Aggro Radius?

Discussion in 'Game Mechanics Design' started by staf00, Nov 8, 2019.

  1. staf00

    staf00 Veteran Veteran

    Messages:
    83
    Likes Received:
    9
    First Language:
    English
    For those of you whose projects have enemies visible on the map, what are your design philosophies when it comes to combat initiation? Here are some of the setups I've come across:
    · Battle starts when the player touches the mob
    · Battle starts when the player comes within a certain distance of the mob
    · Mob chases the player when close enough, but doesn't enter combat until touched
    · Line of sight (Pokemon style)
     
    #1
  2. ShadowDragon

    ShadowDragon Veteran Veteran

    Messages:
    916
    Likes Received:
    282
    Location:
    Netherlands
    First Language:
    Dutch
    Primarily Uses:
    RMMV
    1st and 3rd option (mine is option killing) but some are required to continue.
    because of specific item drop.

    I dont have any leveling going on, but some has items you need or pass a test :)
    it wis a challenging game, so you need to think some and has the items to continue.

    some bosses might be unbeatable or cant enter it without the requirement to be able to beat it :)
     
    #2
  3. gstv87

    gstv87 Veteran Veteran

    Messages:
    1,800
    Likes Received:
    829
    First Language:
    Spanish
    Primarily Uses:
    RMVXA
    ranged aggro and LOS.
    I modified the event movement routes, so the enemy doesn't move outside of their assigned areas, and doesn't engage unless within a certain range.
    this allows the player to sneak around, learn movement patterns and avoid combat if possible.
    it can also lead to some trouble, if you happen to trigger a battle within the aggro range of another enemy that hasn't seen you, while escaping one that has: you end up fighting more enemies than expected, because those aggro mobs will join the fight.
     
    #3
    JosephSeraph likes this.
  4. somenick

    somenick Veteran Veteran

    Messages:
    173
    Likes Received:
    81
    First Language:
    English
    Im making a game with visual encounters. Combat does not start unless the mob touches you (or vice versa). Sometimes they wander around aimlessly and dont really care about you. Others will try to chase you. Most combat is optional. In some cases their position is fixed (like blocking a corridor) and it is generally mandatory to beat them (these will not respawn again afterwards, though, not even if you leave and re-enter the map).
     
    #4
    JosephSeraph and M.I.A. like this.
  5. JosephSeraph

    JosephSeraph White Mage Restaff

    Messages:
    863
    Likes Received:
    725
    Location:
    Santos, Brasil
    First Language:
    Portuguese
    Oddly enough I never even considered line of sight like pokemon. It could be fun. I also have never seen it out of pokemon.
     
    #5
  6. TheoAllen

    TheoAllen Self-proclaimed jack of all trades Veteran

    Messages:
    4,456
    Likes Received:
    5,050
    Location:
    Riftverse
    First Language:
    Indonesian
    Primarily Uses:
    RMVXA
    I have 3 kinds of visual encounter with color indication
    - Blue flame = Slow moving, easy dodge, chase you on 5 tiles, wander randomly otherwise
    - Red flame = Has a slightly longer aggro radius, and chase you faster.
    - Dark flame = Surprise encounter, randomly spawn near your location, smarter and faster than red flame, happens if you evade the visual encounter too much (I won't disclose how I handle this). If you successfully evade this encounter in 2~3 seconds, it will be gone. Counter will reset
     
    #6
  7. Komm

    Komm Warper Member

    Messages:
    4
    Likes Received:
    1
    First Language:
    Latin
    Primarily Uses:
    Other
    Id say mob chases the player whenever possible, it feels more natural that way, but it also is harder to implements :(
     
    #7
  8. Wavelength

    Wavelength Pre-Merge Boot Moderator

    Messages:
    4,457
    Likes Received:
    3,728
    Location:
    Florida, USA
    First Language:
    English
    Primarily Uses:
    RMVXA
    For a "standard" Visual Encounter system where the intention is to allow the player to try to avoid some combats, I think the best option by far is that the Encounter starts to chase the player when they come close enough (and/or when the player enters the Encounter's line of sight), and the Encounter only starts when contact is actually made. This allows the player a chance to use their skill to avoid combat, but doesn't make it drop-dead easy to do so.

    When using this approach, it's very important to find a good balance between being able to outrun the Encounters (even once they're being chased) vs. occasionally forcing the player into combat. If the player's speed is faster than the Encounters' chase speed, you need to make the terrain tricky enough to navigate, or make the number of Encounters dense enough, or give some of the Encounters special chase abilities, so that the player can't just run right past every single one and never fight anything. If the player's speed is slower than the Encounters' chase speed, you need to give the player special abilities on the adventure map (such as limited Dashing or skill-shots that temporarily freeze or destroy an Encounter), or allow the player to somewhat easily get Encounters stuck in terrain or traps, so that the player has a reasonable chance to avoid an Encounter even once they have aggro'ed it. You can see good examples of this in most of the Paper Mario games as well as most of the Tales of games.

    My remake of How Badly Do You Want It? is the first time I'm really making an effort to make a true Visual Encounter system (my early games had Random Encounters and my recent games, including the original How Badly, haven't had Encounters at all). I am working really hard to get the system feeling right (as, like @Tai_MT frequently mentions, execution is King when it comes to Visual Encounter systems), and to make the art of dodging (or engaging) the Encounters fun. Among the features I have either already implemented or plan to implement:
    • Visual Encounters (VEs) will wander around in dungeons; the approximate Power Level of the monsters you'd have to fight (relative to your party) is represented by their size and color.
    • By default, VEs wander dumbly around a room at a slow speed.
    • VEs will start to chase you if you enter their line of sight - a thin cone in front of them with a medium range.
    • VEs also have a small chance to start to chase you every fraction of a second if you are running while in a small radius of them.
      • This is the downside to running rather than walking; the upside is that you'll need to run to avoid VEs that are chasing you, and also that you get bonus rewards for finishing dungeons within a certain time limit.
    • VEs that are chasing you will move faster than wandering VEs. The more VEs that are currently chasing you, the faster each one will move, up to a limit. Therefore, you can easily outrun the first few VEs you aggro, but as you aggro more (without fighting them), they become harder to simply outrun.
      • Pathfinding algorithms allow VEs to find their way around most walls and decorations as they chase the player.
      • VEs that are much weaker than you go into a Flee pattern when they sense you, rather than chasing you.
    • Getting a very far distance from a VE will cause it to stop chasing you and return to its Wandering state.
    • Each party members has Abilities they can use in dungeons, many of which interact with VEs.
      • For example, the Wind Mage can blow nearby VEs away, the Energy Mage can stun VEs in a straight line in front of her, and the Water Mage can create a zone of mist within which VEs' line-of-sight cones are disabled.
      • Abilities cost MP to use.
    • When you make contact with a VE, a battle begins. Any other nearby VEs are also added to the same enemy party, potentially creating very difficult battles!
      • This is a risk/reward mechanic that should encourage players not to get "too cute" with trying to evade all the VEs unless they can handle a stacked battle.
    • If you make contact with the VE from behind, or while it is disabled by an ability, you gain Initiative. This grants you a free turn at the beginning of combat and bestows several other advantages on your team as well.
      • If a VE hits you from behind, or while you are disabled, the enemy party gains Initiative.
    • As a feel-good, after you win a battle, VEs that were included in that battle explode into collectibles like EXP, Gold, and Items that you can pick up.
    There's one other type of Visual Encounter I can think of that comes from an entirely different design intention, where Visual Encounters stand in choke points and you have to fight them to progress. Ostensibly, the battle only starts when the player walks up to the Encounter and initiates combat with it. Here, the player isn't allowed (or expected) to evade any of the Encounters; it's a visual way to force the player through a gauntlet of usually-predetermined battles in each stage. This sounds a bit lame, but it can work well, especially in tightly-balanced, strategic, battle-oriented games that are designed to wear your party down over the course of several battles. This eliminates the "BS" arbitrariness of Random Encounters, and the wide variance that can come from "standard" Visual Encounter systems. It gives the designer very tight control over battle rate, at the cost of feeling a little artificial.
     
    Last edited: Nov 11, 2019
    #8
    JosephSeraph and Kes like this.
  9. Tai_MT

    Tai_MT Veteran Veteran

    Messages:
    5,195
    Likes Received:
    4,370
    First Language:
    English
    Primarily Uses:
    RMMV
    @Wavelength

    Oooo… you're including me in a discussion on Visual Encounter systems? I'm not sure I've got much more to add that you haven't covered. Especially since I'm so picky about how VE's are executed. :D

    Though, I gotta say, it does sound like a lot of work went into your own VE system and it's not "bog standard" like so many I see. I daresay it sounds pretty interesting and I'd love to tinker with it to see if I find it fun. I think my only concern with it would be what the map design looks like around the encounters. Otherwise, the system sounds pretty cool and fairly versatile.
    ---
    So, for those who don't really know my stance on Visual Encounters, I'll give you a brief rundown before I get to the "nitty gritty".

    I am not really a fan of Visual Encounters. One of my favorite games of all time is Earthbound and for me, that Visual Encounter system is the "Gold Standard". I haven't seen a game do their Visual Encounters as well as this game before or since.

    Typically, I advocate players do something like what Wavelength has done. That is, if you're going to use a Visual Encounter system, you need to do more than leave it "bog standard". The "Bog Standard" way of doing it is very boring, often frustrating, and has a lot of issues. There's a few threads around where I cover those things in far more detail. For me, as a player, there is nothing more annoying or frustrating than Visual Encounters that work exactly the same no matter what game I'm playing. The enemy moves at the same speed as me or faster. It can corner me and I can't escape. It blocks my path so I am forced to fight it. It's AI is limited to "move towards the player". It is difficult to grind Visual Encounters because they don't respawn, and I need to screen transition for it. The same enemies will appear on the same screens no matter what, which means there's no variety. There is little, if any, interaction with the enemies on the map (making them flee from you, stunning them, tricking them, luring them to other encounters to fight those instead, being able to debuff enemy encounters on the map before battle, etcetera). There is little, if any, reward for actually evading enemies, so the very act of avoiding combat is detrimental to the player, and thus there's no reason to have a Visual Encounter that can be avoided.

    Put simply, if it's going to be in a game, I want it to be as fleshed out as possible. I want it to be its own system. I want it to add flavor to the game and the world. I want it to engage me as a player.

    I do not want a Visual Encounter system as an alternative to "I took one step, now I'm in battle" Random Encounter system.

    With that being said...

    If we're just talking about how the Visual Encounters initiate combat, I think I prefer "when they touch me". Though, admittedly, Pokémon uses "Line of sight" pretty effectively. Though, that system comes with its own myriad of problems.

    The reason I prefer the "when they touch me" approach is usually because it gives me a lot more freedom to make decisions on the map. I can get the encounter stuck in geometry, I can run off the screen and then back on it to reset its position and try evading again, I can run around it, etcetera. It gives a lot more breathing room. It also allows the dev to do interesting things with its movement route and in programming the AI that will try to engage you.

    The problems that the "Line of Sight" version in Pokémon has is... well... it's too easy.

    Since about Gen 2, I've adopted a strategy for dealing with all Trainer Encounters. Namely, I save before walking into their line of sight. If the battle goes badly, rather than be defeated, I reset the game, level up a bit more, change my party composition, and roflstomp the trainer. I've effectively removed all challenge from the trainers. Now, the older games used to anticipate the player doing this, so it would have "gauntlets" that you could not escape. I'm talking about more than the "Elite Four". There were usually a few places that were "points of no return" and your only option was to trudge forward to complete the area since you could no longer go back. This made trainers far more challenging and forced you to monitor your supplies to keep from losing.

    Newer games and remakes of those older games have removed those Gauntlets (which, in my opinion, is terrible, since these gauntlets were the only time I ever felt challenged, the only time I ever felt accomplished when I completed them, and were enjoyable nail-biting experiences) in the interest of making games more "kid friendly" or "accessible" or whatever.

    However, it has just rendered the "Line of Sight" encounters a moot point again. I just save before each trainer again, fight them, if I don't do well, I reset. There's no threat. No challenge. Sure, newer games have taken to "the trainers spin around in circles uncontrollably, so they have a greater chance of actually spotting you unprepared", but in practice it looks silly and is even EASIER to avoid than when they stood still facing a single direction. My saves are just made "out of line of sight" of any possible direction they can spin now (which is like 2 steps difference than where I used to save). Or, if I don't care about the fight, I just "mad dash" by these trainers, since their line of sight is changing and I've only got a 1 in 8 or 1 in 4 chance of even being spotted. So, you know... even EASIER to avoid while they tried to make them harder. Except... Pokémon has this fun problem: The best places to get XP in the game are Trainer Fights. The only places to get money in the game are Trainer Fights. So... why would anyone evade a Trainer Battle at all? Yeah, I fight every trainer. I beat every trainer. Because, I need the money. Because, they are the fastest way to level up my party. So, why does this "spinning" exist at all? Or alternate routes around the Trainers? The optimal way to play is to fight every trainer the moment you get access to them. Evading trainers is a bad idea all around. Especially since you will have no money to buy Pokeballs that you need to complete the Pokedex or even to catch better Pokémon.

    I could go on and on at length about this sort of stuff and its missteps. Or ways that I, as a player, have exploited existing systems to remove all challenge and tension in a game.

    tl;dr So, yeah... basically... I just prefer touch encounters and good AI to go with those in a Visual Encounter system.
     
    #9
    JosephSeraph likes this.
  10. M.I.A.

    M.I.A. Goofball Extraordinaire Veteran

    Messages:
    812
    Likes Received:
    604
    Location:
    Seattle
    First Language:
    English
    I like to keep the radius predictable in my main. NPC's and Trainers all explain it to the new player.
    - Battles only happen when the player touches an enemy or an enemy touches a player
    - Most monsters just meander about, but will chase the player if the player gets within 3 tiles of them.
    - Monsters will only chase players for a short duration (I set blockades with Region Restrictions for all onscreen monsters).
    - Some Monsters are immobile and are obstacles that you'll have to defeat to continue through.
    - Some Monsters don't chase at all, they just meander around.
    - There are few places where there also "random encounters", but they are few and far between.. and a head's up is given beforehand.

    Hope this helps!
    -MIA
     
    #10
  11. Milennin

    Milennin "With a bang and a boom!" Veteran

    Messages:
    2,212
    Likes Received:
    1,292
    Location:
    Fiore
    First Language:
    English
    Primarily Uses:
    RMMV
    I set a roaming zone for each monster on the map in which they freely roam around at a slow pace. Once the player enters it, it'll chase the player and start battle if they're touched. Keeps monster sprites from going into places I don't want them at or overlapping with each other and feels pretty natural.
     
    #11
    M.I.A. likes this.
  12. CuddleFox

    CuddleFox Furry Veteran

    Messages:
    69
    Likes Received:
    33
    Location:
    France
    First Language:
    French
    Primarily Uses:
    RMVXA
    Personally I make sure that the enemy is visible on the map, has a movement pattern but the fight only starts if the player touches the enemy and not anything else. To give the player the choice to fight or not.
     
    #12
  13. Wavelength

    Wavelength Pre-Merge Boot Moderator

    Messages:
    4,457
    Likes Received:
    3,728
    Location:
    Florida, USA
    First Language:
    English
    Primarily Uses:
    RMVXA
    Yeah, I invoked your opinion because you're probably the most well-known member on this board who's critical of Visual Encounters in general, and as such I felt your view on "how to do it right" would be a really valuable add to the topic.

    I also wanted you to see my post since we've discussed the topic in the past, and this was the first time I've laid out the details for how I'm doing VEs in my own game. I'm glad that it sounds good (or at least intriguing) to you, and I'm curious about this concern:
    What I'm working on is creating randomized dungeon layouts composed of "rooms" with natural objects forming the borders and breaks between rooms, as well as smaller objects in the middle of rooms, in order to make it non-obvious where each room begins and ends. It will also mean very few thin corridors and relatively few dead-ends to get trapped in. I'm hoping this will also allow the VEs to flow naturally around barriers and between rooms.

    What do you think are the best ways to make room/map layouts for games with VEs, or conversely, what do you think are the biggest pitfalls to avoid?

    I really like this. While it's not necessary to throw the Kitchen Sink of features in, it is really good to do what Tai described and make sure that the player has lots of ways to interact with any kind of system you add into your game.

    VEs are a great example of this - there are a lot of "hooks" where you can add ways to interact with the system, from the Encounters themselves, to the terrain that the player and VEs will run over, to the way that battles start when they touch (or spot) you. As the designer, figure out fun and rewarding ways to interact with a system like this that won't get stale over time. :D
     
    #13
  14. Tai_MT

    Tai_MT Veteran Veteran

    Messages:
    5,195
    Likes Received:
    4,370
    First Language:
    English
    Primarily Uses:
    RMMV
    @Wavelength

    I'll quote the previous post to sort of illustrate what I'm talking about with the map design. Hopefully, it will make a lot more sense that way.

    Depending on the size of each room, or how much clutter is in each room, seeing the "pathing" of these encounters may be interesting or predictable.

    While I do appreciate a "predictable" route in order to plan escapes and movements around the enemies to avoid detection... after a while, the "waiting for enemies to move" thing gets tedious.

    By the same token, if the encounters use entirely "random movement", then it becomes more difficult to get around the enemies or avoid detection. It could even be frustrating.

    But, how effective either option is would likely depend on the map design itself. Are any objects able to be used as cover? Any of those cover objects able to be moved "at will" or through some other means? How much space does the player have to maneuver in?

    By screen size, we're only using like a 17 by 13 tile distance. It's not a lot of room to include "space for maneuvering" as well as some of the other features you listed below. I'm imagining rooms at probably twice that size to cater to movement and highlighting the features listed. But, room "clutter" will reduce available tiles you can use for maneuvering.

    So, I think the map design will contribute a lot towards whether the "wandering" of the enemies remains interesting.
    Does the radius go over objects? For example, could I aggro an enemy on the other side of a 1 tile thick wall, if it's "aggro radius" is 3 tiles in any direction? Does it then try to pathfind a way around the entire dungeon to get to me? Does it check certain pieces of geometry more or less in this radius? For example, could it check over some debris, but not over a wall tile?

    Depending on execution of the map design, you could do some interesting things here. For example, if some things like foliage which would dampen sound exist, I could "run" through it and not be detected unless I was very close to the enemy already. But, if I'm running on a tile floor in a room that carries echoes well... running in any part of the room could aggro the enemies to my position.

    It sort of depends on how your maps actually work and what your tiles do, or any other considerations you put into the map design itself (including obstacle layout). If it is little more than "you are running within X tiles of enemy, regardless of what is between you" in order to trigger, then a player like myself might find this exploitable or never take the chance of "running" as there's little upside to doing so except after you've already passed the encounter.
    I rather like this idea, but I worry about its limitations in regard to map design. For example, could you get VE's stuck on other VE's if you use a choke point? Is that something you intentionally designed into some of the maps, or something a player may just happen to exploit? It's a tactic I've used in a lot of shooters. Force enemies to come at me through a choke point instead of trying to evade and engage in a large open space. How wide are the corridors? If they're 3 tiles wide, there may not be any opportunity to "choke point" the enemies and only a means of "outrunning" them.

    Likewise, if "outrunning them" is an option as it reads it is, I wonder how effective that would be depending on the map layout. Would I be able to "run a circle" around obstacles to slow them down? How fast do the enemies get? Is there a point where they will inevitably catch up with you? If the enemies can find their way around walls fairly effectively, would it even be worth it to try to use the terrain as an advantage to "escape" them?

    Without a good map design, I could see just accepting that you were "spotted" as a player and engaging in the fight without attempting to "bypass" the enemies at all. After all, if this one spotted me and is running, and I begin to run away, I'm going to aggro any other enemy within range, most of them in front of me, which will likely block off any escape route. Unless, there's terrain between them I can use to "double back" and get all the enemies behind me, but now I'm running circles around a map, unsure if there's even a way to lose the enemies aside from a screen transition, and if I'm now trying to evade all the enemies in this room, when they get so fast I can't outrun them at all, the combat is a foregone conclusion anyway. Would it be less aggravating and time consuming to just accept the combat once I was spotted, or is there a possibly to evade it entirely through clever usage of the map layout?
    This is kind of what I wonder about. If map sizes are simply double what the normal screens are (26 by 34), how far would I need to get away from the enemy to keep it from chasing me? Is there a time limit? A set amount of spaces away I must be? If they get faster, the more I'm dragging behind me, will I ever be able to evade it? Is it exploitable? Can it be used to accidently trap yourself? What if I evade it and the final tile I used to evade it, put me in a corner or something where the very next step starts the chase again? Or the next movement of the wandering enemy starts the chase again?

    Does this distance also account for objects in the way? Other encounters?

    There's a lot of variables in here that makes me wonder how it would all work in practice, with some actual maps to work within.
    ---
    I'm going to not include the rest as they generally fall under the things I've already quoted. How does that stuff interact with the map geometry? Can you cast across obstacles? Through walls?

    I might also add that with a system in which you can "lure" enemies around, you could potentially use enemies to trip switches for you, or blunder into traps ahead of you so that you don't hit them.

    It all sort of depends on the map design you've got going for you and how complicated this all becomes when you have to take all of that into account.

    Mostly, I just worry that there's not enough "room" or "opportunity" to really enjoy these features because of the map design.

    Which, in itself, might also be something to consider for everyone using a Visual Encounter system. You have to pay a lot more attention to your map design than you otherwise would. Hand craft nearly every escape point, choke point, and open room while making it appear that it's natural and not part of some "puzzle" the player will immediately recognize. Making each room a series of options instead of a linear puzzle to be solved.
     
    #14
    Wavelength likes this.
  15. somenick

    somenick Veteran Veteran

    Messages:
    173
    Likes Received:
    81
    First Language:
    English
    I think it is like everything else, a matter of taste. In this visual encounter example, you can either wait for the enemy to move or straight up go to battle with it. Just like I could keep running away from battles in a random encounter system.

    Wether you use a visual encounter system or a random encounter system, that is only half the battle (pun intended); the other half being the combat system itself. Is it enjoyable? Boring? Drags on for too long? Involves any strategy? Final Fantasy 1 may have started everything, but boy, system sure was slow enough and dull enough.
     
    #15
  16. Tai_MT

    Tai_MT Veteran Veteran

    Messages:
    5,195
    Likes Received:
    4,370
    First Language:
    English
    Primarily Uses:
    RMMV
    Except we ignore the obvious. :D in a Random Encounter system, one need only equip an item to turn off all Encounters. An item it takes about 10 seconds to program and can be given to the player immediately as they start the game.

    In a Visual Encounter system, such an option to "turn off encounters" rarely, if ever, exists. So, if you want to actually avoid combat, it can become QUITE tedious to try to do so when you have to wait for enemies to move on their patrol routes.
     
    #16
  17. Wavelength

    Wavelength Pre-Merge Boot Moderator

    Messages:
    4,457
    Likes Received:
    3,728
    Location:
    Florida, USA
    First Language:
    English
    Primarily Uses:
    RMVXA
    @Tai_MT First of all thanks so much for explaining your thoughts in more depth, and for the thoughts on how my aggro/chase VE system might succeed or fail. These are some very important questions for me to keep in mind as I continue to flesh out its implementation.

    One thing I do want to address right off the bat is that while the original How Badly was a VX Ace game, I'm creating the remake on GMS2. Ultimately I decided I needed more "help" from my middleware with the action elements than the RPG elements for this particular game, so after trying out both platforms I moved ahead with development on the more action-friendly platform and built the RPG mechanics up from scratch.

    So while concerns about things like Tile-Based Movement, and Tile Size vs. Map View Size, are certainly relevant to anyone reading this topic, I just want to say that I've personally got them covered - rooms are large, obstacles are large, and the player will be able to see a screen area that's larger than one full room at a time. Characters and VEs aren't confined to grids or tiles in my system; they can move in pixels and curves.

    By the way @staf00 if you feel we're diverging too much from your topic, please let me know and I'll gladly take it to PM with Tai, but I hope this can provide an interesting and useful look into the kinds of nitty-gritty details that designers need to think about when they're building Visual Encounter mechanics based on the kinds of systems you mentioned (line of sight, aggro radius, touch encounters, etc.).

    Very interesting and realistic thoughts. I've only just begun to implement the mechanics of each movement mode's mechanics (the first step was to implement the changes between modes, e.g. wandering -> chasing -> wandering). Currently, what I have the VEs do is wander in a direction for a few seconds, turn, then wander in another direction for a few seconds. Once they turn, you'll know where they're moving for a few seconds, and you can act on it.

    Perhaps more importantly, what I'm planning to do with the randomized maps is to generate them using large objects, so that it feels like one continuous space rather than a series of rooms and corridors. One thing I've noticed with some games that use randomized dungeons is that the player often loses a lot of agency when a lot of things end up together in a tight corridor. I want to mostly avoid that, allowing players a lot of breathing room to run around most VEs. The tricky part will be when there are several VEs in the same place, and you have to figure out more careful, clever paths through them if you want to avoid fighting.

    (Also worth noting is that, ultimately, I want players to find themselves in at least a few battles per floor of the dungeon, even if they are trying to avoid combat! The slow drain of hard-to-renew resources is part of what makes these dungeon runs engaging, and since defeat in combat is merely the loss of an in-game day rather than a Game Over, I can worry less about players being forced into battles while they're low. The VE system is more about being a fair skill element than it is about being a clear "fight or don't fight" choice to the player.)

    Right now, I don't include things like "cover objects", and yes, aggro radius can go through walls. However, it wouldn't be hard to implement a check to see whether certain "tall objects" were in the way of a VE's Line of Sight - it would just mean a small hit to framerate and performance.

    I remember playing Recettear (one of my favorite games) and feeling like even though it was weird you could aggro and shoot monsters through most walls, it still felt pretty good. So my instinct is that I don't need to make things too complex or physically accurate. I could change this pretty easily. What's your position on this - do you feel things like cover objects are necessary for a good VE system, or was it more of an example of one out of many features that could be included to "flesh it out" into a fun, interactive system?

    Obstacles with long tails or sharp concave curves seem to create choke points that player can intentionally get VEs stuck in if they're trying to escape. Of course, once the player moves far from that point, they will revert to a "wander" pattern that may get them out of the choke point. The best tactic is usually to try to outrun them but I've had some success snagging the VEs on walls. Feels like a good balance to me, but what do you think? I haven't designed the real obstacles for the dungeons (just placeholder shapes), so I'll keep this front-of-mind when designing the real things.

    The maximum speed for enemies is slightly faster than the player's speed, and that will only happen when you have something like 6 or 7 VEs all chasing you. If you are trying to avoid combat (not all players will or should - winning battles comes with real benefits), the idea is that if you aggro one, you've either screwed up a bit or willingly accepted a risk (running past it to save time against the floor's time limit), so now your job is a bit harder because a VE is chasing you while you move. You can still outrun it and keep moving on if you route well. Not that hard. If you aggro two, now they're moving a little faster, as a slight penalty for screwing up again (or a counterbalance against another accepted risk). Now it's more challenging. And so on, until after aggroing several VEs without getting away, the battle becomes nearly inevitable unless you pull a great move like lining them all up and stunning them with an Energy Beam ability.

    Until now I'd been thinking more about the player mechanics and the VEs' movement mechanics, so thank you for bringing up the importance of map layout as a vital third factor in the equation! That's a really good point.

    Most of this doesn't apply to my game in particular (though again they are really good questions in general), but one thing that does is the "End Chase" range for VEs, which is about 700 pixels (two-thirds of a screen) of diagonal distance away from the VE. No time limit on the VE chasing you - it will pursue you across the entire map if you let it, but the more likely outcomes are either you achieve 700 pixels of space and it stops chasing you, or it gets pulled into an Encounter and then it's no longer on the map after you win the fight.

    It does not. I feel like this kind of system works best with simple, clear, transparent rules, and that trying to adjust distances for things like objects/other VEs in the way, while realistic, would be getting too cute.

    Do you feel that it's impossible (or nearly impossible) to do this well with randomized dungeons, then?

    A valid concern - I'll try my best to make sure my map design doesn't run into that pitfall!

    Once again, thanks for your thoughts, man.
     
    #17
  18. Tai_MT

    Tai_MT Veteran Veteran

    Messages:
    5,195
    Likes Received:
    4,370
    First Language:
    English
    Primarily Uses:
    RMMV
    No problem. I just always seem to have questions when someone explains how something works. I get this visual in my brain of how it works when explained to me, and then I start running scenarios in my head on how it could work in an actual situation as well as potential ways to exploit it.

    What I read sounded really cool and interesting, but I was wondering how it played with map design, because I could see the AI either getting very complex or the maps needing to be very simple. If the AI was complex, I wondered how that would affect map design as it'd need to be carefully thought out and somewhat intricate. But, if the AI was simple and so the map design itself had to be simple, I wondered how any of the cool features you had been talking about would work. Namely, would they even get a proper chance to shine and reach their full potential?

    Got it. :D I'll try to keep that in mind going forward.

    Okay, quick question. Do you have programming involved that keeps them from choosing the direction of say... an obstacle and walking into it for several seconds?

    It could be fairly exploitable if not. Or, a VE may end up in a "corner" where they have a 50% chance of choosing a direction that is "walk into a wall", which gives the player a lot more extra time to evade the encounter, which could affect any projected difficulty you have in evading an encounter (I have no idea how easy or how difficult you'd like it to be for players to simply bypass your enemies, I'm using the assumption that combat takes a lot of time in a timed dungeon, and evading as much combat as possible ensures better rewards if you finish the dungeon quickly).

    I think a good map for VE's is usually a mixture of both kinds. Some spots where it's cluttered, some spots where it's open, and sometimes they mesh well together, sometimes they create an interesting problem, or sometimes you know you just aren't going to evade anything due to the map layout.

    But, I enjoy a bit of variety. Not every room needs some choke points. Not every room needs a lot of dodging room to evade an encounter. Sometimes, you just need to mix various levels of clutter and open space together. All you really need to maintain is likely a rule about "multiple paths" through a space. Not just one route, but probably at least two. It gives the player a little bit of flexibility and makes them feel like they have some agency, even if you decide that this particular section, they've gotta fight at least a couple encounters before you let them go (like say, Route A has two encounters, a Wolf and a Bat, but Route B has three encounters, two of which are super weak slimes and the third is a strong Bandit. Both directions could add up to the same amount of time spent or the same amount of XP and Loot, but the player feels like they've got a choice in the matter).

    To my mind, I've been just approaching it from two angles. The first is, "I need to get to the end by using as little amount of time as possible, so evading as many encounters is preferable to engaging in a lot of combat". The second is, "I want to fight some enemies, but I want to set up the fight on my terms, so how could I manipulate the VE's to reliably produce favorable conditions?".

    I don't typically worry about players needing to run away from combat. But, I suppose some of this could translate to just needing to run away due to being unprepared and near death.

    My overall questions have been geared more towards that first aspect. You've got a timer (presumably to give the running function a purpose as well as to reward players who actually do evade combat), so I've been trying to think up ways to finish as quickly as possible. How many encounters could I evade reliably? What's the terrain like? Could I use to help me out? What are the rules of the map objects?

    I figure there's probably some places where I just would not be able to evade combat and it would eat into the timer. But, I was looking at it from the perspective of how to fight as few as possible to gain the best reward possible for making it to the end in time. Could I nail an enemy through a wall that might be a future issue and then get out of range quickly while they're stunned and never even need to "sneak by" them? Could I aggro a bunch, get 'em in a line or similar, and hit 'em with a stun that would give me time to evade them and reach the next part of the map?

    Things like that is what I was thinking of. How to use the tools you've given the player, as well as the programming of the AI and the design of the maps, to cut that time to as little as possible.

    I just asked in case you had some kind of "stealth" element in the game. If a player wanted to be stealthy, they could use an object or something to block line of sight and perhaps create an opening for a "surprise attack" or something.

    Mostly, just spitballing. Trying to figure out how the map interplays with the VE's and how I could possibly interact with it to gain an advantage.

    I mean, cover could be a good feature, but I just have no idea what your map design is actually like, so I just started thinking of possible uses of map clutter.

    I don't know if it would necessarily flesh it out or make it fun and interactive, but I think making the map clutter and map objects do more than simply block movement certainly wouldn't be unappreciated.

    I'm a little torn. I like that VE's can be snagged on objects in most games so you can use that to "evade" or "run away" or what-have-you. But, at the same time, it feels really cheap to me. Feels a little like I'm cheating. But, if you've got a good balance, then it might kind of fun that there's no reliable way to get your VE's stuck on geometry, but you could still attempt to do it and occasionally be rewarded for doing so.

    I'm not sure how other players might feel about it (whether they'd savescum a lot until the AI finally did get stuck), but I think I'd like it. AI is smart enough to evade obstacles, but still sometimes gets stuck and I get a reprieve.

    It's good to know that there's a lot of leeway in there. Each mistake means the player must think faster and faster, until... well... combat is a forgone conclusion. Seems pretty reasonable to me.

    What's the radius of enemies dragged into fights? Will it drag all 7 into a fight? Could it potentially drag more than 7 into a fight, if new enemies are close enough when you are caught? What about using an "escape" command?

    I guess, what I'm asking is: Could I gather up 7 encounters, they all bunch up together on the combat screen, and then I hit "Escape" and they go invisible for a bit to let me get away from their touch? I know many VE's will give the player some "temporary immunity" after escaping from a VE in order to give a little "leeway" to the player, so I wonder if you have that same feature and if I could exploit it in a way that would render your VE's somewhat pointless?

    Just sort of curious how that all works.

    I was thinking if the encounters didn't disappear, I could just drag everything with me through the whole dungeon, let it tag me, escape from it, then get a few seconds to essentially run away before needing to do it again. It would certainly make clearing the floors much faster than engaging in combat, as well as getting to the end of the dungeon in a good amount of time.

    It sort of depends. Without knowing the "ruleset" you'd use to place random objects/terrain down in a given area, it's hard to tell whether the map design would land on the good or bad end of the spectrum.

    By and large, I think randomizing dungeons would work very well for keeping the encounter system fresh as well as making it feel like the dungeon is more "naturally" created. However, there's a whole new slew of pitfalls to worry about depending on how this is done.

    The major ones I can think of would be:
    1. Familiar room shapes. This can get old to a player quickly. The minute the player can tell where the "edges" of each randomly spawned section occur, the player is always aware of it. It drops their experience when they know where each "section" begins and that they will inevitably see pieces of that "section" again in slightly different configurations.
    2. Accidently creating layouts that just don't mesh together well. For example, a waterfall on one section, into a river in another section, and then a cave in that blocks the river, then the river continues at the same size on the other side of the blockage with zero overflow of the water. It... tends to look strange. I'm not sure a lot of players would care about this particular example, but I'm sure there are others that would break immersion worse you might be able to think of.
    3. Accidently creating "unwinnable" dungeons due to the randomized layout. This is usually the first thing anyone using a random dungeon generator tests for, but it's still a valid concern. Especially if you are doing things like collecting keys to unlock doors or solving puzzles. The remake of "Link's Awakening" lets you create dungeons of your own from selected tiles and it does a rudimentary check to see if every room is accessible, but it does not check to see if your dungeon can actually be completed. It is possible to create a dungeon with a lot of locks and not enough access to the keys needed to open them.
    4. VE's spawning in unfavorable positions that spike difficulty too far in either direction.

    Basically, I could see each section of the random dungeon looking and feeling more natural. However, depending on how the edges are placed together for the randomized parts... You may have some issues that destroy that illusion and make the game feel more like a puzzle of some sort.
    -------
    I hope that helps a bit more.

    I've probably spent too much time thinking about VE's, how they work, how they can be improved, neat features to add to them, and such for how little I actually enjoy them. I guess I'm a strange sort of character.
     
    #18

Share This Page