Tuesday, September 1, 2015

Do Three Stars Equal a 6.0?

When I finish a book or movie or CD or game, one of my first thoughts is, of course, processing how I felt about it. Did I like it? Would I like to experience it again? How would I compare it to others? Quite often for me, this boils down to a single rating, which somewhat reductively describes my experience with the item. In this way, I can look back later and remember first, if I’ve already done something (read that book, for example), and second, did I enjoy it? One concept I struggle with, however, is the multitude of rating systems. You have stars, half-stars, 10-point scales, 100-point scales, and many others.

Somewhat obsessively, I like to catalog what I’ve experienced. I have multiple accounts that track my music listened, games played, movies watched, books read, and so on. I like to go back and look at what I’ve accomplished. It’s also a good way to compare artistic tastes with friends. Most often, the best way to track this kind of information is by assigning it a rating, and as we know, rating systems can vary wildly.

The confusion of different systems arises because so many publications try to utilize unique ways of reviewing items. Websites and magazines that review products are in constant competition for views. Numeric ratings are often used in conjunction with a longer written review, obviously far more informative. Longer reviews, however, have the larger problem that most people would rather look at the number without digging into the article. Personally, I glance through a review only when I’m interested in the item in question. So, the rating system is a quick way to grab attention for the item.

I have had many debates with friends about what is the most effective system. Largely, these discrepancies revolve around the degrees with which items can be assigned. Five stars obviously gives five levels of definable quality; a 100-point scale gives 101 levels (including 0). As you can see, these get more precise with more levels. My personal preference falls somewhere in between.

100-point scales are a bit excessive. Though they carry a sense of scientific nature, it is mostly impossible to assign a sense of value with so many options. Questions arise like: is a 68 really better than a 66? These statements are obviously impossible to discern. However you feel about aggregate sites like Metacritic, these actually are take an average rating based on a collection of ratings and you can actually explore the differences between a 66 and a 68. But in the end, there is not much of a difference.

Ten stars is tricky. It gives twice as many choices as five stars, but not nearly as many as 100. In this way, we can get more specific responses, without getting ludicrously deep into ratings that differ by 2%. But I have issues with 10-point scales as well. Although I like the scaling, there is an inherent nature of interpreting quality, where any rating below about a 6 is considered ‘below average.’ Perhaps it is largely a problem in the U.S.--where a 60% is generally considered failing in the education system--but this has an enormous effect on how we comprehend low scores. Because of what we (at least in the U.S.) have learned, even something that earns 7 stars is considered mediocre.

The argument for five stars is a good one. The best way I have heard it described is that three stars is average with two levels above average and two below. I believe this is the most popular preference because it is most akin to how we think about quality. Was it very good? Five stars. Just okay? Two stars. The ease of use with a rating system like this makes it easily translatable. Furthermore, three stars is perceived as a better rating than a 60 out of 100. My issues with this system is that there is just not enough to differentiate from other ratings. I don’t enjoy every four-star item the same way; some are better than others. This is why I believe the best way to rate includes five stars with half stars.

I’ve been told that my issue with half stars is that I compare items against each other too much. Simply because I enjoyed something more than another is not reason enough to assign it a higher rating. I acknowledge that this is an issue, but I can’t help it. It’s only natural to compare media, so why wouldn’t you want to say something deserves a higher rating than something else? To me, half stars is as granular as I need to be to denote how I feel about a piece of entertainment.

Obviously, the rating is totally dependent on the rater. I know when it means when I’ve given something four stars, but I also know what it means when I assign something a 7 out of 10. The rating doesn’t matter; it’s how you ultimately felt about the item in question.