Datasets

All data used for the following plots can be found here:

https://docs.google.com/spreadsheets/d/1L18xN2bOUAqCSYOcg8kGGi6i47Iszq8tdFVKHUDxBis/edit?usp=sharing

Descriptions and Timeline

Character Count Across All Korean Chapters

Ridibooks shares an estimate of the number of Korean characters in each chapter, shown below. These were the original counts, before the author did a major edit of the work for ebook publishing. Those newer character counts are included in the spreadsheet above.

Word Count Across All English Chapters

Webnovel’s coin cost is based on estimated English word counts, explained further below.

Note that in Book 1 on Webnovel, chapters 193 and 194 (the later being described as an epilogue) were published separately. They are combined here to more easily compare to the Korean counts above, as in Korean these chapters were published together.

Release times of Korean Original and Various English Versions

The black vertical line is today’s date, 2025-02-24. All chapter publishing dates afterwards are estimates. The Webnovel dates are based on their established publishing pattern: 3 chapters per week, on Wednesday, Thursday, and Sunday at 10PM EDT. The maximum level of Webnovel Privilege is currently 20 chapters, therefore 7 weeks ahead of the Webnovel Free dates. Amazon Kindle dates are based on projected 3 month break between releases.

English Word Count

Explanation of English Word Count Estimate

I have used a program to pull 38 of the official English chapters, to get their exact word counts. Using these word counts, I was able to determine how the Webnovel coin cost coincides with the English word count. Word counts are binned into blocks of 200, and the number of coins depends on what “block” that chapter’s word count lives in. An extra coin is also charged when a chapter count goes over 1000 words (why it is not fully “linear”).

Red points are actual, exact chapter word counts.

Black points are a random spread of possible word counts for chapters based on their coin cost. The exact word counts all fall into their predicted “bins.”

There is also a strong correlation (if not perfect, as the languages are very different) between the Korean character count, and the English word count estimate.

Predicting Final Book Count

Using the estimates based on coin cost for the available chapters, we can get a very accurate English word count the published chapters. Then using the relative Korean character count from what has been published of Book 2, we can estimate the word count of the remaining chapters. I’ve also added the total estimated “coin cost” on Webnovel. Below are those calculations.

Book english_not_published sum_word sum_char sum_cost percent_of_book predict_word_total predict_cost_total
Book 1 FALSE 356100 885000 4107 1 356100 4107
Riftan POV FALSE 94100 207300 1100 1 94100 1100
Book 2 FALSE 381400 940200 4502 1 381400 4502

This gives the final estimated word count of all of what’s been written by Kim Suji of the Under the Oak Tree as:

sum_word sum_character
831600 2032500

Comparing Book 1 to Book 2

There have been several discussions comparing book 1 and book 2, and specifically the amount of sex scenes there are, an the amount of time Maxi and Riftan are interacting.

These concepts are of course debatable, but I’ve tried to quantify this based on the number of chapters that contain a sex scene, or a scene where Maxi and Riftan are interacting.

I’ve defined a sex scene broadly, but also as a scene that contains genital contact by the other person.

I’ve defined interacting as either speaking to each other, or physically contact.

I’ve also added a count for the number of chapters that features Riftan’s perspective!

Comparing the Books to the Comics

Comic episode publishing pace

Plotting Comic Episode to Book Chapters

Here is plot of which book chapters have material covered in each comic epsiode.

Estimated number of Book Chapters covered per Comic Episode

Based on the above, it does appear there are distinct periods. Estimating the slope of each of those lines gets us a estimate of how much of the book is covered each episode, in each period.

group Comic.trend
S1 early 2.0462963
S1 late 1.1410475
S2 0.9551403
S3 1.3028158
S4 1.0739913

Adjusting for cut material

However, one critique of the above estimates is that this assumes that all of the material is covered from each chapter. As we well know, the comic skips sex scenes, which is especially applicable in season 1, where the longest and most detailed of these scenes from the book are covered. The black vertical lines separate the seasons.

Here are my estimates of the number of chapter content skipped in each period:

group sum
S1 early 3.0
S1 late 0.9
S2 0.3
S3 2.0
S4 1.2

Here I have another way of looking at the data, specifically taking percentage of each chapter covered in each episode, removing a percentage of a chapter if it is skipped. This can be somewhat subjective, but feel free to take a look at my estimates in the dataset.

Above analysis adjusted for differing chapter length

Another critique might be that as we saw, there is variety in the chapter lengths. difference in the length of chapters. Below I adjust the number of chapters to reflect the “average” length of a chapter adapted so far.

Comic Megabyte Size

Ridibooks posts the megabyte size of each of the webcomic episodes. This is approximately the “size” of the episode.