Hacker News Reader

Comment by naillo | Hacker News Reader

Devils advocate but not sure how not allowing reusing mechanics leads to killing off novelty. If anything not allowing resuing mechanics should lead to developers being forced to invent new non cookie cutter mechanics to get around the patent.

View on HN

It prevents new games from coming up with novel application/combinations of these mechanics.

For instance if Minecraft or any of the multide of crafting games had a patent on sticking together objects we wouldn't have BoTWToTK in the first place.

The example I go to is the Nemesis system of the Mordor games. Such a cool system for that type of game, and it's easy to imagine ways to use it in a ton of similar games.

My understanding is that it's patented, though. So no one else is making anything like it. No other experiments about what else those mechanics can do.

I'm not sure if even that Nemesis system is really something completely new. You could probably get a similar emergent behaviour from older systems like goal-oriented action planning, and I wouldn't be surprised if games like Dwarf Fortress already had something similar.

That's probably true, but the specific implementation in the Mordor games is pretty unique. It's not just emergent gameplay, there's elements of deliberate action in there. The game selects certain enemies to become "resurrected" to come back stronger and become your longterm rivals.

The enemies are also generated with perks and flaws based on the available game mechanics. These orcs return to fight you again, they often have upgrades themselves to resist or be immune to the ways you killed them previously.

Taken individually none of these mechanics are novel I don't think. But this particular implementation is pretty unique. I'd love to see similar things in other games. Like Mech pilot rivalries in a Mecha game or something similar. But games aren't copying it, despite how well received it was. I can only think it's due to some kind of patent licensing issue.

One of the techniques when you file a patent is to come up with as many variations of an idea as possible. You don't even need to have built those ideas into an implementation. So maybe I say "This patent is for playing a sound right when a button is pressed" but I expand it to "and right before and right after, based on some cadence, or a configuration", etc etc etc. This makes patents both precise and very broad, leaving little room for someone to go "ah but I did it differently".

Couldn’t this be fixed by requiring an implementation to patent?

I think that would have its own issues, such as creating a significant burden on both the person filing (who the system is designed to protect) as well as the patent office having to verify the implementation. I think it would be very complex.

Really, patents should just expire after 5 years by default, possibly with exceptional grants for 10 years. I think 5-10 years is plenty of time to establish yourself in a market, and if at that point your competitors start competing, well, you had your grace period.

No, because one of the traditional use-cases why patents were made is so that an inventor of a novel solution can get investment to get that invention actually implemented without risking that the investors will just steal the invention; or that the inventor can contract a manufacturing company to make the prototype without risking that they will just include the invention in their own products.

So being able to secure a patent before its first implementation is made is pretty much a non-negotiable requirement.

Patents are often not directly exploited by the holder; instead, the holder sits and waits decfor someone else to build a working tool, then sues.

So, yes, folks should be able to secure a parent before they're able to build and sell a working product, _but_ it should be a requirement that they build and sell within a shorter time-frame than the length of the patent, e.g. the patent is for X years and I have to build and sell within 0.2*X years or the patent is rejected.

It seems to make sense, but is problematic in practice - for example, for pharmaceuticals it might take half of the total patent time until you are able to sell a product based on the patent, as getting FDA approval could take something like 8 years if they ask for more trials.

Also, patents make sense also for things that are not for sale to general public - for example, an invention to improve some machines that you use in your factory, but aren't selling to other factories i.e. your competitors; consumer products are the thing that is visible, but B2B products matter even more.

It could also be fixed by making the penalty for filing an invalid patent prohibitively high. If someone proves prior art, you have to pay them "a mountain of cash". That way, there's A) a real risk in making your patent any broader than it needs to be and B) an incentive for law firms specialize in killing patents for the reward payout.

That's going to make it way easier for large companies who can afford a team of lawyers to hunt through potential infringements (and defend them). The cost of filing a patent is already really high (1000s, 10s of thosands) even without that.

A game consists of many mechanics combined. A great game needs all of them to work and licensing deals have very high transaction costs. You also have game mechanics that only work in the presence of other game mechanics, so patenting these can actually _decrease_ novel game mechanics.

No game mechanic in Minecraft, Dark Souls, Baldur's Gate 3, etc is new. Their success comes from a well-executed mix of known features. Trying to create a game with a new mechanic is like writing a story with an unknown plot device: theoretically possible but quite difficult and the result likely won't be any better in terms of quality.

Try and name just one game released in the last decade with a completely novel game mechanic.

Then I doubt you would approve as a novel game mechanic anything that Nintendo attempts to patent within that claim. An exemplary new mechanic recent decade game is Baba Is You. It isn't obvious. If it was, someone would come up with Baba Is You pretty soon after Sokoban was made 40 years ago.

I also don't think nintendo should be granted the patents it is seeking, other games have done similar enough before that it feels wrong to give it to the bigger company because they asked.

Oh, Baba Is You is a good example, I'd say that counts.

"flOw" certainly seemed novel (and really nice) when it was released back in 2006 (not the last decade, obviously):

https://www.youtube.com/watch?v=tTVDSOnPLns

It's a pity nothing else since seems to use the same mechanic / concept.

Well written.

Reply on news.ycombinator.com