Patrol Routes in FMs

[grayman]

Besides, it's not like AI sneak up on you...part of the reason they have loud footsteps is so that players can hear them coming and prepare for it.

 

Though it would be interesting to have an AI sneak up behind you and stand there. You turn around, he says, "Lookin' for something'?", and then he clobbers you.

 

We still to this day retain the original FPS behavior where melee enemies run to you when they see you. Something I'd like to work on at some point is to have more collaboration, or at least some more intelligence. A guard spots you crossing a balcony, and he knows where the balcony stairs come out, so he runs to a spot just around the corner from there and waits for you. Since the player can always walk around a corner and come upon a guard, I don't think this new guard behavior would make the game more difficult. To give something back to the player, perhaps the guard could yell, which alerts the player. The guard disappears, and the player spends the next few moments wondering where he went. Is he coming for me? Is he waiting somewhere? Did he run off to sound the alarm?

[Sotha]

@radiator:

I find it very fascinating how the community is divided on many aspects of sneaking gaming.

 

I'd say it is not about the guards being stupid and the player being clever. It is about the player doing his best to avoid an intelligent and powerful opposition. It is more satisfying to beat a clever opponent rather than a stupid one.

 

@grayman:

is he coming for me? Is he waiting somewhere? Did he run off to sound the alarm?

 

That would be most unnerving and awesome! Will he find more guards, which intelligently come in the location he saw me from multiple directions. I love those rare moments in a game when I suddenly lose track of what the AI is doing and I fear I'll get suddenly attacked from behind.

[Springheel]

Something I'd like to work on at some point is to have more collaboration, or at least some more intelligence. A guard spots you crossing a balcony, and he knows where the balcony stairs come out, so he runs to a spot just around the corner from there and waits for you.

 

Yes, back in the early days of development we had talked about behaviour like this for elite guards. They would move more quietly and wait around corners and things to try and ambush you. I think that would be great.

 

It is more satisfying to beat a clever opponent rather than a stupid one.

QFT.

[Professor Paul1290]

I prefer dynamic patrols where they make sense. Obviously guards assigned to certain places aren't going to deviate very much, but not everyone would be be tied down as such, like guards on break and random civilians.

 

I think there is an issue in that many players won't be expecting the NPCs to deviate from their routines, but if it is eventually present in enough FMs then players will learn to prepare for it.

Personally I think it would be nice if something was there to pressure players into trying to leave themselves more room for the unexpected whenever possible. When you have to "prepare for the unexpected" it can lead to a very different style of play.

 

I think it's worth mentioning that many missions in tactical shooters like Rainbow Six 3: Raven Shield, SWAT 4, and the ArmA series benefit from a LOT of replayability just from semi-random placement and behavior of hostiles. In these games it works very well because individual enemies have such a significant effect on the player. A single bad guy in the right place can cause a lot of damage if the player isn't careful. Because of this the player must always leave themselves room for error whenever they can and maintain "security" as best they can in case of hostiles coming from unexpected directions and this leads to much more cautious and tense gameplay.

 

While it's a completely different genre, The Dark Mod presents some similar conditions such that I think it could also benefit from semi-random placement or movement of NPCs. Like a tactical shooter, a stealth game like The Dark Mod has enemies that have a significant effect on how the player approaches a situation just by where they are, and again a single enemy in the right place can make things very difficult for the player if the player doesn't prepare for them. If there exists the possibility of NPCs coming from unexpected directions then that would make things very tense for the player and cause them to adopt a much more cautious style of play. Devices like mines, noise arrows, and perhaps even random junk items that make sound when stepped on would probably see much more frequent use. Player will start to pay more attention to whether they leave certain doors open or closed, how they "inspect" rooms they cannot see through sound and other means, and other such subtle details.

 

I suspect that if dynamic patrols became more common in The Dark Mod then it would spark a lot of deep discussions of tactics among players, especially regarding how to deal with such unknowns. Should you leave doors behind you open so you can better hear or see approaching guards or should you close them to keep concealment? When should you break away from lockpicking? How can you best locate a guard you cannot directly see? How can you best place small noise making object mines in a given common hallway configuration to ensure maximum safety for yourself before proceeding? Dealing with an unpredictable threat can cause your tactics to become much more sophisticated than they would be otherwise.

 

 

On the other hand, while having more dynamic patrols would change the gameplay in a way that I would personally find very appealing, I'm guessing not everyone would agree with me.

I'm the sort that would take failure due to a random guard walking in on me not as unfair but as a sign that I need to improve my tactics to minimize the probability of such failure. Not everyone is going to take the random guard walking in on them quite as well.

Edited July 14, 2011 by Professor Paul1290

[Baddcog]

To play devil's advocate, I don't think Radiator was implying guards should be dumbed down.

 

Rather pointing out that the guards are generally 'simpleton's' compared to a sly master thief. Thus the reason for the entire game. Sneak in, outsmart 20 guards, get out alive.

 

Also the point about patrols being static is that people generally settle into a routine as others have mentioned in the thread.

I do it too, I wake up generally about the same time everyday, get something to drink, drive or ride my bike a certian path to my destination....

 

And I can see a guard doing that at work too. But that's the base patrol, going room to room in a certian order. Even a static patrol doesn't have to be an exact 10 second stop at this corner...

But that's the beauty of TDM's path system. You can set up a static patrol. then at certian spots have one branch out to one point that makes him stall 10 seconds, and that part is completely random, yet the entire patrol route is static.

That mixes it up while giving player a very good idea of what that guards routine is.

-----------------

Still on the whole does it really matter either way? Of course to the author it's always:

"Oh the player is gonna be amazed at this awesome random patrol I set up. They'll be blown away when the guard picks up that apple and takes a bite."

 

But truth is in game unless they hide in the kitchen and watch the guard patrol through several times they won't see a pattern or break in the pattern. So whether or not it's random they won't know unless his patrol routine is very small. Say a cook who never leaves the kitchen, but does a lot of random stuff.

 

If it's a long patrol it's almost a waste to set up random stuff because the player may never even see the guard (in which case patrol might be too long/wasted) or the player will only see him pass once and if he does or doesn't do the random thing they will never realize it anyway.

---------------

 

So a better way to use random patrols might be like this:

 

long patrols - static (same loop points with very little variation).

 

one room patrols/stationary guards - random (ai does lots of tasks in small area so player witnesses the brilliance of the patrol)

 

medium patrols - semi-random (player might see guard in hall several times, and once witnesses them do something out of the ordinary)

----------

I think Spring is looking more at the flavor of 'conversations' which is what you used in Thief 2 to 'spice' things up.

Whereas some of us are thinking more about the actual patrol 'pathing'.

 

I think all patrols, wheter static or random or in between can always use 'random flavor', but that still doesn't describe the type of patrol. ie: picking nose, drinking, eating apple, turning speed/angles... all add flavor but don't really change patrol enough to throw player for a loop.

[Sotha]

While it's a completely different genre, The Dark Mod presents some similar conditions such that I think it could also benefit from semi-random placement or movement of NPCs. Like a tactical shooter, a stealth game like The Dark Mod has enemies that have a significant effect on how the player approaches a situation just by where they are, and again a single enemy in the right place can make things very difficult for the player if the player doesn't prepare for them. If there exists the possibility of NPCs coming from unexpected directions then that would make things very tense for the player and cause them to adopt a much more cautious style of play. Devices like mines, noise arrows, and perhaps even random junk items that make sound when stepped on would probably see much more frequent use. Player will start to pay more attention to whether they leave certain doors open or closed, how they "inspect" rooms they cannot see through sound and other means, and other such subtle details.

 

I suspect that if dynamic patrols became more common in The Dark Mod then it would spark a lot of deep discussions of tactics among players, especially regarding how to deal with such unknowns. Should you leave doors behind you open so you can better hear or see approaching guards or should you close them to keep concealment? When should you break away from lockpicking? How can you best locate a guard you cannot directly see? How can you best place small noise making object mines in a given common hallway configuration to ensure maximum safety for yourself before proceeding? Dealing with an unpredictable threat can cause your tactics to become much more sophisticated than they would be otherwise.

 

On the other hand, while having more dynamic patrols would change the gameplay in a way that I would personally find very appealing, I'm guessing not everyone would agree with me.

I'm the sort that would take failure due to a random guard walking in on me not as unfair but as a sign that I need to improve my tactics to minimize the probability of such failure. Not everyone is going to take the random guard walking in on them quite as well.

 

I couldn't agree more. In fact I think you bring up the most interesting point yet mentioned in this whole thread!

In many games which you mention, the enemies could be anywhere. This increases replay value and keeps the player on his toes.

 

Left for dead 2, for instance, offers zillion hours of fresh gameplay as each playtime is an unique experience.

 

In TDM the FM is sort of 'movie' in terms of guards. Once you've played the mission you know where the guards are and how they move. Later if you replay the mission you remember/know where they are and the mission is not pitting the player in an unexpected situations anymore: "Hey I know the noblewoman passes this corridor at some stage of her patrol." True, the guards might be a bit different phase of their patrol, but you know their routes.

 

Let's think what would happen if the guards had their patrol slightly altered between each game? Every time the player plays the mission, he is getting slightly different experience in terms of the guards. Sometimes getting in is easy as the guards move favourably, sometimes it is utterly impossible to get past the guards without expending equipment.

 

This replay effect is something I'm very interested in achieving. I honestly believe it will make an FM much more interesting.

 

I tried it in a simple level in Mandrasola: when the mission starts, the guards first target a node that sends them randomly to any path of their normal static patrol route. This resulted in a slightly different moments after mission restart, but it still increases replay value as someone mentioned in the mission thread.

 

In my current WIP I'll try to develop the idea a bit further. If someone wants to test the idea in their maps before I get my map done, it is rather simple. When you are gonna place path nodes, design two or three slightly or totally different patrol routes for each of the guards. (with interesting lifelike side tours with path_anims, etc.) Then when the map starts have each guard target a dynamic path node, which sends the AI randomly to one of the routes you have designed for him. Repeat for each guard.

 

The end result should be significant variation of gameplay experience between map restarts, as the AI's not only move in different phases, but in different routes altogether. The new-gameplay-feeling should be even stronger as all the AI paths seem to change between restarts. If you have two routes for each guard and 10 guards, the mission play experience is most likely never ever exactly the same. (Makes betatesting tricky, probably..)

 

Someone might argue that it will be a huge burden to create alternative paths, but I think it will be negligible when considering the overall effort required for a FM. Plus if it makes the FM worth more playing hours among the community, it probably is worth it.

[nbohr1more]

Quiet now Sotha...

 

If the "Professional Game Industry" finds that not only are quality games being offered for free but also they have infinite replay value (!!!) they will find a way to stop this.

[Professor Paul1290]

The player does have to be willing to accept some level of unpredictability with this. Inevitably some combinations of routes are going to result in a much more difficult or much easier playthrough depending on which routes end up being used. On the other hand, if they don't result in some noticably varying difficulty then it could be argued that there isn't enough variety in the routes.

Given that I don't see players comparing exact stealth scores all that often in The Dark Mod I don't see this being that much of a problem.

 

I feel the biggest thing this would add is some degree of uncertainty that requires the player to prepare for what "might" happen rather than what "will" happen. I feel that not enough games have this. As I mentioned earlier, a lot of depth can come out of trying to deal with somewhat unpredictable threats. In many games this gets subdued by the fact that most missions/levels can be played through a certain way that's always best.

 

When you start trying to prepare for what "might" happen rather than what "will" happen I think it encourages a different way of learning the game. Instead of coming up with ways to conquer a specific missions, you take commonly occuring situations and find ways to best deal with them. You focus less on making a walkthrough in your head and focus more on developing tactics. Such methods are much more versatile and can be applied to many missions, so you're more likely to carry these lessons with you when you play a new mission. Learning from failure becomes more about improving a set of tactics that you've carried with you and have been refining over the course of several missions rather than figuring out how to beat that specific mission.

[Baddcog]

@Sotha, I don't feel the need to have say 2-3 different routes for guards. As i described you can have one base route, then have random splits off of that.

 

So all you really need to do is have a split right at the start of the main patrol route but the guard will end up in different places/ the timing will change per map load. ie: he could start by walking towards or away from player.

 

Still it makes it predicatable enough for the author to keep the ai where they need to be, and yet it changes it enough players will be thrown between loads.

[grayman]

If you rely on the "guard randomly picks his starting path node" approach to extend the life of an FM (aka "replayability"), keep in mind that--even with lots of nodes to choose from--each mission could start the same way, because computer-generated random numbers follow a sequence. Play N+1 of the mission would generate the same sequence of random numbers as Play N.

 

Once the game is underway, the effect of duplicate sequences fades as more and more events take place, especially the player's actions. However, duplicate sequences at entity spawn time should always produce the same results.

 

I haven't looked at how TDM seeds its random number generator. If someone else has, perhaps they can speak to whether anything's been done to guarantee the sequences aren't always the same. For example, seeding with the current time would produce a different sequence, and thus a different playing experience each time.

[Sotha]

Once the game is underway, the effect of duplicate sequences fades as more and more events take place, especially the player's actions. However, duplicate sequences at entity spawn time should always produce the same results.

 

Yes, that makes sense. But since the the path decision is not made at spawn time, it should be okay, right?

Spawn time: AI spawns, path node spawns.

Later: The AI reaches the path node. The path node sends the AI to one of the 2-3 path options.

 

I am assuming the random path decision is made the instant the AI reaches the path node with multiple targets.

 

In my Mandrasola experiment I think the AI's chose their starting nodes randomly.

 

@Sotha, I don't feel the need to have say 2-3 different routes for guards. As i described you can have one base route, then have random splits off of that.

 

So all you really need to do is have a split right at the start of the main patrol route but the guard will end up in different places/ the timing will change per map load. ie: he could start by walking towards or away from player.

 

Yes, probably. But like someone said here, they would like the path be somewhat predictable during the game.

 

Your suggestion would, in essence, have the AI split from the main road to all the possibilities in a single game play time.

 

In my suggestion the single play time experience will not confer the player any information of all the possibilities that could occur on another play time.

 

Basically, I'm probably gonna do like you say (ans I said in post #2):

• Make a single static patrol route with dynamic branch points to make the AI look more lifelike
• Then I repeat this two or three times per AI.
• Then a map restart makes each AI to choose between one of these routes.

Here will be the benefit of

• Rich interesting AI behavior in a single playsession
• The prospect of still having a totally different playing experience the next time the player picks up the mission. In a single playsession, the player cannot get information what the paths would be in the next playsession.

I think my method will keep the FM more interesting (in terms of guard activity) for a few additional playsessions. Heck, if the thing works like it should, I think the mission will pose a challenge even to it's creator as the combinations of the patrol routes may surprise even him.

 

The other thing I am dreaming of is to have mission-critical-objects to choose their location among a few logical places. But that's probably achievable by scripting. Can I have something like

Roll a d6.
If 1-2, teleport mission critical object to location A
if 3-4, teleport mission critical object to location B
else
, teleport mission critical object to location C

[grayman]

Yes, that makes sense. But since the the path decision is not made at spawn time, it should be okay, right?

Spawn time: AI spawns, path node spawns.

Later: The AI reaches the path node. The path node sends the AI to one of the 2-3 path options.

 

1 - The longer you can delay the AI reaching that first path node where he randomly decides, the better your chance of having different patrolling pattern combinations per session. If the first path node is right next to where each AI spawns, then the number of different patrol patterns is reduced.

 

2 - You'll have a better chance at randomizing the patterns if you have lots of choices. If an AI has 2 choices, the player could easily play the game 5 times and see the AI walk the same route, because the random number generator happened to pick the same place to go 5x in a row. If the AI has 10 choices, the odds of seeing the same pattern in 5 sessions goes way down.

 

 

I am assuming the random path decision is made the instant the AI reaches the path node with multiple targets.

 

Yes.

 

 

Can I have something like

Roll a d6.
If 1-2, teleport mission critical object to location A
if 3-4, teleport mission critical object to location B
else
, teleport mission critical object to location C

 

Yes. Though with so few choices, a player could play the game 3x and find the object in the same location.

 

One method of guaranteeing a different playing experience each session is to capture the choices made in the first session and don't consider them in the next session. If the object was in A the first time, then the next time it will randomly appear in B or C. If B, then the third session would place the object in C. A fourth session would go back to considering all locations again. Ditto for the initial AI patrol choices.

 

 

 

[stumpy]

it depends on the map terrain which patrol type you use.

[fllood]

FMA's may use the difficulty implemetations to make differences.

This may cover challange rates and preferences which do correlate imo.

 

For example like this:

 

Easy: static behaviour (with random wait times)

 

Medium: static with a few dynamic twists added

(going to the kitchen / toilet, turning directions, a few random forks to path corners and anims)

 

Hard: much dynamics included

(lots of random fork to path and anim corners)

Edited July 15, 2011 by fllood

[Baddcog]

FMA's may use the difficulty implemetations to make differences.

 

 

This is also something that hasn't been discussed. You can delete one of the random path nodes for easy difficulty and that ai will be more predictable for those players.

ie: hard has 3 random paths. easy will have only 2.

 

---------

 

@Sotha,

 

I still prefer the ai to have one patrol that deviates. Randomness is good, but there's a point when too much won't even be noticed. Like I said before, a really long path can have a lot of variety, but how much of that will the player even see?

 

The thing with Thief is the author typically places guards where they want security. So if you make too many deviations in paths, etc.. you loose control over making tough spots/easy spots where the player has a challenge/break.

Of course this could be discussed ad neaseum(sic). But my point is too much randomness can make the events unpredictable even for the author.

 

As far as replays, yes some people do play maps multiple times. typically I don't, so I wouldn't notice the random patrol. Probably only random actions on a small patrol (ie cook in small kitchen).

 

Sometimes I do forget to quick save at mission start though and end up restarting several times. So would probably notice alterations of paths with the first few guards encountered but not with guards further into the mission.

---------------

 

As authors (as stated above) we can have illusions of granduer. ie: the player will be amazed with the variety of paths this ai has. Because we know it has them, because we did the work on them.

But when it comes down to it I don't think the majority of players, even players ho replay several times will even notice.

 

I certainly remember a guards path during a mission. But if I came back a week later and he has an altered path would I realize it? Most likely not, I would just recognize his new pattern within that play time.

--------------

 

Random 'main objective objects'. In some cases I guess it could add a bit. But in most cases a mission is built around specifics. ie: Baffords scepter is always in the throne room. The throne room is your goal. If it appears in his bedroom you'll have to have an entire backup story for that possibilty. Why is it there? Why did the player go to the throne room? When they get there do they know why it's in his bedroom?

It's one thing to have scripting to tell the player it might be there.

It's another just to have it elsewhere randomly.

 

Again, I think it's easy for authors to wring their hands and think "muhahaha, I'll trick those players good!" and another to look at it from player ONLY perspective. While it's cool to have random loot, how many players will actually replay enough/memorize loot location? How many will realize it's happening.

[lost_soul]

I like surprises so...

[Baddcog]

I like surprises so...

 

But it's only a surprise if you play them map several times. Do you?

 

This is my point. We can say it's random, so players will be surprised. But will they if they haven't memorized the mission from another play through?

[Sotha]

@BC

We'll have to see until after the experiment.

 

If FMs would be single-use entertainment for everyone, I'm not sure how mappers should stay motivated to create more maps. I would be delighted if people saw my work as more than disposable single-use fun.

 

My aim is exactly to offer a relatively simple mechanic to increase the replay value of a single mission.

 

If I manage to create a mission that stays fresh and surprises even the author, myself, I know I was successful in my goal.

 

Since you have chosen to play a given mission only once, I'm fascinated how you actively undermine the mission restart randomness as that will not concern you (because of your decision) but may bring a lot of fun to those who really DO replay missions.

 

But let's see how it goes. I may fail miserably in my goal. But I find it interesting and worth the effort to try.

[Tels]

But I find it interesting and worth the effort to try.

 

I can just underwrite this. Making missions having a higher replay value is easy, and yet brings a fresh thing to the mission. In basically extends the value of the misison with very little efford.

 

And you know, once you have beaten all missions, you want to beat some of them again (you know, all these people who still play T2

 

So, Sotha, please go for it!

 

Oh, and using the random entity from the SEED system will help, too:

 

http://wiki.thedarkmod.com/index.php?title=SEED_-_Premade_entities

[Professor Paul1290]

Sure not everyone re-plays missions, but I think more dynamic patrols can have significant enough effect on those that do at all.

 

I think enough players are going to memorize where things are to some degree, at least enough to know which rooms are safe and which ones aren't in later playthroughs. Even if I replay a mission I haven't played in months I may not memorize exactly where the guards go, but I still have some semi-conscious memory of which spots are "safe" and which ones aren't. I already fall into a sense of security when going through certain hallways and rooms in missions because even if I don't memorize the guard patrols exactly, I am likely to remember which rooms don't get any guards at all.

 

The thing is, players don't really need to conscously acknowledge whether the guards are "random" or not. All they need to know is that there is the threat that they will encounter a guard where they might not think they would, and by accomplishing and reinforcing that you've already influenced how they play.

 

Going back to the SWAT 4 and Rainbow Six 3: Raven Shield example. When me and my friends play a mission in either of these games we often don't think about the fact that the hostiles are randomly placed as much as we think of the fact that a hostile may come from any doorway or be in any position they can reach at any time. That in itself really affects how we proceed through the mission. We're always careful about rooms we haven't gone through yet and even rooms that we have already been through but have not checked within the last minute pose a significant threat to us.

The threat that a hostile might come into the room or be in position to pose a threat to us affects how we play the game even when they don't actually do so.

 

Amnesia: The Dark Descent is also a good example of this. Many players don't know this, but a few of the monsters in Amnesia aren't always there, and may or may not be there on a given playthrough depending on chance. This only applies to a few of the monsters and most players don't actually know that they may or may not show up, but that doesn't stop them from occasionally getting the crap scared out of them when they run into a monster they somehow didn't remember was there, in some cases they really weren't there the first time.

 

Guards in The Dark Mod have a just as great or greater effect on the player than terrorists or suspects or monsters in either of these games just by being where they are, so I don't think there has to be quite as much randomness in this case to achieve a similar effect.

 

 

The point is that the players don't necessarily need to know which guards are random nor know all that much about the patrols themselves, though it would be nice if they did.

You don't necessarily need the player to consciously know the guards are random. You just need to make it clear that a guard may become a threat to them when they aren't expecting it and that they should prepare for such a scenario. The threat of a guard walking in on you while you are reading or looting is enough to affect the way you play the game even when they don't actually walk in on you.

Edited July 16, 2011 by Professor Paul1290

[fllood]

Devianting from gameplay considerations, I also think guards would have a much more static predictable route than commoners or monsters. That could be also a way for the mapper to add both static and dynamic paths to a FM but keep consistency to the player in that some AI are predictable on their routes (guards) while others won't (undead, nobles, etc.).

[Sotha]

I can just underwrite this. Making missions having a higher replay value is easy, and yet brings a fresh thing to the mission. In basically extends the value of the misison with very little efford.

 

Hmpf. Actually it is not that easy. I made an experimental single guard setup with two path possible routes, which covers a street and dead end. The problem is actually the patrol area: there are only a few logical places the guard would go to. Thus the patrol routes are not very different from each other. Such type of guard does not apparently benefit from choosing between alternate routes. I think it is more beneficial to do such a situation with a simple static route, with a few dynamic deviation points.

 

I need to still check how the random paths will work inside a building, but at present it seems that increased amount of dynamic nodes may indeed be the way to go.

[Sotha]

Okay, I found a neat system how the author can surprize himself and the map has higher replayability in terms of guard activity.

 

It works like this:

 

The map consists of areas, like a street patrol area. The mapper places ordinary static patrols in there (green arrows) which contains a few dynamic branch loop points to keep things interesting (red arrows). These static guards are the basic staple of the map protection, the mapper can control which areas are constantly manned with these.

 

But with only these, everything is expectable. There will not be any random surprise guards around.

 

But then the mapper adds something I call the dendritic dynamic matrix. It is essentially a bunch of branch points, which branch even further. Note that the direction is set so that the AI never do a (bad looking) U turn and go backwards, but the overall structure of the matrix of waypoints lead the AI's overall circling around the map. The AI appears to be walking intelligently forward around a patrol path, but essentially the movement is as random as the mapper built the matrix.

 

The "chance" parameter can be used steer the matrix walking AI's towards the more important areas.

 

The mapper could use the same matrix, but reverse the flow direction to have a few the AI's go to the other way too.

 

Now, when the mapper drops a few AI on this path, they will get dispersed in the matrix and walk in totally (well, mostly totally) unpredictable path through the map. The in-going guards can even target the same node initially, but they will eventually get lost in their own paths.

 

The mapper should not put too many AI's to walk the matrix, but only a few totally unpredictable AI's to avoid the mission becoming too difficult. If the mapper gives these few AI's some torches, the AI's have significant gameplay impact. Mission difficulty setting could really easily adjust the amount of matrix walking guards and whether they have torches or not.

 

The player can always be nervous about a torch wielding AI to happen to pass by his way. Also the matrix walking guards are very effective in finding bloodstains and hidden bodies.

 

If I will not encounter unexpected (haha) problems with this approach, I'll use this system for my mission instead of the mission-start-random-path method.

 

Comments, ideas, suggestions?

[Baddcog]

umm yeah that's basically what I said, lulz

[Baddcog]

@BC

We'll have to see until after the experiment.

 

If FMs would be single-use entertainment for everyone, I'm not sure how mappers should stay motivated to create more maps. I would be delighted if people saw my work as more than disposable single-use fun.

 

My aim is exactly to offer a relatively simple mechanic to increase the replay value of a single mission.

 

If I manage to create a mission that stays fresh and surprises even the author, myself, I know I was successful in my goal.

 

Since you have chosen to play a given mission only once, I'm fascinated how you actively undermine the mission restart randomness as that will not concern you (because of your decision) but may bring a lot of fun to those who really DO replay missions.

 

But let's see how it goes. I may fail miserably in my goal. But I find it interesting and worth the effort to try.

 

 

I'm not ACTIVELY UNDERMINING the idea.

 

Simply pointing out that the majority of the work will be wasted on the majority of players. It's only a few players who will replay missions, and even fewer who will replay any specific mission depending on their favorites list.

 

I've been doing this (rather lackluster lately I admit) for over 10 years now. It's not that I haven't thought it through and haven't got plenty of feedback over the years.

 

It's easy for an author to KNOW how cool said features are, but in reality most players just will never realize such small details even exist, even if they do replay.

 

The point is not to spend 6 months on a map that 10 people will replay constantly, but trying to get 400 people to play once. Is one more valuable than the other.

 

You can do as much as you want, work on a map for 10 years making every tiny little detail perfect and magical if you want. Doesn't matter to me, but there's a point when you gotta ask yourself, is this REALLY worth it?

Do you want to script out 12 hours of an ai's action so a player can sit and watch them perform every daily task they do realistically from waking to bedtime? Or do you want to map around the 30 seconds to 5 minutes of gameplay that the player might actually see that ai do something.

 

This is why patrols don't need to be amazingly random, because in a 2 hour play session (ie: big map) the player is not likely to notice any small deviations in patterns, much less huge deviations.

And gameplay to gameplay seperates them from that even more.