One unique thing about anime fans is that they will often publicly track how much anime they watch. Sites like my anime list (MAL) exist mostly for this purpose.

This raises the question: how much anime have anime fans seen?

The Result

One day is defined to be 24 hours. So someone with 2 days of watch time would have watched 48 hours of anime. Here we can see the maximum amount a typical anime fan watches goes up to about 350 Days and then there are a few people in the tail which goes out pretty far. For reference only around 1% of the data is greater than 355 Days.

Notes on the data

The data was collected from MAL online users. The data contains a total of 351,019 entries. Users of MAL have likely watched a bit more anime than the typical anime fan, so the results are likely a bit higher than if the data were sampled from all anime fans. Additionally, MAL is mostly aimed at western anime fans, and the results for other groups might be different.

For the analysis I excluded the top .1% of the data because there were some unrealistic numbers such as those where someone reported watching over 50,000 days of anime, which is over 60 years of watching anime nonstop. I assumed these people were using MAL for something other than tracking the anime they have seen because that is impossible. As a result the maximum time spend watching anime in the analysis is 793.70 days¹, which is still slightly unrealistic in my opinion, but it is at least plausible².

I excluded all data for people who had exactly 0 days for both anime and manga for a similar reason.

The data is located at ./data.tsv and the data for the favorite anime is located at ./data_top_10_anime.tsv. The script to make the exhibits is located at ./make_exhibits.R. Please audit the data yourself, and if you find I have made a mistake let me know.

For all graphs with confidence intervals the confidence interval is 95%.

Manga read time

A substantial proportion of users have read no manga at all. In my experience, it is not uncommon for anime fans to not read manga or light novels. Also, "My Anime list" is clearly indented for tracking anime, so some manga readers might not use it.

Also, it should be noted that it is harder to relate manga read to actual time spent reading it.

Anime vs manga

There seems to be a correlation between time spent watching anime and time spent reading manga. While there is undoubtedly a relationship (the p value is less than 2.2e-16), the R² is only about .12 so it is not super predictive. For all relationships that will be discussed this is true.

For those with exactly zero hours of anime watch time the average manga read time was a bit higher. This probably reflects the group of people who only read manga, but don't watch anime or just use MAL to track their reading.

Watch time vs Anime score

MAL allows users to score the anime and manga they watch. It seems like people initially start with a lower score, then go up higher and have a gradual decrease as watch time increases. This could be because early on people learn about the existence of a lot of anime, and watch the ones that appeal to them most early on. Alternatively it could be because as people watch more anime they develop more sophisticated scoring systems leading to a lower average score.

A similar relationship seems to hold for manga score. This suggests that the way people score manga and the way people score anime may be similar. In particular, I think there are a lot of manga that are well known amongst regular manga readers that are very good, but are otherwise somewhat obscure. For example, the works of Satoshi Mizukami tend to be highly rated, but are not widely popular. In my opinion this is evidence that the initial score increase is due to discovering some very high quality media.

Similar to the Anime vs Manga chart there is quite a bit of variation within each group.

Popularity of different types of anime by watch time

I was curious if different types of anime are popular with different groups. To figure this out I picked a few entries for each genre and checked how often they appeared in the top 10. For anime that are in a series only the first entry was tracked except for Monogatri where the third Kizumonogatri movie was also included³.

Obviously you could pick entries for each category to change the results a lot, but these were just anime I thought of. For all categories the probability of it being liked goes up. This makes sense because people who have not watched much anime might not even have logged 10 favorite anime, so there are more total chances for one to appear as time goes on. Most categories were similar except for top-rated anime and Shounen anime which both went up significantly more.

Shounen anime tends to be very popular overall so probably more people have seen them in every group allowing it to go up more.

Conclusion

I would be interested to see how this compares to other sites like Anilist. I would assume they would be similar, but it would be interesting to test.