12 réponses à ce sujet

[Lotion Soronarr]

Do you know what makes great quests and decisions less great? Predictability. Peeking at the guide and doing a second playtrough and all the mistery is gone.

What I propose is to have consequences be somewhat random (when appropriate of course).

Let me give an example:

REDCLIFFE.

You got 3 choices of which one (getting help from mages) is superior because there is no danger in it. You KNOW nothing bad will happen when you leave. Something that is a risk, a chance, ceases to be.

But what if you didn't know? Waht if - no matter how many uides you read, how many times you play - you can never be certain that everything will be OK once you get back?


Basicly, when approprite - usually when events are outside of  player character control - have the outcome randomized to a point.

Now, if you had a great victory at Redcliffe and prepared everyone, chance of something bad happening is 40%
If you had a good victory, chances are 50%
If you did bad, chances are 60%

If the bad consequence happens, even the extent can be randomized.
If all knigts survived - high chances of Teagen being alive, but some knights died protecting him (for example)



This makes unpredictable things unpredictable. Taking a risk is always a risk - you can increase your chances, you can take some precautions to reduce the fallout - but you will never KNOW. You will enver be sure.


***

Now he question is - what is stopping the player from reloading?

That depends on whow it's done. When is the roll determined? Once rolled, does it become static?

So let's say the roll is made the second you head out towards the Circle Tower.

Let's also say it's not static. You'll have to re-do the Broken Circle quest (unless you did that already) and travel there and back  for a CHANCE to change the outcome.

Let's say it is static. Roll sez Teagan dies. Too bad. No re-loading will help you now, Teagan dies.

Unless maybe, you load an even older save and change some of the other variables (like how well the Redcliffe milita is equipped). In this case, a roll would be static UNLESS some variables change. In which case it would be rolled again an then become static again.
Which would make it both fair and at the same time incredibly frustrating to save-scum.

[Allan Schumacher]

Just to continue discussion: Why shouldn't it be deterministic?

Why would actions play out differently in the event of not being present?

[Allan Schumacher]

General User wrote...

Allan Schumacher wrote...

Just to continue discussion: Why shouldn't it be deterministic?

Why would actions play out differently in the event of not being present?

I see this idea more in terms of being a way to bridge the gap between the player and the player character.  In other words, if the player character doesn't know how any given event or mission is going to resolve itself, why should the player?

Plus there's the idea that the player/player character might be able to influence the outcome of a given mission, but ultimately they shouldn't be able to determine it.  Afterall, we human beings can only influence the actions of others to greater or lesser degrees, why should our digital avatars in Thedas be any different?

Who's to say they are determining it?  Simply because that's the consequence of all those preconditions being met?

Lets say there are 3 different preconditions that can be met, and each permutation is a unique outcome (so 6 outcomes).  If I satisfy the first 2, I have put my influence into the equation.  Everything being equal, why would the offscreen actors allow my influence to affect them one time, but in the other time, my exact same influence does NOT affect them?

Modifié par Allan Schumacher, 29 octobre 2012 - 07:48 .

[Allan Schumacher]

In a way... yes. When there's only one character (the player character) whose takes action(s) to meet the preconditions that inevitably lead to any given consequence, I don't see how anyone could say that they aren't determining it.

Keep in mind the only perspective you're shown is the players.

If you tell me to be careful the roads are dangerous and don't interact with me again, and I drive more carefully, doesn't mean that you alone determined my behaviour. Who's to say that after trying the King to not execute a prisoner, that after you leave the Queen echoes what you've been saying a dozen times over to show the King the error of his ways.

Because the fact is everything isn't equal (or it shouldn't be anyway). To stick with the Redcliffe scenario, perhaps (after the Warden left for the Circle) Bann Teagan decided to assign Guardsman A instead of Guardsman B to watch the door Connor locked himself behind, and Guardsman A makes decisions or has vulnerabilites that Guardsman B wouldn't have.

What would motivate Teagan to make this decision differently? If it's a dice role, then his decision starts to become arbitrary, which is maybe okay if Teagan is considered an arbitrary character.

Modifié par Allan Schumacher, 29 octobre 2012 - 09:29 .

[Allan Schumacher]

Lennard Testarossa wrote...

Reality is a very complex system in which the tiniest of changes in the initial conditions can have a huge impact on the outcome. Reality very often seems arbitrary. Having that in a game does not harm the game's credibilty in any way.

That reality may seem arbitrary (a notion I'm not sure I agree with) doesn't mean that it necessarily is.  It just means that you may not know all the influences.

For something to happen arbitrarily, it needs to happen truly randomly, or perhaps more clearly: without reason.


What I've been getting at is, why would the events play out differently all else being equal?  If you have set things up so that there's 10 guards to defend the king, why would they make different decisions (right down to their attacks and tactics) without some external force (player or otherwise) influencing those decisions.

The problem with comparing this to real life (since I've seen other also state that this is "more realistic") is that we never get the opportunity to completely replay out reality the same way that the characters in a video game do.

If you look at my history, I made decisions and I could have made different ones along the way.  But I didn't, I made the decisions that I actually did.  Why would I make different decisions in my past in the event that nothing of my past is presented differently in any way?

[Allan Schumacher]

And creating Always-Win scenarios.

Cake or Death isn't an interesting choice, but that doesn't mean it always has to be Death or Death. It can be Cake or Cake too.

Agreed, choice of equivalent value.

Although, even in the case of "Cake of Cake" there's still going to be some level of opportunity cost. I.E. Getting the sweet two-hander as a reward, or getting the sweet crossbow. You still don't get both.

[Allan Schumacher]

Lennard Testarossa wrote...

Allan Schumacher wrote...
What I've been getting at is, why would the events play out differently all else being equal? I

But everything else isn't equal. On every single playthrough, things will be a little different. Maybe you took three minutes longer than on your last playthrough to get to Connor. The randomization would simulate such small differences.

It actually wouldn't, because it'd be random and arbitrary.  If there's a reason for things to play out differently due to time differences, why should it not be accounted for more deterministically?  The next time I play I take 5 minutes longer, but now I get a different (perhaps even more ideal!) solution?  How does that make sense?

You're getting into one heck of a butterfly effect type of argument here, and if randomization is simulating the small differences, why shouldn't it be reproducible?  Unless you're getting to the idea that a player using a sword instead of a mace is sufficient to alter the outcomes where the player isn't even present in the future.

So you'd still have to convince me that the NPCs are going to make choices that are signficantly different enough to actually result in different decisions that alter the outcomes of situations that the player cannot observe nor influence.

Because you took 3 minutes longer (or two minutes less, or 15 seconds longer, or 30 hours longer) there will be an arbitrary force influencing how other characters make their decisions.

[Allan Schumacher]

Minute decisions made in combat by unnamed characters, even the somewhat comical truth that sometimes things happen to Person A and not Person B just because A happened to be standing on the left and not the right.

I agree! My question is: "Should we expect that thing to NOT happen to Person A if we were to replay the situation?"



The impression I am getting is that people are wanting the NPCs to be bound by the same dice rolls the player is when the player is making decisions. Is this somewhat accurate at least?

[Allan Schumacher]

I would say "no, we shouln't." Because, even with dedicated conscious effort, it is all but impossible to precisely replay any given situation. Perhaps you attack the darkspawn on the left with your bow and arrows vs hitting the darkspawn on the right with a sword. Things like that.

Why would the person decide to attack stuff in a different order with no outside influences affecting his decision making process? Do you make the decisions that you make "randomly." Yes, you can choose to attack the guy on the left or the guy on the right. But whichever one you choose is the one that you choose. Is there any reason to assume that your choice was random and not the sum of conscious and subconscious thought?

Simply because an action is not predictable, however, doesn't mean it is not deterministic. I understand what you're saying, as it goes back to Chaos Theory. So yeah, it can make sense for slight differences in initial conditions to result in different outcomes, but not in a random way.


And this all then boils back to: "Does this actually make a superior product?"

[Allan Schumacher]

Note, I just wanted to clarify why I was contributing to this thread.

This is a philosophical debate, and any that state one way or another better serves "realism" will find such a position is not certain (at least at this time).

I was discussing the topic from the school of thought of biological determinism and the general idea that free will is an illusion. If a person fires an arrow at the guy on the left is because all the conscious and subconscious effects on said person led to that decision being made.

There are a lot of other schools of thought on how we make our decisions. An advantage of the computer simulation is that we get exact replicas of said situations. It's not unreasonable for someone to take this information and have a set of assumptions that goes along with it. They aren't wrong (or even right) for believing this way.


In terms of whether or not this actually provides for an interesting game mechanic is not something I entirely agree with. I disagree with the notion that it makes the games more replayable, because for some gamers part of what they enjoy about replayability is going through and making different choices and seeing how the story/narrative plays out from there. So while this might be a "very good idea" from one perspective, it may not be from another.


From a QA perspective, random chance starts to become a nightmare to verify. I'm not a huge fan of it because it's not a situation where you can quickly set up a unit test to verify that the system is working as expected (automated testing for plots is mindcrushingly frustrating from a maintenance perspective). The problem is exacerbated the more complicated the system becomes.


I have to go, but I'll check back on the thread when I get back from Volleyball.

Modifié par Allan Schumacher, 29 octobre 2012 - 11:59 .

[Allan Schumacher]

I'd say it would. The idea of a video game that reflects the reality that many persons and events truely are unpredictable (no matter how or why they do the things they do) is something that would be very interesting to me.

That people are unpredictable does not mean they are not deterministic.


I wouldn't get too hung up on the word "chance" since anything arguably has a chance, even if not explicitly stated. To solve this conundrum all you need are some of the "high risk" situations to not work out very well. If the player gets the impression that the chance may not actually work, suddenly those options become more interesting.

[Allan Schumacher]

Which is admirable, however this also means that there will never be a
'wrong decision for the right reasons' thrown out again. Harrowmont is a
perfect example of this - the guy who isn't a murdering, lying,
conniving bastard is actually the worst guy for the job. Go figure.

I'm not sure why you feel this can never happen again.

[Allan Schumacher]

It sounds fun to me. You want ASSURANCE. You want guarantees. Which is something that doesn't exist in real life. It is that that gives weight to many of our choices.

You also have a notion that "doing everything right" MUST guarantee a super outcome. Which btw, isn't true in Real Life either.
Of course, given the outcry ME3 had for not allowing everyone their super-duper outcomes annd blue babies, I can't say I'm surprised by the responses.

The realism argument is irrelevant. What you describe isn't realism. The only guarantees come from metaknowledge, which is irrelevant to the game world. The only reason why you "knew" that going to the circle the first time would be ideal is based on prior experiences with other games.

Also, if you bothered reading - which I can only assume you didn't

The attitude isn't necessary.

However, you have control over the state of Redcliffes defenses.

So regardless if A or B happens, you can have damage migation and some buffer.



So let's say Connor does flip out.

If you posted guards, equipped the milita, made every preparations
before departing - Connor gets contained and only a few soldiers die.

Aha! So what this really is, is a request for more content. This is a very different type of request. Furthermore, the situation you describe is still something that can be done deterministically too.