Patterns in my AO3 stats
Dec. 20th, 2021 10:50 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
Because I had fun playing with the data about my statistics from flamebyrd's bookmarklet when I was doing the AO3 Averages meme.
In short: Hits seem to not be very closely correlated to word count except for very long fics, which I think is caused by people coming back to multi-chapter fics multiple times, eg each time a new chapter is added.
Fanart seems to get about 2/3 to a half of the feedback that fanfic does by most metrics, and even fewer when it comes to bookmarks. But the feedback on fanfic is also more inconsistent, possibly because most of my fanart was posted for exchanges, where you're generally guaranteed SOME feedback. The most popular works by all metrics are long, multi-chapter fics.
And the thing people on the AO3 like least is Mass Effect fic with an unflattering depiction of Commander Shepard.
At least based on the data to hand. YMMV ;)
Some caveats: My division into "fanart" and "fanfic" by word count is inaccurate in ways that will affect the data. My fanart with more than 260 words tends to be comics, which are more popular, though you could also consider them to be a type of fic. And my few fics with less than 260 words tends to be very unpopular.
I have fudged some of these numbers slightly, for clarity and because I only noticed a couple of incorrect lines of data after making half the graphs >.>
If I remembered any statistics I might have done in-depth quantitative analysis but I don't so you get graphs. But I think the general patterns are accurate, at least for my works.
So! I'm working with 220 works ordered by wordcount. The bottom 110 works, with 0-259 words, are mostly art. The top 110 works, with 260-114,000 words, are mostly fic.
Graphing word count versus position in the list, I get this roughly exponential shape, starting flat and shooting up at the end:

The first 38 works have zero words. Work 180 has around 4000 words, while work 200 has 10,000 words and work 220 has 114,000.
Now to compare some other stats:
Hits:

Hits vary between 20 and 16,000. Below about 4000 words there doesn't seem to be much correlation between word count and hits, but then it goes up dramatically, which I assume is because that's when works start having multiple chapters and getting multiple hits from the same reader.
There's a bit of a bump at around 60 words, I checked it out and those works are a varied collection of fanart, maybe having a few more words implies more complex art ideas that get more hits? Or it's just random variation!
Kudos:
Kudos is much more strongly correlated with wordcount, and there's a very distinct jump from fanart to fanfic. Fanart generally hovers around 20-30 kudos, with one outlier at 545 kudos(*) and the rest generally below 100. Meanwhile fanfic varies more but has more than double the mean number of kudos, at 84, and there's quite a lot of works with hundreds of kudos, especially for works with very large word counts.
Comments:

Comments again divide pretty starkly: Fanart generally hovers below 10, while fanfic often goes higher, especially for longer works. The maximum for fanart is 61, while the maximum for fanfic is 186. Comment counts start to rise dramatically at about the 10K word mark, I wouldn't be surprised if they're closely correlated to chapter count since people are more likely to leave new comments when there's a new chapter.
Hits/Kudos:

To counteract any effect caused by fanart having lower hit counts, I checked out the hits divided by kudos. A low value means the work got more feedback proportional to its hits, and a high value means it got less.
As you can see, the difference between art and fic is again quite pronounced. For fanart, the mean is 64 and the median is 48. For fanart, the mean is 46 and the median is 27. That said the highest values are all for fanfic, there's a bunch of spikes up into the hundreds for what turned out to mostly be Mass Effect fic with an unflattering take on the personality of the player character.
Here's the table of results from my previous post again, plus hits/kudos:
I didn't graph bookmarks or subscriptions but look at those numbers! With subscriptions it makes sense that people will generally only subscribe to multi-chapter works, which for me means large word counts, even for comics. But it looks like people really don't bookmark my fanart much at all, even the stuff they like enough to leave kudos or comments on.
I had always gotten the feeling my fanart got less feedback than my fanfic on the AO3 but was not expecting it to be so blatant! Though this really can't be extrapolated to the AO3 as a whole: the art I post is not the same as the fic (more gen, more often for exchanges etc) plus I am just one person. Like, my most popular works are mostly femslash, and you definitely can't extrapolate from that.
Still, it's interesting. I would be curious to hear from other people who post both fic and art to the AO3, as well as people who post vids, podfic etc. Also anyone else who writes Mass Effect fic where Shepard is a jerk haha.
(*)The Sound of Gears Turning, Katerina/Maria, My Next Life as a Villainess. It doesn't stand out particularly in terms of subject etc but I guess it came out pretty well and it's one of those pairings where people are very grateful for content.
In short: Hits seem to not be very closely correlated to word count except for very long fics, which I think is caused by people coming back to multi-chapter fics multiple times, eg each time a new chapter is added.
Fanart seems to get about 2/3 to a half of the feedback that fanfic does by most metrics, and even fewer when it comes to bookmarks. But the feedback on fanfic is also more inconsistent, possibly because most of my fanart was posted for exchanges, where you're generally guaranteed SOME feedback. The most popular works by all metrics are long, multi-chapter fics.
And the thing people on the AO3 like least is Mass Effect fic with an unflattering depiction of Commander Shepard.
At least based on the data to hand. YMMV ;)
Some caveats: My division into "fanart" and "fanfic" by word count is inaccurate in ways that will affect the data. My fanart with more than 260 words tends to be comics, which are more popular, though you could also consider them to be a type of fic. And my few fics with less than 260 words tends to be very unpopular.
I have fudged some of these numbers slightly, for clarity and because I only noticed a couple of incorrect lines of data after making half the graphs >.>
If I remembered any statistics I might have done in-depth quantitative analysis but I don't so you get graphs. But I think the general patterns are accurate, at least for my works.
So! I'm working with 220 works ordered by wordcount. The bottom 110 works, with 0-259 words, are mostly art. The top 110 works, with 260-114,000 words, are mostly fic.
Graphing word count versus position in the list, I get this roughly exponential shape, starting flat and shooting up at the end:

The first 38 works have zero words. Work 180 has around 4000 words, while work 200 has 10,000 words and work 220 has 114,000.
Now to compare some other stats:
Hits:

Hits vary between 20 and 16,000. Below about 4000 words there doesn't seem to be much correlation between word count and hits, but then it goes up dramatically, which I assume is because that's when works start having multiple chapters and getting multiple hits from the same reader.
There's a bit of a bump at around 60 words, I checked it out and those works are a varied collection of fanart, maybe having a few more words implies more complex art ideas that get more hits? Or it's just random variation!
Kudos:

Kudos is much more strongly correlated with wordcount, and there's a very distinct jump from fanart to fanfic. Fanart generally hovers around 20-30 kudos, with one outlier at 545 kudos(*) and the rest generally below 100. Meanwhile fanfic varies more but has more than double the mean number of kudos, at 84, and there's quite a lot of works with hundreds of kudos, especially for works with very large word counts.
Comments:

Comments again divide pretty starkly: Fanart generally hovers below 10, while fanfic often goes higher, especially for longer works. The maximum for fanart is 61, while the maximum for fanfic is 186. Comment counts start to rise dramatically at about the 10K word mark, I wouldn't be surprised if they're closely correlated to chapter count since people are more likely to leave new comments when there's a new chapter.
Hits/Kudos:

To counteract any effect caused by fanart having lower hit counts, I checked out the hits divided by kudos. A low value means the work got more feedback proportional to its hits, and a high value means it got less.
As you can see, the difference between art and fic is again quite pronounced. For fanart, the mean is 64 and the median is 48. For fanart, the mean is 46 and the median is 27. That said the highest values are all for fanfic, there's a bunch of spikes up into the hundreds for what turned out to mostly be Mass Effect fic with an unflattering take on the personality of the player character.
Here's the table of results from my previous post again, plus hits/kudos:
Words | Hits | Kudos | Comments | Bookmarks | Subscriptions | Hits/Kudos | |
---|---|---|---|---|---|---|---|
All | |||||||
Max | 114920 | 16098 | 830 | 186 | 223 | 248 | Infinity (approximated by 500) |
Mean | 4175 | 1695 | 59 | 6 | 10 | 9 | 55 |
Median | 253 | 1011 | 27 | 2 | 2 | 1.5 | 35 |
Fanfic | |||||||
Mean | 8332 | 1922 | 82 | 8 | 16 | 13 | 46 |
Median | 2412 | 1071 | 35 | 3 | 4 | 2 | 27 |
Fanart | |||||||
Mean | 56 | 1470 | 36 | 3.5 | 3 | 2 | 64 |
Median | 29 | 1069 | 22 | 2 | 1 | 1 | 48 |
I didn't graph bookmarks or subscriptions but look at those numbers! With subscriptions it makes sense that people will generally only subscribe to multi-chapter works, which for me means large word counts, even for comics. But it looks like people really don't bookmark my fanart much at all, even the stuff they like enough to leave kudos or comments on.
I had always gotten the feeling my fanart got less feedback than my fanfic on the AO3 but was not expecting it to be so blatant! Though this really can't be extrapolated to the AO3 as a whole: the art I post is not the same as the fic (more gen, more often for exchanges etc) plus I am just one person. Like, my most popular works are mostly femslash, and you definitely can't extrapolate from that.
Still, it's interesting. I would be curious to hear from other people who post both fic and art to the AO3, as well as people who post vids, podfic etc. Also anyone else who writes Mass Effect fic where Shepard is a jerk haha.
(*)The Sound of Gears Turning, Katerina/Maria, My Next Life as a Villainess. It doesn't stand out particularly in terms of subject etc but I guess it came out pretty well and it's one of those pairings where people are very grateful for content.