Alex Olteanu
Apr 21, 2017 · 2 min read

That is a really nice suggestion. I will try to see how can I actually do that when I will have the time (I’ve just started working on another project).

For me, the tomatometer was quite problematic for two reasons:

  1. It was difficult to integrate it in the framework of my analysis (bad-average-good).
  2. It is vague.

When I researched for the article, I read that the tomatometer represents the percentage of professional critic reviews that are positive for a given film.

I found the term “positive” vague, so I researched more, without much success.

On the contrary, my opinion was reinforced by scenarios like these, where two identical ratings are interpreted as positive or negative:

2/5 => positive
2/5 => negative

Now, addressing another issue. I don’t know what you actually understand by “biased”, but to me it sounds a bit harsh for what I’ve done in the article. I’ve made two assumptions:

  1. Movie ratings should reflect movie quality.
  2. Most people experience most movies as being of an average quality.

For 2), the other two main options were:

  • Most people experience most movies as being of a low/high quality.

These other two options seemed really unlikely for me. Plus, for the first one I could actually provide one argument: my own experience.

So the main idea is that I chose to an assumption over another. I wouldn’t call it biased.

This kind of thing happens a lot in science, and it doesn’t seem reasonable to call it a biased approach. For example:

  • In cosmology, when you construct your theoretical model, you may need to pick between these two assumptions: the universe is infinite/finite. You cannot prove true or false neither, so you’ll have to pick one and build a model.
  • In all sciences, it assumed that all phenomena have a cause — sometimes, you observe a phenomenon, and you start searching for its cause, assuming thus that the phenomenon has a cause. You have no proof that that particular phenomenon has a cause, you just assume it does, by induction. It could be that we live in a supernatural world where not all phenomena have a cause — this is the alternative assumption.

I wouldn’t call these approaches biased. For me, “biased” is always associated with something negative. And I tried to do something positive, constructive in this article.

However, if you enlarge the domain of the term’s reference, I guess you can call these approaches biased, as well as my analysis.

    Alex Olteanu

    Written by

    I write data science courses at Dataquest.io

    Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
    Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
    Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade