What are the ways to translate user ratings into a point system?

Question

Analytics

Greetings.

A survey was conducted and it was in an open form. The question sounded like this (example): " What criteria are important to you when choosing a fishing store? " One person could name any number of criteria important to him. When a criterion was found in the answer, it was entered into the table with the value 1, and when an already recorded criterion was found, a point was added to the table in another answer. As a result, a table was obtained, consisting of 10 criteria with different scores. The criteria were divided into two groups: attraction, equipment.

Attraction
3 - work schedule
9 - reviews
11 - live worms
17 - advice from friends
3 - discounts

Equipment
3 - availability of marmosets
1 - the presence of bells
4 - all types of hooks
7 - work at night
4 - quality of goods

How now to correctly translate the number of points received into a ten-point scale?

It is possible to divide the largest number in each group by ten and thereby get a unit of the group's point system, and then divide all the numbers from the group by the received conventional unit and thus get an estimate on a ten-point scale for each criterion, but is it true - the question!

Answer 1

I do not think that repeated repetition should infinitely increase the level of the assessment for this score, especially if the assessment is then normalized by the number of all assessments, in which case information about other merits will have little effect.
Can each criterion be given a significance factor? Is the criterion 'presence of jig' is equivalent to 'discount'? it’s just difficult to come up with such coefficients, .. as an option, you can initially collect data, and then set the value of the coefficients inversely proportional to the number of mentions.
Another option is to score for each criterion up to a certain limit, i.e. up to 10 (choose experimentally, maybe changing periodically) or take the logarithm of the number of times the criterion is mentioned, i.e. 2 mentions are not twice as bad as 4
I hope your analytics understands that if a client complains - 'why don't you have live worms?', this does not go well by this criterion?
ps the goal is what in the end?