Top 100 Anime According to MyAnimeList (by Weighted Score)
Did you know Reddit compresses images even further when you add two or more images to a post? I didn't, hence the reupload.
A few months back now, during the discourse surrounding how to fix MAL's Top 100 Anime list, a user compiled this post that compressed MAL's Top list based on average score for a franchise. This is a decent solution for fixing the sequel spam, but still runs into the issue of treating all seasons equally, so later seasons that will inevitably have less votes actually get weighted higher than maybe they should. The logical solution is to then calculate a weighted average based on number of votes, and that's what we get here.
This is MAL's Top 100 Anime compressed down to single entries per series using the weighted average formula:
>SUM(entry score * number of votes on entry)/(total votes across entire series)
I decided to calculate this number twice, hence the two lists. The list on the left looks at all entries in a given series including spin-offs, movies, and non-canon material that a diehard fan of a series might watch. If this kind of averaging would be implemented, this would likely be what happens, but is maybe a little unfair. To remedy this, I also calculated (to the best of my ability) the same score, but only using canon content along the main storyline. For example, for Hunter x Hunter (2011) this would exclude the two non-canon movies and just cover the main series. It's hard to compare how these spots fare against the vanilla list, but for the "canon only" list I added in the amount of spots and points a show moved between calculations. Entries that pop up from below Top 100 aren't given a position rise (instead marked "NEW") because I only calculated as low as the 100th spot, so while I can guarantee Top 100, below that it technically gets sketchy.
A few other notes on what was counted:
- Commercials featuring the characters were not included because that seemed like a little much.
- Remakes are counted as separate continuities and thus separate entries (FMA 2003 and Brotherhood are separate).
- Franchises with multiple "parts" are split in a way that one could theoretically watch as stand-alones. For this data set this meant that Uma Musume's main series, ONA series, film, and Cinderella Gray were kept separate. Each Jojo part was taken in isolation (which works out given that the only parts that appear, 4, 5, and 7 are fairly stand-alone-ish). Fate/Zero is separate from each of the Fate/stay night paths which are also treated separately.
- Recap movies are counted as a separate timeline if they terminate on a novel ending or feature new scenes (i.e the Madoka Magica films, Lelouch of the Resurrection, and Gurren Lagann films are separate from their respective series).
- Donghua is included. Should it? Probably not, but MAL counts it so to stay true to that, it's also counted here.
- Data for the newest shows might be slightly outdated. I kept it current through the date listed, but scores kept fluctuating and eventually I had to cut it off. Smoking Behind the Supermarket With You is also not listed but with a score of 8.51 (as of writing) would fall in 60th and 68th on each respective list.
- I did not hard recalculate each number and instead just took the stated three-digit number on the main page. This leads to a bunch of ties for single-entry shows. MAL would be able to break these with more precise numbers, but I opted to use "most votes" as the tie breaker in cases like this. Final numbers listed are also rounding the average that I got from the seasons after sorting. There's some statistical crimes going on here, but we'll just ignore it. This means a show listed as an 8.60 can technically be as high as a 8.604 and as low as a 8.595 which further factors into "tie-breaking".
A Few Conclusions
- This list still kinda sucks. There's a lot of things I certainly wouldn't include that become unfortunate benefactors of the numbers. However, if you don't take the list too literally, it is a pretty good representation of what your average anime fan (who I will remind you does not largely watch older anime) would say the best anime of all-time are. If we view the list as a "guide for a beginner to crash course anime" then it serves that end surprisingly well.
- The list is surprisingly competitive, with fluctuations of only a few hundredths of a score being the difference of around 10 spots in the lower ranks, making rising easy but falling just as easy. This also means there's still plenty of room for new entries to tighten up the list further.
- Newer shows do tend to deflate in score once the hype dies down and more scores get added on, leading me to have to make multiple revisions to the Winter 2026 and Spring 2026 additions. This doesn't really come across in the data set proper, but is an observation I noticed while making it and I'm sure if you checked some of these scores now, they'd be even lower.
But I leave it now to you guys. I don't have the data set on me right now, as I'm away for the holiday weekend, but can find a way to post it on Monday if people are interested. In the mean time, what other changes would you suggest adding to further clean up the list? And I don't just mean "include X series" or "downrank Y series".