Can you perfect chain without learning levels?

[nullpointer]

Hi all,

Ok lets explain what i mean. Obviously you can't humanly chain a perfect run of ikaruga without knowlege of the level design and patterns.. Only through practice and strategy can you work out a route and order to run a perfect chain..

However, imagine if you could pause the game at any point you like, or move in ultra slow motion, so you can always react in time to whatever you see on screen. Would it then be possible to chain a perfect level? I know this is an odd question (and purely hypothetical) but could a computer keep a perfect chain going just by analysing what is onscreen, without knowing what is coming up?

In game theory it kind of related to the idea of games with perfect information or imperfect information and how far into the future needs to be known to perform perfectly in the present..

Sorry if this sound a bit academic (but i guess it is!), just wondered what people think

[ShmupSamurai]

Hmmm...Not sure.

Personally it depends on your hand/eye coordination and reflexes I'd say.

[Plasmo]

Personally it depends on your hand/eye coordination and reflexes I'd say.
Hmmm...Not sure.

Personally it depends on your hand/eye coordination and reflexes I'd say.

That's exactly what is NOT present if the game would run in ultra slow-mo or is even stopped at all.

In Ikaruga it would theoretically be possible, you could take your time and figure out the patterns. In DDP however it would NOT be possible, because you need to know what comes next onto the screen.
Of course you wouldn't be able to do a perfect chain in Ikaruga, i.e. getting ALL chains, but you would do a lot than in DDP.

[ave]

If Ikaruga ran in slow motion, it would be of no need if the ship also just as slow. If you didn't know the stages, you would be maybe on top of the screen, forgetting chains that pop up near the bottom and you can't get down there in time to get them.
If the game would slow down massively (maybe 5fps) but the ship would move in 60fps, I think a very good run would be possible without too much knowledge of the game. Such an experiment would require an experienced player though since ship movements and and color switching stuff is hard to apprehend at the beginning, esp. for a newbie.

[Slump]

Uh, maybe I'm off base here but wouldn't that be highly dependent on the chain system of the game? Regardless though, I'd say for the vast majority of shmups out there it's a no. You need to learn levels, especially enemy patterns, before you can perfect chain in most cases. And slowing down the game could certainly work to your advantage but even then I'd say it would work to your advantage in learning the levels, not necessarily for keeping your chain.

**Edit**
Also, I think it would be perfectly possible for a computer to perfectly chain by only analyzing whats on the screen if the odds were tipped in it's favor (like as you mentioned slowing the game down vs. ship speed). Still would require eloquent code I'd imagine.

[ShmupSamurai]

Agreed..

[Jeneki]

Also consider medal chaining games such as Battle Garegga or Raiden Fighters series. Sure, slow-mo and pause would help you collect all the medals that you see on the screen. To maximize the value of those medals though, you really need to understand the game mechanics of increasing medal values, and which enemies drop them.

[Udderdude]

You could probably do a "perfect chain" if the time for the chain dropping was increased. Of course that's not going to happen unless you're playing one of those games that are really lenient with the chain dropping.

[kengou]

I get the question you're asking from a philosophy perspective, although I don't know if it really applies to shmups in the way you want it to. I doubt any human could realistically perfectly chain Ikaruga on his first time through no matter what speed the game ran at. There are several very confusing enemy patterns that take time to be worked out, and just pausing the game and thinking about them wouldn't suffice because you really need to see the enemies in motion to figure out their patterns and how to chain them. Particularly I'm thinking of 2-1 or 4-2. However, I do believe that it is POSSIBLE, even if nobody could realistically do it. Perhaps if you got the smartest, quickest conceivable person (or even AI) to play, they could very quickly figure out exactly how to chain every stage. I don't know, really. For something like DDP with time-based chaining, I doubt it is possible at all. You really need to know what is coming up next in order to know your timings and delays and everything.

And as I understand it, Game Theory is all about the best way to act based on imperfect information about the future, isn't it? If you know what will happen in the future, game theory isn't really much help, it would be pretty much indistinguishable from subjectivism/egoism.

[8 1/2]

Dodonpachi Stage 5 and 6. So long as you stay alive, you'll probably come close to chaining the whole stage whether you try or not.

[Keade]

It's not possible with most games, some games being "worse" than others in that registry.
I think a fair balance is what makes the game fun.

Dimahoo, for instance, requires just too much planning before a single (efficient) run for me - I still wonder how the top players play that game. Plan everything beforehand and have to restart each time they miss an important item ? Or reprocess their "route" while in-game when they do minor mistakes ? That seems to require a LOT of thought, something I could never pull off ingame (maybe if I knew the game SO well that I could have a mental map of every incoming item in the level and the next ones ? That sounds like a whole lot of information).
I'd go for the "plan everything beforehand" but maybe I'm underestimating the players.

[ncp]

Dodonpachi Stage 5 and 6. So long as you stay alive, you'll probably come close to chaining the whole stage whether you try or not.

No you won't.

edit: 2-5 and 2-6 you probably would, though.

[Keade]

More ennemies in the 2nd loop ? I thought there were only more bullets.

[ncp]

chaining bar is a lot longer