Timed puzzle

[Prozail]

Quick question.

How much time do you guys think is fair to give the player to act in a timed puzzle?
I mean the time from his GUI gets activated, to the time where he actually dies, if he doesn't do a specific thing.

[DoorKnobHandle]

You know, something tells me this has to with... I don't know... the actual puzzle?

:p

[Prozail]

Hehe maybe..

Since the rest of the game is pretty much about repeating scenes until he gets it right.. this particular one is not "virtual" so to speak, and will actually kill him.
So its more about the time it takes for him to realize that his GUI is back and he can actually do something..
I started out with 5 sec... that was to short.. now im up to 10.. and it's still short.. but any longer that that makes me feel like
the bad guy is actually waiting for the player to do something..

[Crimson Wizard]

Being more a player than a game creator I find this even more important .
According to what you say it is mainly a problem of making player be ready in time. I may suggest using some visual and/or audio signal, like screen border flashing accompanied by short sound. Or maybe flashy cursor animation.

[DoorKnobHandle]

What I was trying to say is: what does the player have to do to survive? Is it just moving somewhere? Or a complex inventory item or hotspot interaction? Without knowing, there's no way to give a good time estimate.

[Prozail]

If you know the solution it's just one action.

It's basically a hotspot interaction.
I'm using the 9-verbs interface so its just one action (two clicks)

I've decided to extend the time a bit and keep the badguy doing some SayBackground-ranting about how good he is... (the downfall of every badguy.. )

[Miguel R. Fervenza]

Actually I don't think kill the player is a good thing in adventure games, I hate that. You have to load a game and learn of a death, which is a nonsense in adventure context, even after several deaths, the emergency or the fear are lost because the player knows he can load until he can overcome the situation. You can die in Super Meat Boy, there's no problem, is part of game dynamic, you restart immediately and try again; but the adventure is a special genre, has an enormous narrative charge, you have to know the environment, peculiarities of characters, the universe... and think about the way to advance in story, that's the key, no ability, if you have to restart for learn to a death is like in a film the director decide to include several bad takes before the good one, disastrous for rhythm and a bad narrative solution.

[Snarky]

I'm not convinced a timed puzzle that is just "do one particular action" is a very good one (because you either know it and it's trivial, or you don't and you die), but it's hard to say without knowing more details.

Ron Gilbert once wrote about timed puzzles (something like) that you should aim for 10% of players solving it easily with time to spare, 10% of players failing, and 80% of players feeling like they just made it. To achieve that you can use a bunch of tricks, like escalating warnings, elastic time (depending on the puzzle progress, or a "countdown" that goes fast at first but then more slowly, or just giving you a couple of extensions if you don't solve it by the initial deadline), a "grace period" (so you have a second or so to finish it even when it seems like you've failed), and of course playing the "time's up" animation as soon as the puzzle is completed (e.g. the rope bridge that snaps as soon as you've made it across). You can also put in a simple puzzle that gives the player more time at the last second (e.g. in Resonance where a monster is breaking into your bedroom, you should lock the door, and the game gives you a reminder of that when it's almost too late).