 1859 coin. Frequency thus 53%.

I would like to see number of members owing the type as well.

Am extreme example.

Line 1 - 0%

Line 2 - 50 %

Line 3 - 100 %

Nobody that owns a type of this entry owns a Line 1, everyone owns a Line 3 and every 2nd person owns a Line 2 too.

You still know nothing about the population owning the coin. In your example, it might be only two, four, six, eight, etc. In my opinion, the percentage is no good without the total number of owners of the coin type.

Sjoelund

Enfin I understand it, but it's explained extremely badly in numista: 

Frequencies show the percentage of numista users who own each year or variety among all the users who own this coin.

 

The bold percentage is for me the percentage of all numista user! So when 53% is shown it doesn't really mean that at all.

 

I suggest to use Rsirian1's explanation, since it's clear and precise, the first line is enough:

 

If 100 members have this coin then 53 members of those have the 1859 coin. Frequency thus 53%.

 

I would like to see number of members owing the type as well.

That explanation is only really bad because you highlighted a specific part of the sentence without seeming to care about the rest.

Rsirian1's explanation is for middle schoolers. And I don't see why Numista should become like one of those government websites that also have a dumbed-down version for weak readers.

Trooper8

Sjoelund

Enfin I understand it, but it's explained extremely badly in numista: 

Frequencies show the percentage of numista users who own each year or variety among all the users who own this coin.

 

The bold percentage is for me the percentage of all numista user! So when 53% is shown it doesn't really mean that at all.

 

I suggest to use Rsirian1's explanation, since it's clear and precise, the first line is enough:

 

If 100 members have this coin then 53 members of those have the 1859 coin. Frequency thus 53%.

 

I would like to see number of members owing the type as well.

That explanation is only really bad because you highlighted a specific part of the sentence without seeming to care about the rest.

Rsirian1's explanation is for middle schoolers. And I don't see why Numista should become like one of those government websites that also have a dumbed-down version for weak readers.

Thanks, I think, I better stop here.

Would it be so hard to change the algorithm to one which added up to 100%? All we need to do is to add up all the coins registered for a type, then divide the number for each year line by that number. It would be far more meaningful than the current method and would prevent many people from simply ignoring the value due to its rather contrived basis.

Why not put the amount of people who (claim to) own a certain coin,
  instead of this confusing percentage-thing?

Yvon, Well spoken. 

Hello everyone, in relation to the topic I think that for the first time I disagree with the opinion that has been given here by some users with whom I almost always agree:

1. As a collector for years and variants, I love the frequency tool that Numista introduced some time ago, both for its usefulness (so I can see which years/variants are more common and which are more difficult to obtain) and the format and description I think which are the correct ones. I think it is perfectly understood that of all the people who own that coin/banknote, a certain percentage owns that year.

2. Changing the algorithm so that it adds up to 100% makes sense when, as in the case of the coin shown by Ole, there are only a few lines of years. When there are many lines, the value would be very distorted if it had to be a total of 100% between all of them, and 25% of a line could become 1% or 2%. To me that doesn't make sense.

3. Putting the total number of users who have each specific year does not indicate anything, while with the % you have a more concrete idea of ​​whether that year is common or difficult to achieve. Maybe something similar to the total number and the percentage could not be too bad Ex. 1342 (23%)

You make a good point about types with many year lines. One way around that would be to have 2.3% rather than rounding it off to 2%, so as to give a better representation of the frequency. The problem with the current system is that it doesn't actually show the frequency with which a particular year is encountered (the obvious meaning of the term) but you only know that if you read the explanation. I would guess it was developed when Numista was being designed primarily as a tool for collectors to record their collections, rather than as a catalogue in its own right. The rarity index is meant to show how common or rare a coin is but that only applies to the type as a whole (and is quite well hidden). The fact is that we have a database of many collections and we could use this to give the community some useful statistics but we really aren't doing that at the moment.

Why can't we have both the percentage (if you insist) and the number of people as well? 

You are right on the subject of statistics, I am a lover of these and I would love for Numista to use the data it has to show more individual and collective statistical data.

Again, this is not something I can change, so I will formally reject, however, if you want statistics change, please use website forum with a ticket as “suggestion to improve Numista”.