So here's some crude assumptions to establish a baseline:
Let's assume that Jiraiya spent about 10% of his waking hours writing notes and stuff,
except for Sealing notes which would probably include more writing.
We know that about 20% of his output is sealing notes, and if we assume that Sealing as an activity creates 3x as many notes for the time invested then Jiraiya spent about 6.7/86.7 of his time Sealing. In other words, averaged across J's life we can crudely assume 7.7% of his time is spent Sealing and 92.3% of his time is spent doing the many other things Jiraiya did.
I hear '40 years' tossed around so let's assume that's how long he's been doing things. 40 years is 233,600 waking hours (assuming 8h a night). Using the above numbers that's 17,969h doing Sealing and 215,631h doing other stuff.
Again, we assume 10% of his time is spent writing notes, except Sealing which is 30%. Thus we get 21,563h of note-writing from other things and 5,391h from Sealing, for 26,954h spent writing overall.
We can use this as a SAN check as-is, since the time spent to write it should be comparable to the time needed to decode it. This would suggest something on the order of 4.5 years to decode it all. Another way to take it would be to use the writing speed numbers described above (1,800 letters per page, 100 letters per minute) to conclude that Jiraiya wrote about 89,647 pages over his career, which would take (using same numbers) 26,954h (because same numbers), i.e. 4.5 years to decode it all.
These values are highly sensitive to things like 'how much of J's time did he actually spend writing things?' You can see that the 4.5 years pretty much maps to '10% of his career, plus some change from Sealing generating more notes'.
Another factor you could introduce is that Jiraiya might take more time to think up what to write, and thus write at a slower pace than can be decoded. This could easily reduce the number of pages by half or even more. A reverse factor could be shadow clones, if J made use of those to write more.
The main point, though, is that 90k pages is the rough upper limit of reasonable approximations of Jiraiya's lifelong output. The previously-estimated 2,700,000 pages does not fit with what Jiraiya could have reasonably written in his lifetime.
How does this jive with what Hazou said about the quantity of boxes left to us? Well, assuming Hazou was telling the truth about how many boxes there are (1000-1500), then they're only 3.3% as space-efficient as google says should be possible for modern paper. Factors that might contribute to this:
- Thicker paper
- Larger script (fewer letters per page)
- Boxes not being completely full
- Scrolls or books or other paper-containing objects bulkier than sheets of paper.
These can go a long way, but eyballing things I can't personally see them reaching less than 10-20% space-efficiency, so it might be worthwhile to decrease the total number of boxes to some degree. If we cut the total number of boxes by a factor of 5, reducing it to 200-300 boxes in total (of which 40-60 are Sealing (of which we gave 3/5 of the next box)) then the overall space efficiency rises to 16.7%, and the above factors could properly explain the discrepancy. Alternatively, if Hazou was bluffing about how many boxes there were, then that could also neatly solve the problem.
Edit: ROUND 2
After some (surprisingly unhelpful) conversations on Discord, I'm going to recalculate a little bit and try to establish a broader range of reasonable values.
It was mentioned that J's Spy Network work would likely involve a lot of writing, perhaps more than his Sealing work. Let's assume that, averaged across his entire ninja career, about 30% of Jiraiya's time was spent doing spymaster stuff. We now have a ratio of 7.7:30:62.3 between Sealing:Spymastery:Misc.
The useful information I did get from the conversation persuaded me that Jiraiya would have spent less time writing in his non-spymastery, non-Sealing day-to-day life. Not
zero, because of things like his philosophical writings and Icha Icha, but maybe less than 10%. So let's ballpark 5% of the time for Misc., 30% of the time for Sealing, and, say, 40% of the time for Spymastery. Taking the same 40 years timeframe, that maths out to 5,396h of Sealing-writing, 28,032h of Spymastery-writing, and 7,277h of Misc. writing.
Now, let's add a discount to that. Writing is slow and 'writing time' isn't going to go as fast as you can copy text (which is the source of the numbers used above). My intuition says that for Misc. the discount should be large, as J is doing things like writing about philosophy and stuff, that the discount for Sealing should be middling-large, as it's technical work and putting already-brainstormed interactions and stuff down to paper, and that spymastery discount should be middling-small, as much of it is presumably managerial-style work that probably isn't as difficult on the text-generating-speed.
So, let's assume a discount of 80% for Misc., 60% for Sealing, and 40% for Spymastery. The numbers then turn into 2,158h for Sealing, 16,819h for spymastery, and 1,455h for Misc. (messes with the 1/5 sealing ratio but whatever). Now that these discounts have been applied, we can calculate about 68k pages in this version, which would take about 3.5 years to transcribe.
Given that discounts were not applied to the first estimate of 4.5 years, it can be assumed that the 4.5 years would likely become less than 3.5 years after discounts. I think it reasonable to estimate 3-3.5 years to properly transcribe everything. This does not account for time spent cracking codes, or any slowdown Kagome might experience in the decoding process (that is, it assumes that Kagome knows every key and can decode in his head without slowing down at all. This may be far or not far from the truth.)
Using the above calculations that Kagome has been working at this hobby-speed for 57 days and full-time for 40 days, if we assume Kagome's version of 'hobby speed' is 75% full-time speed, then we get about 83 full-length days of work on this, out of 1095-1278, meaning Kagome is about 7% done decoding, and that he has 1012-1195 days left until he is finished.
Given the large amount of spy network contributions in this model, Kagome could probably save a lot of time by only decoding until he knows that the notes in front of him are spy network notes, and then setting them aside alongside the cipher key so the Tower can decrypt them on their own. If so, we could apply another discount to the spy portion of the notes, say 75%, and reduce the total time needed to about 1.3 years (~489 days), of which he has completed 17% and has 406 days remaining.
