Patterning

Jun. 24th, 2019 08:49 pm
migmit: (Default)
[personal profile] migmit
You know, it seems like this thing is repeated often enough — not sure if anybody noticed this before.

Here it is: let's look at The Matrix. What we have here is a very solid first movie — that could just stop there, it did not require sequels. Then there was a second movie, which would be pretty forgettable, if it didn't have a very well done action sequence — the highway chase (and I don't mean Neo's antics going in parallel to that, those were totally pointless). And then there was a third one, which is a total mess that looks like it was drunkenly assembled from various parts pulled out of a recycle bin — and I don't mean that icon on the Windows desktop.

Then there are Pirates of the Caribbean. A very solid first movie, that could stop just there and did not require sequels. Then a second movie, which would be pretty forgettable if it didn't have a very well done action sequence — this time a three-way duel. And then — a third one, which is a recycled mess again. They didn't stop there either, but everything else is more or less an official fanfic.

And then there is How to Train Your Dragon. A very solid first movie that did not — you guessed it — require sequels. Then a second movie, which, again would be forgettable without an action sequence — Drago's attack on the Dragon Island. And, capping it, a recycled mess.

So, is it some law about that? Or a mind-controlling alien switching between different directors and making them repeat the same structure? What's going on?

Date: 2019-06-24 08:55 pm (UTC)
thedeemon: (Default)
From: [personal profile] thedeemon
Thank goodness, Back To The Future did not follow this pattern. The 2nd movie was the best!
Terminator 2 was also not worse than the 1st. Plenty of other counterexamples.

Once you've got a pattern you can fit a lot of entities in it, having a lot of source data.

Date: 2019-06-25 09:03 am (UTC)
thedeemon: (Default)
From: [personal profile] thedeemon
Yep, that's what I'm saying. You pick a pattern, like "n and n+2 are prime", you take a large amount of source data (like "all natural numbers") and you see a lot of cases where this pattern works, and a lot of cases where it doesn't. Nothing surprising.