Hacker News new | comments | show | ask | jobs | submit login
Who wrote the 'Death Note' script? (gwern.net)
183 points by gwern 1603 days ago | hide | past | web | 79 comments | favorite



" the studios initially wanted to give the main character Light Yagami a new background story to explain his downward spiral as a villain. The new background would have had a friend of Light murdered when he was young. When Light obtains the Death Note - a notebook with which he can put people to death by writing their names - he uses it to seek vengeance"

This really makes me cringe. This is precisely the sort of thinking that gave us so many mediocre/bad movies. Only someone who has no idea of what the story is about could suggest such a crude alteration of the story. And for what? Do they really think the movie would be better because of it? Or at least draw more audience?

Anyway it's great that the director managed to hold off these changes, I saw Kiss Kiss Bang Bang and it was absolutely great so I'm anxious to see what he'll make of it :)


When I read that I thought of Kevin Smith's conversation regarding Superman and a Giant Spider - http://www.youtube.com/watch?v=vgYhLIThTvk#t=8m0s

My current theory about movies at the moment is that all these crappy movies get made just to put a trailer to a big "block buster" at the start. After cost + advertising - box office revenue is taken into account, the end cost is probably less than it would cost normally to mass market the initial trailer. -- The reason why bad movies get made :D


> A forger could have randomly changed Charles to "Charley", looked up an appropriate address, edited the metadata, come up with all the Hollywood touches, wrote the whole damn thing (quite an endeavour since relatively little material is borrowed from DN), and put it online.

Most people would have used that pile of circumstantial evidence to close the case. But you can't fool gwern (or L).

The best thing about this analysis is that it reflects the thinking of the one of the main characters in the anime. This seems like how the detective L would solve the same problem. has always been one of my favorite fictional characters because I greatly admire his (fictional) rigor and inductive abilities. Leave it to gwern to apply the same level of super detective skills to answering a question about the movie. It seems like a lot of work for a small point, but we'd all be better off if more people were this thorough in eliminating possibilities.

I saw the japanese movie and it was good, but probably best for people who were already fans of the anime. An americanized version might open up a much broader audience, though they'd have to put a lot of money behind it.


One thing I found strange about his analysis is that he kept on looking for new approaches after he got an upper boundary of 6.6% for the style result being random. I'd have introduced a more realistic model and looked at whether it get's me below 3% and be done with it (probably). He seems to like analysis just for the analysis sake. Of course it's still refreshing to see after reading tons of today's superficial blogosphere blabber (Google has ALWAYS blocked WP!!!11! ... wait no, sry).


I don't think you can consider the stylometric result as yielding such an upper boundary; yes, that's some strong evidence against one possibility and in favor of another, but we could easily find some more evidence which goes the other way - and indeed did.

So, it may be an upper bound _ceteris paribus_, but ceteris is not paribus here.


Warning: his whole site is an incredibly wonderful time sink :)

I can recommend his extensive research on Nootropics [1] for actionable advice.

[1]: http://www.gwern.net/Nootropics


A few of these are not nootropics but cognitive enhancers. Difference being that nootropics are extremely non-toxic or neuroprotective. Another way to put it is that nootropics are a subset of cognitive enhancers (safe cognitive enhancers, basically).

The following are not nootropics:

* Modafinil and Armodafinil

* Adderall (Amphetamine)

* Caffeine

* Kratom

* Nicotine (this is a border case, in my opinion)

* Selegiline

Stay away from the above, unless taken very occasionally, if you value your health. Drugs like these rapidly build tolerance and become a liability in the long run. If you insist on using stimulants such as amphetamine, learn about how to reduce tolerance with things such as NMDA antagonists and take things to prevent neurotoxicity.

He should also get on with trying Noopept - it's one of the best nootropics of recent times.

From experience, anybody who is a programmer or does similarly demanding intellectual work will stand to benefit from learning about and using nootropics. They're not a silver bullet but quite a few of them are most certainly effective.


  If you insist on using stimulants such as amphetamine,
  learn about how to reduce tolerance with things such as
  NMDA antagonists and take things to prevent neurotoxicity.

  He should also get on with trying Noopept - it's one of 
  the best nootropics of recent times.
Care to expound on those points?


Amphetamine neurotoxicity:

http://en.wikipedia.org/wiki/Amphetamine

>The increase of cytosolic dopamine appears to trigger neurotoxicity, as dopamine readily auto-oxidizes, so that amphetamine or methamphetamine's increase in cytosolic dopamine can lead to oxidative stress in the cytosol that in turn promotes autophagy-related degradation of dopamine axons and dendrites

NMDA antagonists such as magnesium (as mentioned by rms), zinc, DXM, MK-801, Memantine and Ketamine can reduce/prevent tolerance as well as nerotoxicity.

http://www.acnp.org/g4/gn401000166/ch162.htm

> Methamphetamine toxicity is inhibited by a variety of drug treatments, including: ... 3) NMDA receptor antagonists

http://www.socialanxietysupport.com/forum/f30/evidence-for-n...


Magnesium and blueberries seem to be the easily within reach stack to take with amphetamine.

Noopept is quite good. I stopped taking it because for me it seemed acutely enough anxiolytic to cause rebound anxiety.


Combine with L-Theanine (increases GABA activity), Lemon Balm (slows down GABA breakdown) and L-Tryptophan (good for social anxiety).


If Noopept causes rebound anxiety, isn't that a sign that it could build up resistance/dependence over time?


Are you intentionally classifying caffeine the same as amphetamines, or did you overlook it?


Intentional, I believe. It is habit forming and addictive, compared to say, taking Omega-3s before bed every day.


Just because something is physically addictive doesn't mean it has negative health consequences. Many studies have shown that long term, low to moderate caffeine consumption (especially coffee) has a positive impact on overall health.

Caffeine, amphetamines, and modafinil are not neurotoxic. I believe the others aren't either but I haven't read up on them. Methamphetamine is neurotoxic but non-methylated amphetamines are not.


The problem is that these studies are narrow and only look at a handful of the aspects of consumption.

http://en.wikipedia.org/wiki/Health_effects_of_caffeine

> Short term side effects such as headache, nausea, and anxiety have been shown as symptoms of mild caffeine consumption.

When assessing the health impact of a drug, it is important to look at the whole picture.

Amp is indeed neurotoxic, though the exact effect is a little nuanced and dose-dependent.

http://www.ncbi.nlm.nih.gov/pubmed/15019431

> In addition to the social, cultural and indirect medical complications of amphetamine analog abuse, this class of drugs is also known to have the potential to damage brain monoaminergic cells directly.

http://www.ncbi.nlm.nih.gov/pubmed/17606768

> In early studies, high doses of amphetamine, comparable to amounts used by addicts, were shown to damage dopaminergic pathways. More recent studies, using therapeutic regimens, appear contradictory. One paradigm shows significant decreases in striatal dopamine and transporter density after oral administration of "therapeutic" doses in primates. Another shows morphological evidence of "trophic" dendritic growth in the brains of adult and juvenile rats given systemic injections mimicking "therapeutic" treatment. Imaging studies of ADHD-diagnosed individuals show an increase in striatal dopamine transporter availability that may be reduced by methylphenidate treatment.

I did not claim that caffeine and modafinil are neurotoxic, just amphetamine.


The stranger point is including Modafinil, which is not addictive at all.


You do however build up a tolerance to modafinil, and it has various negative side-effects. But I agree that it's weird to make some sort of hard separation between various types of nootropics, and say that you should never touch some. Everything potentially has its use, and I think that Gwern covers this topic exceedingly well at his site.


I did not advocate staying away from them but using them sporadically as opposed to regularly:

> Stay away from the above, unless taken very occasionally


any particular reason to take them before bed? on my bottle of fish oil caps it says to take them with a meal?


In what way is modafinil not a nootropic?


I believe he's using modafanil in the sense of broadly neuroprotective.

However selegine in microdoses is a true neuroprotective in that sense..


Then he's using a strange and novel definition of 'nootropic'. 'Neuroprotective'? Heck, lithium is neuroprotective at some doses, but I don't see many people suggesting its use.


Modafinil has too many side effects to be considered nootropic. It is a cognitive enhancer.

http://en.wikipedia.org/wiki/Modafinil#Side_effects


Too true. Gwern's whole site is an awesome example of applied critical thinking.


I`m currently lost in that time sink. So many things to read and learn :)



Hollywood, please don't ever try to make live-action remakes of anime. Why can't they just show the anime itself? It was already good.

I highly recommend watching Death Note in its original anime form (preferably in Japanese with English subtitles). It's excellent. A little culture never hurt anyone. Every story doesn't have to take place in the U.S.


The anime form isn't the "original", but itself an adaptation of the manga. I don't have a huge problem in principle with media forms being adapted into other media forms, though, so that isn't really a knock against it. Imo whether an adaption is good is more interesting than the status as original/adaptation/remake. Some adaptations and remakes are good!


Yes, the original format is manga, and I have a few volumes -- the overlap is remarkable. As if Madhouse studios took the pictures, enhanced them, gave them a flow and breathed life into them with excellent voice acting and music.

The manga is outstanding because of the plot, but nothing beats the anime.


The extreme example is when the translation is actually better than the original, for a book.

Edgar Alan Poe's tales translated by Baudelaire are said to be one of such cases.


Maybe someone should publish a translation of Baudelaire's translation


My first reactions to this were "Thank god this abomination hasn't been green-lit" and "*@#& you Hollywood!"

Getting over that I can see how a looser translation could work. But don't call it Death Note. Calling it Death Note while whitewashing it and getting rid of many of the major plot characters and plot elements entirely is just asking for a colossal failure and fanbase revolt. It's like they didn't pay attention to the DBZ and Airbender fiascos.


The Japanese TV Drama industry is a much more frequent offender of such adaptations (as you may imagine, the outcome is almost always catastrophically bad).


I'll never understand the people that are hurt by changes to something that they could just avoid. Why does an American movie hurt it? Will it make you enjoy the original less? I'll also never understand the fascination with subtitles and the Japanese voice, which anime fans seem to think is more dramatic.

The people who read the original manga probably feel a certain way about the anime. The same way that people who read books feel about movie adaptions because they know there will be differences in it. A movie can bring things to a wider audience and then a lot of them may check out the original, which they may or may nor prefer.


>I'll never understand the people that are hurt by changes to something that they could just avoid. Why does an American movie hurt it? Will it make you enjoy the original less?

Because it spoils the 'reputation' of the original work? If the movie is done crappyly, then people will assume that the anime/manga will also be crappy.

>I'll also never understand the fascination with subtitles and the Japanese voice, which anime fans seem to think is more dramatic.

It's not that anime watchers hate english dubs. It's just that they hate crappy dubs. I watch a ton of anime and I usually prefer japanese with english subtitles, but sometimes the english dubs are better than the original and in that time, I'll stick with the english dub.

It's just a matter of which version has better quality voice actors and dialogues.


The usual claim is that it "ruins their childhood".


> Why can't they just show the anime itself?

The rationalization most moviegoers would use is one of legitimacy. Many people see film as the only "serious" medium for storytelling (live action TV is too cheap, books are too academic) and categorically dismiss newer media (video games, comics, etc.) as having no cultural value whatsoever. Thus, a good story is not actually good until it's a movie.

Hollywood loves this attitude and may be entirely to blame for it. It's hard to tell how many studio execs actually think this way though. I get the impression there are also a lot of people in Hollywood who don't believe in the concept of art.


Because an entire TV series is much longer than a movie.

* goes back to waiting for more of Rebuild *

Edit: Huh, if the last thing in a comment is an asterisk it becomes invisible.


i think The Matrix shows that it's at least possible to make a live action remake of an anime. i think what's actually needed is a director who understands the anime essense that well. eg the Wachowski bros

the Wachowski bros also directed Speed Racer (2008). it's one of my favorite movies and was 1,000,000x more awesome than it got credit for


Ah, Death Note, one of my favorite manga. I read it in the original Japanese and it was very thrilling. Unfortunately, the anime adaptation was rather poorly done. The second half, IIRC, really went off the rails.

Edit: the author of the linked post also has an analysis of Death Note Endings[0], which talks about some of the reasons for the manga and anime endings and the incentives for why they ended the ways they did.

0: http://www.gwern.net/Death%20Note%20Ending


Although it went a bit off the rails, I wouldn't say it was poorly done. As an anime compared to others it stands amongst the best in my opinion.

I agree though it is always a pity when something has the potential to be perfect, and some directorial decision is made to make a big dent in it (see almost every book to movie adaption).


> As an anime compared to others it stands amongst the best in my opinion.

Well, my opinion may have been tainted by having read the manga first and expecting too much faithfulness to it out of the anime adaptation - a rather common problem. And on the whole, it's definitely one of the better anime in recent years. Unfortunately, there has been a strong trend these days to go for the moƩ/fanservice crap, likely in an attempt to drive DVD/Bluray sales among the hardcore otaku in the face of declining mainstream viewers and ad revenues. I think I lost my patience somewhere around when Highschool of the Dead came out.

Nowadays I watch a lot more K-dramas, which have rapidly picked up in cinematography and acting over the last few years - they are definitely in a Golden Era right now, and the internet makes it so easy to rapidly obtain them, no matter where you are living. The K-drama fansubbing community is absolutely amazing, and it has been a TON of fun learning Korean through Japanese and seeing the similarities (and differences) between the languages.

But there's so much good anime from back in the day that doesn't get enough attention. Legend of the Galactic Heroes particularly stands out in my mind as a wonderful OVA (though it is quite atypical among OVAs, at 110 episodes). The use of a space opera setting as a backdrop for political commentary was very poignantly done. I've been meaning to read the novels the series is based on, but as science fiction novels are wont to do, they will no doubt be chock full of vocabulary that will probably require me to break out my Japanese-English dictionary.


Yeah if you're a drama fan I can imagine the past few years have not seen a good trend in Anime lately :P

I'm more of an action adventure kind of guy, there's not many good new stuff in that genre either as far as I'm aware. The most recent one I really enjoyed was Steins;Gate which is really quite brilliant!

Also I'm jealous you're able to read Japanese, I've not had the discipline to really learn it yet.


> I'm more of an action adventure kind of guy, there's not many good new stuff in that genre either as far as I'm aware.

Take a look at Gaksital (2012). I've not had a chance to watch it yet, but it's got action and is apparently quite good.

Another excellent action K-drama is City Hunter (2011), which is actually based on a late 80s/early 90s anime series, and has a corresponding anime as well that was first released in 1987. City Hunter was probably one of the first K-dramas where the production values and special effects really struck me in the face. Things have only continued to improve since then. Unfortunately, most J-drama producers haven't opted to continually increase production values in the same manner (and the Japanese wonder why K-dramas are so popular in Japan).

Another good K-drama in the last year (which I've seen) was Ghost, which (while it has some action) is more about crime and computers/hacking, but since you're on HN, you might like it. It had a surprisingly realistic portrayal of computers/hacking - not your usual CSI crap. Plus, the lead actress, Lee Yeon Hee, while still rather young/inexperienced in terms of acting, is one of the most beautiful (in a uniquely understated way) I've seen in K-dramaland in quite a while.

I certainly enjoy melodramas as well though, especially the makjang dramas[0] that Korea is famous for. Nice Guy (2012) is a thrilling example of this, and one of the best K-dramas overall from last year. And if you're interested in historical dramas, I hear Arang and the Magistrate is quite nice.

Dramabeans[1] is a good (and surprisingly professional) blog for keeping abreast of K-drama news in general. They have good year-end roundups[2] that allow you to quickly spot the good shows.

> The most recent one I really enjoyed was Steins;Gate which is really quite brilliant!

Yes, Steins;Gate has been a notable exception - I've been meaning to play the visual novel the anime is based on, but that would require me to boot into Windows :/

> Also I'm jealous you're able to read Japanese, I've not had the discipline to really learn it yet.

Learning how to read/write Japanese is all about making it fun. I used the standard RtK/Anki approach, but what really motivated me was using a brush instead of a normal pencil/pen. It really improves your handwriting and makes practicing a lot more fun. Once you've got all the jouyou kanji down, it's just a matter of picking up jukugo in context, which is inherently a lot more fun, if you can find interesting stuff to read and make Anki flashcards out of. I started out with lines from dramas (there are Japanese subtitles available on D-Addicts[3]), and then moved onto websites, before going to actual novels.

0: http://belectricground.com/2011/04/11/korean-language-and-cu...

1: http://www.dramabeans.com/

2: http://www.dramabeans.com/tag/year-in-review/

3: http://d-addicts.com/forum/subtitles.php


Have you seen the late J-dramas like 'kasefu no mita' or 'doctor X'? I find those to be pretty good entertainment, however I don't know K-dramas. What I like a lot about Japanese media is, they don't tend to stretch their stories out just for a quick buck. Series with massive success like kasefu don't get sequels if their story is done.


> Have you seen the late J-dramas like 'kasefu no mita' or 'doctor X'?

I've heard of both, but haven't seen either. I've been meaning to, but as I mentioned before[0], part of the reason I was watching J-dramas was for Japanese reading/writing practice, so my dorama watching has dropped off after passing that phase of my studies. Hopefully I'll have some time to watch them this year, as they seem decent enough.

> I find those to be pretty good entertainment, however I don't know K-dramas. What I like a lot about Japanese media is, they don't tend to stretch their stories out just for a quick buck. Series with massive success like kasefu don't get sequels if their story is done.

The same policy exists for K-dramas. Very few get sequels, and those that do have generally turned out quite well, such as 2011's Vampire Prosecutor[1]. Sequels seem to be more popular on (the recently created) Korean cable networks, but even then they're not the stretched-out 5- or 6-year sagas you regularly see on American TV.

In any case, I would highly recommend checking out K-dramas - I mentioned a few good ones in another comment[0]. The production values, cinematography, set design, and acting are generally far better than they are for J-dramas. I find the actresses significantly more attractive as well, which never hurts ;)

0: http://news.ycombinator.com/item?id=5011440

1: http://wiki.d-addicts.com/Vampire_Prosecutor


Interesting- I only watched the anime back in high school (darn, has it been that long?!), and absolutely loved it.

I just ordered the manga box set on Amazon :)


I enjoyed the original Japanese films more than the manga. Despite the mediocre acting, the movie was much better than the anime.


I've never seen the live action films, I'll have to check it out. I see that Toda Erika, one of my favorite Japanese actresses, stars as Amane Misa. Apparently it was her debut film, so it'll be interesting to see how her acting back then compares to now (I just started watching the spring 2012 dorama Kagi no Kakatta Heya yesterday, in which she isn't all that bad).


Agreed. The films ended far more sensibly, and the plot was better thought out.


But keep Death Note III away yourself. It's just bad, like really bad.


I will pretend that I never saw someone mention this. I had never heard of this movie, and still haven't.


I liked the way they resolved the conflict, but it wasn't enough to make up for the terrible acting.


[SPOLIERS]

> Unfortunately, the anime adaptation was rather poorly done. The second half, IIRC, really went off the rails.

I think(for me) it went down the moment L dies. Is the manga ending good?


It's identical. The only real change the anime did was to double the pace for the second half due to how much worse it is.

The entire second half was tacked on at the request of the publisher due to how popular the series was, so it's not surprising that it didn't work all that well.


Funny thing about Death Note. My friend who is big into anime recommended this as my 'first time' anime I should watch. To make a long story (very short), I gave him shit about anime for a good 2 years if not longer. It was all light-hearted but eventually he snapped, he just couldn't take it anymore.

"FINE!!! F&#$&$ YOU. DONT WATCH IT. SEE IF I CARE". I can see 2 years of pushing his buttons and joking finally caught up with him.

I watched it, turns out I loved it more than I could have imagined. We tease each other about that to this day.


A Death Note fan would write an analysis this detailed.


In a barefooted couch hanging sugar rush no doubt!


While fixing up the final errors the other night, I consumed a bag of Gummi Bears.

And half a box of chocolate.


Oh? Do you have a stochastic analysis? ;)


- Rob Malda joines HN... Check.

- anti-ms league attempts to rewrite history... Check.

- Anime on the front page of HN... Check.

Fears of HN becoming reddit are overrated, but the chance of becoming slashdot instead looks to be > 0.


A detailed, exacting analysis, worthy of L himself.


Awesome Article (and site)! Just one correction, look like you used the word memsahib in correctly (I think its a respectable salutation for a woman)..


Looks like an analysis someone from the Singularity Institute would write...


That's an impressive analysis - very interesting!


I get the strong feeling I've read this article before, but I can't find a strong indicator this is an old article. Did someone else write about the leaked script?


You may've read the much shorter original version back when it lived in http://www.gwern.net/Notes or read it sometime in the past 3-4 months when I was working on expanding it.

(The first version was basically the non-quantitive first third, but about half that size, so I added a ton.)


It's a bit hard to find, but in the left sidebar there's a date string "2 Nov 2009". This would match up to the article's timeline; script appears on the Internet in May, add a few months for the author to discover it and perform extensive research. Inspecting the HTML verifies that that date is probably when the article was first published:

  <meta name="dc.date.issued" content="2 Nov 2009"/>


This particular post is a new version of an older article, not a completely new article, so it seems mostly likely that you read the old version already.


Ah ok that makes sense then :)


The dates on the sidebar indicate it was first published in 2009 (and last revised today). Most of gwern's site is an indefinite "work in progress".


I've seen the Death Note anime twice and it is one of the best Animes, with probably only Code Geass above it. I would love a real-life make of it but I believe that is impossible because everything about Death Note can only be captured in a lengthy series. In order to feel Death Note and truly understand the genius behind it, you must submerge yourself into its world for an extended period of time.


"probability a fake script will have copies taken down = 5%+5% = 0.10"

So the probability that the automaton who combs file sharing sites for "Warner Bros" and spams DMCA requests would both: A. be in a position to know whether or not the script was real, and B. would refrain from issuing a takedown if it were not is 90%?

This does not seem to fit historical evidence.


"how many unproduced screenplays get leaked?"

Pretty much all of them, to a greater or lesser extent.


Just an observation:

Political threads usually get several comments about being off topic for HN, some times quite strong. However, not one descenting comment for this, which to me seems equally, if not more, off topic.

I personally really don't mind either subject, but how is this considered more relevant to HN than politics? Seems people are just as interested.

I don't think this is a problem with this topic, even as a non fan its an interesting discussion. Its more interesting to me that the group reaction is different.


Even if the topic is not something you care about the techniques and methods gwern uses are not often talked about here and useful for a variety of text processing tasks.

His methodical analysis and scientific rigor is always what I wish other posts could have. If nothing this is a post which will motivate others to write better.

The cluster analysis is cool - not seen that before. Wonder what the results of Latent Dirichlet Allocation might have gotten...


This is an article about Bayesian statistics, which is usually considered as quite hip in these places. You do remember who wrote the "A Plan for Spam"-paper, after all?


This is the most intense fan fiction.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact

Search: