Estimate the reading time for each manga

[TheLunarCarver]

The idea is to look at the dimensions and amounts of pages in each chapter and calculate an estimated total reading time of a manga (maybe also separate values for each language too).

Obviously this can't be calculated as precisely as books with just words in it, since some pages are denser than others, but it would be better than trying to guess just based on the number of chapters. I would start with something like 20 seconds per regular manga page, 10 seconds per 9:16 page scroll of a web comic, and 40 seconds per 4-koma manga page.

 

[GipoScribantino]

Going to translate a 4koma in rhyming hendecasyllables with as many figures of speech and other poetic devices so even if the estimate says 30 seconds the real reading time is 10 minutes.

 

[Angry_Panda]

Going to translate a 4koma in rhyming hendecasyllables with as many figures of speech and other poetic devices so even if the estimate says 30 seconds the real reading time is 10 minutes.

Yes, please. Would read.

 

[Manko-sensei]

This would cause a great clash between those who "read" manga to look at the pictures and those who try to enjoy the titles. My suggestion is visiting the appropiated thread for bad suggestions.

 

[TheLunarCarver]

This would cause a great clash between those who "read" manga to look at the pictures and those who try to enjoy the titles. My suggestion is visiting the appropiated thread for bad suggestions.

That is a funny thread. I still think my suggestion is feasible though. You could try to account for differences between readers by asking them how fast they typically read a page, and account for differences between titles by having the website time how much time users spend reading them instead of X seconds for each type of page. It's less that I want to do robotic time management on my reading, and more wanting to know how long different mangas are compared to each other, because the chapter count doesn't tell me jack when it's a webcomic.

 

[GipoScribantino]

You could try to account for differences between readers by asking them how fast they typically read a page, and account for differences between titles by having the website time how much time users spend reading them instead of X seconds for each type of page.

I'm going to respond the questionnaire saying I'm a fast reader and then stay on a single page a significant amount of seconds, maybe minutes, in addition of my actions from my previous post.

 

[UnnamedPlayer]

Separate the hentai from the rest, their "reading" time would skew everything.

 

[BzzBzz]

You could try to account for differences between readers by asking them how fast they typically read a page,

People don't even bother to rate manga even though it literally takes 2 clicks. Do you expect them to answer 10-min survey how fast they read this or that?

and account for differences between titles by having the website time how much time users spend reading them instead of X seconds for each type of page.

I often switch to other matters leaving manga chapter opened for hours. How will you teach website to time it? Do I read single page 3 hours or just do something else?

 

[Trobilobada]

Read? 👨‍🦼

 

[jecroisenlasuperioritedubel]

Put More trackers , collect More data , analyze More data , remove ſome restraints , chꝏse an Average-creation techniquͤ.

 

[sanctuary]

Separate the hentai from the rest, their "reading" time would skew everything.

Ha! a funny french, now I've seen it all

 

[garagex]

User behavior could totally skew the data if people mess around with the timers or just read in unusual ways. I guess any system like this would have to allow for a big margin of error or only show rough averages.

 

[WindOcean]

I don't necessarily mind such estimations, but I don't really like being tracked either. Even if it is anonymized.
If there is the possibility of abuse, it will be abused. (maybe, lol)

 

[TheLunarCarver]

Responses to problems pointed out:

• "10-min survey how fast they read": I meant this to be a single adjustable setting for users that would simply be a global multiplier on estimated reading times.
• Users trying to create junk data: I don't see a motive to abuse the system unless they also add a "sort by length" option. If it actually does become a problem, maybe try using median instead of mean.
• Users genuinely lingering on one page (going AFK or "doing research"): Should be easy to detect outliers and either set a maximum value or just discard them.
• User privacy: Ask users to opt in to having their reading timed.

 

[undeed]

What an overly complicated solution.
Just pick a number like 6 or 7 minutes,
times highest count chapter currently showed,
and be done with it.

That is how many other reading apps and sites do it anyway.
No need for overly complicated justification.
It's not like getting the most accurate "estimation" possible will cure cancer or something.

 

[GipoScribantino]

Responses to problems pointed out:

• "10-min survey how fast they read": I meant this to be a single adjustable setting for users that would simply be a global multiplier on estimated reading times.
• Users trying to create junk data: I don't see a motive to abuse the system unless they also add a "sort by length" option. If it actually does become a problem, maybe try using median instead of mean.
• Users genuinely lingering on one page (going AFK or "doing research"): Should be easy to detect outliers and either set a maximum value or just discard them.
• User privacy: Ask users to opt in to having their reading timed.

You are underestimating a lot of factors.

• a lot of people don't care about how fast they read (I don't, for instance) and they will either set it to "unknown" or use a random value, effectively making it as useful as using an arbitrary fixed number with no user feedback;
• anything that can be abused will be abused, even just to be "funny". Some will even make it a matter of principle because they are a master troll of snafu and chaos. Changing which indicator you use doesn't change this fact;
• the amount of readers on the site might have decreased after the Great Chapters Purge, but it is still larger than what you are thinking of. Outliers are anything but easy to filter out, at least automatically. Removing anything above a threshold will delete important data, especially if the threshold is too low;
• anything opt-in will be activated only by a small amount of people, doubly so if it's for tracking. It will be so small to be useless, in fact.

I understand you want to have a genuine conversation and flesh out an idea, but this is simply a bad suggestion with way too many flaws to ever be useful and even if implemented, the inaccuracy would be so evident people will just outright ignore it.
I'd rather spend effort to provide proper sorting in lists.

 

[TheLunarCarver]

What an overly complicated solution.
Just pick a number like 6 or 7 minutes,
times highest count chapter currently showed,
and be done with it.

That is how many other reading apps and sites do it anyway.
No need for overly complicated justification.
It's not like getting the most accurate "estimation" possible will cure cancer or something.

That is why I led with a simple solution when I started the thread .

 

[undeed]

That is why I led with a simple solution when I started the thread .

yeah, I don't hate the idea
it's fine, it's whatever
any other reading app also do it

it's not that complicated in my mind, no need to overthink it
who cares if in actuality you can read 100 pages in 1 minutes
who cares if someone read a 4-koma page in 5 minutes
it's called estimation for a reason

you already have the number of chapters (or chapter groups) on the page
multiply by say 6 minutes or whatever, don't worry about it
put it beside that mark everything read/unread
and it's done

 

[WindOcean]

who cares if in actuality you can read 100 pages in 1 minutes
who cares if someone read a 4-koma page in 5 minutes
it's called estimation for a reason

If this metric has no accuracy, it has no meaning.

But we could show the accuracy of the estimation, via box plots.
Every manga one fat plot

 

[undeed]

If this metric has no accuracy, it has no meaning.

But we could show the accuracy of the estimation, via box plots.
Every manga one fat plot

this is not a peer-reviewed paper, my dude
there's no need for ablation study
it's all about user satisfaction to the UI
just like random in spotify is not actually random
I trust readers are not that stupid and can adjust the "written estimation" accordingly

"oh, it says 6 minutes per chapter, I can read faster than that. so if it says 1 hour to read all chapters, it shouldn't take me half an hour"