Poll
Question: Should we adjust the RNG to represent THC probability more accurately?
Yes - 19 (41.3%)
No - 27 (58.7%)
Total Voters: 44

Pages: 1 2 3 [4] 5   Go Down
Print
Author Topic: Design Topic #2: RNG  (Read 25271 times)
anubioz
Neophyte

Posts: 1


View Profile
« Reply #45 on: June 05, 2018, 02:53:46 pm »

Path of Exile already did RNG for evading right: https://pathofexile.gamepedia.com/Evasion

Quote
Evasion is not purely based on chance in the sense that each hit is independent. Instead, it uses a system of "entropy" to ensure that enemies won't get long strings of hits or misses by chance. To summarize, these are the calculation steps in each attack:

    1. If it is first time an entity is attacked or if the time between the last attack and this one is larger than 6 seconds, randomise the entropy from 0–99.
    2. Calculate chance to hit of the attacker using the above formula, and add this integer to the entropy counter.
    3. If this is 100 or greater, the check counts as a hit. Subtract 100 from the entropy counter. Otherwise it is a miss and the entropy counter doesn't change.
    4. A critical hit is evaded on a separate random roll and will not affect this entropy value.

This system is designed to evenly spread out hits and misses such that players always evade an average number of times according to their ‘chance’ to evade. It is still based on chance because of the entropy counter's randomization in the first step. Note that if multiple mobs are attacking one character, all of the mobs' attacks share the same entropy counter. Also note that the chance to evade calculation on the character page is based on the average accuracy of a monster at the player's level.
« Last Edit: June 05, 2018, 03:00:26 pm by anubioz » Logged
old_school_gamer
Apprentice

Posts: 50



View Profile
« Reply #46 on: March 07, 2019, 02:16:30 pm »

1)   Should we rig the RNG to meet players’ expectations?
2)   If yes, how? Meaning what should we aim it? What outcomes should never ever happen when your THC is 70-80%?

I don't know how to vote exactly since I'm in a demographic that loves plain old RNG that isn't rigged, but if I find fault, it's in the way and context in which RNG is used.  lol Used correctly in a setting like combat, I see it as a way to expand on the uniqueness of an experience and the freedom of a player as well as expanding on the variety of outcomes that can unfold instead of looking at damage numbers and misses all day. It is not interesting to me to take a few turns on a repeat of a battle and the AI does the exact same thing and the exact same thing happens, and I never liked the idea of moves being strictly "correct move" vs. "blunder" and prefer to think in terms of "risk vs. reward" and especially avoiding the moves that maximize risk while minimizing reward.

One of the reasons I love tactical combat with a positional factor and a grid/board is because it tends to favor a shorter but more lethal type of combat where every turn counts. I'm not spamming attack/heal/revive/buff/debuff over and over in such a repetitious way, and instead putting more thought into each move. I like when every turn feels important. And such games make choices like melee vs. ranged and AoE vs. single target so much more interesting and meaningful. If every move feels very meaningful and not repetitious and requires very careful thought with every single turn (especially in challenging battles where my characters are outclassed), then I tend to be a happy camper. If I feel like I'm just spamming the same sequence of actions repeatedly and eager to get the battle over with, then I tend to not be as happy.

I also don't mind missing several times in a row with 70% chance to hit. That's not so improbable from a mathematical perspective (almost 1/30 times), and the analogy of weighted dice (ex: something which produces a bell curve instead of a horizontal line when graphed) could risk making the experience feel a bit more "static" if it's taken very far. I think we're psychologically biased to thinking that happens more than one in thirty times even if gathered statistics would reveal this to be the case, since we tend to associate more emotion to unlucky events and recall them more vividly than when we luckily hit an enemy 3 times in a row with 30% chance (something I imagine would also no longer be possible if the RNG used an algorithm which avoided multiple consecutive numbers in a row).

What I actually find tedious and frustrating is not because I missed several times in a row, but because I'm bored just exchanging damage numbers and hit points with an enemy repetitively (and I was rarely ever bored in AoD combat). If a particular game lacks interesting events that could happen beyond hit/miss, like a game that lacks a variety of criticals, or criticals are so bland and just do double damage each time and nothing else, then I'll easily get bored more easily with more heavy RNG involvement because I just want to get to the next part of the game. Meanwhile when a game is so random that it has such a wide range of things that could occur, however improbable, I find myself constantly amused while adjusting my whole tactics to what happened. In particular I especially enjoy games where losing in the most unfortunate way is entertaining in itself (ex: Blood Bowl where my opposing team managed to throw a halfling halfway across the field and score a touchdown one turn before the game ended). I can't help but laugh at that astronomically improbable situation even though I'm a bit frustrated at the same time.

This idea is at risk of frustrating certain demographics even more but one way to mitigate the repetition of staring at misses and damage numbers for lengthy periods of time is to actually make combat even more lethal in nature, where a single hit or two could take down even the most powerful character. That would shorten the experience and make it more gripping and tense, but for those who base fun factor on their ability to win as consistently as possible, it would probably frustrate that demographic even more. For me it'd likely heighten the amusement and tension of every single move and make it so I'm thinking even more carefully about every single move I make.

Another that involves tackling elements around the RNG rather than the RNG itself is to put more emphasis on the strategic elements of gameplay being around the types of moves made where no RNG is involved (putting emphasis on allowing many risk-free moves before taking a risk). I learned very quickly from Blood Bowl that trying to do things that risk causing me to waste an entire turn like running to grab a ball which is subject to RNG is a terrible beginner mistake. Meanwhile positioning my units involved no RNG (unless they were moving away from a tackle zone). So the deep element of strategy in that one in spite of its incredible involvement of RNG is tactically positioning your units into an ideal position which isn't subject to any chance before taking any chances, and that's where the heavy skill factor comes in where skillful players can start to win far, far more consistently than lucky players. X-Com is similar that way but doesn't take the idea quite as far as Blood Bowl.

Yet another that could make even easy battles more interesting is to have more varied outcomes than black and white "win vs. lose". I vaguely recall some in AoD which were like that where I was a fairly good fighter but a very fragile NPC was involved, for example, so even if I easily won the battle, I might have failed to save the NPC and that becomes something unique and interesting. Jagged Alliance and Darklands had that sort of thing where wounds were very difficult to heal even outside of battle, so even an easy battle that lead to heavy wounds was interesting and somewhat costly to the player even if he won. X-Com had that somewhat as well since the loss of squad members or even the entire squad didn't lead to game over but was a costly setback. Fallout had some permanent damage effects (I remember permanently losing an ear to a fighter who curiously resembled Mike Tyson in a boxing match). That makes victory/loss more gray and not strictly black and white, and that grayness might help players more readily accept their fortunes and misfortunes.

What I'm trying to get at (forgive my verbosity) is that I tend to feel like complaints about RNG might have to do more with surrounding factors than the nature of the RNG itself. It is like if players complain that an RPG lacks a respec option, perhaps the game lacks a sense that every choice the player can make is interesting and worth accepting and living with the consequences.
« Last Edit: March 07, 2019, 03:14:04 pm by old_school_gamer » Logged
menyalin
Artisan

Posts: 502


View Profile
« Reply #47 on: March 12, 2019, 07:32:17 am »

Reading previous post and come to some thoughts about RNG badness. What RNG situation deal the biggest pain? I'd call two types:
1) RNG wall, when you need to repeat something again and again to progress.
2) Really long engagements when you fall near end just because of bad luck.
So if you making some hard RNG balance, it can be better to make shorter engagements or possibility to save\load somewhere during it. Maybe add posibility for save-load during battle on medium to low difficulty? It will ease the life or intermediate player on one side and hardcore gamers not so picky about RNG hardness on other side.

What I'm trying to get at (forgive my verbosity) is that I tend to feel like complaints about RNG might have to do more with surrounding factors than the nature of the RNG itself. It is like if players complain that an RPG lacks a respec option, perhaps the game lacks a sense that every choice the player can make is interesting and worth accepting and living with the consequences.
Completely agree.
Logged
Wizard1200
Colonist
*
Posts: 743


View Profile
« Reply #48 on: March 14, 2019, 08:11:26 am »

Another that involves tackling elements around the RNG rather than the RNG itself is to put more emphasis on the strategic elements of gameplay being around the types of moves made where no RNG is involved (putting emphasis on allowing many risk-free moves before taking a risk). I learned very quickly from Blood Bowl that trying to do things that risk causing me to waste an entire turn like running to grab a ball which is subject to RNG is a terrible beginner mistake. Meanwhile positioning my units involved no RNG (unless they were moving away from a tackle zone). So the deep element of strategy in that one in spite of its incredible involvement of RNG is tactically positioning your units into an ideal position which isn't subject to any chance before taking any chances, and that's where the heavy skill factor comes in where skillful players can start to win far, far more consistently than lucky players. X-Com is similar that way but doesn't take the idea quite as far as Blood Bowl.

A game can be interesting without RNG in my opinion. Two examples:

Invisible Inc.:


Tactical Breach Wizards:
« Last Edit: March 14, 2019, 08:14:36 am by Wizard1200 » Logged
Scott
Developer

Posts: 2415



View Profile
« Reply #49 on: March 14, 2019, 10:42:10 am »

I don't have an issue with "long" engagements. All out, room-clearing brawls are some of the most memorable fights of all. OTOH, drawn out exchanges between low-talent or high-defense fighters can be a drag. Vince may already have addressed this, but what about scaling aim bonuses after multiple misses. For example: two misses, give the next swing/shot +5% and additional stacks until a hit is made and the counter is reset. This would apply to player and enemy of course, and isn't that farfetched. I assume someone firing at a stationary target is eventually going to come at least slightly closer after several shots.
Logged

contributed to: Age of Decadence | Dead State | Dungeon Rats | Battle Brothers | Fell Seal:Arbiter's Mark | Stygian
working on: Colony Ship RPG | Encased | ATOM RPG | Realms Beyond | Fell Seal:Missions&Monsters | Urtuk: The Desolation
Wizard1200
Colonist
*
Posts: 743


View Profile
« Reply #50 on: March 14, 2019, 02:31:31 pm »

Vince may already have addressed this, but what about scaling aim bonuses after multiple misses. For example: two misses, give the next swing/shot +5% and additional stacks until a hit is made and the counter is reset. This would apply to player and enemy of course, and isn't that farfetched. I assume someone firing at a stationary target is eventually going to come at least slightly closer after several shots.

I would go one step further:
+ 5 % THC for every attack after the first against the same target. Stacks up to 3 times (5 times with a feat) and resets as soon as you attack another target or reload a weapon. This would make fast weapons more useful, melee weapons more useful, Perception less powerful and larger groups more dangerous.
« Last Edit: March 14, 2019, 05:07:24 pm by Wizard1200 » Logged
old_school_gamer
Apprentice

Posts: 50



View Profile
« Reply #51 on: March 23, 2019, 03:55:51 pm »

A game can be interesting without RNG in my opinion. Two examples:

I have to check those out. They look very interesting to me but one of things I found somewhat unsatisfactory in general for my personal preferences when RNG is eliminated outright is when I repeat the same battle starting off with the exact same move before, and the AI does exactly the same thing in response because there's no RNG involved. That level of knowing exactly what the AI will do in response to my moves starts to make it feel so much like a puzzler to me.

I think the reason is because I start to feel heavily constrained in terms of player freedom. There is no longer choices like, "That was a good move -- low risk, high reward, but it failed. I would repeat that move again if I had to repeat the battle because it was a good move even though it failed. Now let's think of plan B for this battle in response to the fact that it unluckily failed." Instead it's like, "That move failed, therefore it is a bad move. In subsequent playthroughs, I will no longer make that move ever again," and that feels so black and white -- I like something more fuzzy and grey.
« Last Edit: March 23, 2019, 03:58:21 pm by old_school_gamer » Logged
Wizard1200
Colonist
*
Posts: 743


View Profile
« Reply #52 on: March 23, 2019, 04:03:19 pm »

Invisible Inc. is a tactical puzzle in my opinion, but the the risk vs. reward management is very important and i have lost many games, because i was too greedy Grin

Tactical Breach Wizards is currently in development.
« Last Edit: March 23, 2019, 04:08:14 pm by Wizard1200 » Logged
old_school_gamer
Apprentice

Posts: 50



View Profile
« Reply #53 on: March 23, 2019, 04:10:45 pm »

Vince may already have addressed this, but what about scaling aim bonuses after multiple misses. For example: two misses, give the next swing/shot +5% and additional stacks until a hit is made and the counter is reset. This would apply to player and enemy of course, and isn't that farfetched. I assume someone firing at a stationary target is eventually going to come at least slightly closer after several shots.

I never played Hard West but from what I've gathered of videos watching it, it has a system like that to discourage players from taking cover in one spot indefinitely. It makes a whole lot of sense in that one that ducking under one piece of cover indefinitely would eventually make you more and more susceptible to enemy fire due to your stationary nature, and instead encouraging the player to take cover temporarily in one spot and then move to another in an intuitive way.

I think it's trickier to make that as intuitive in one that has a lot of focus on melee combat in a turn-based grid format. In that case it might not actually make as much sense for a stationary target to become increasingly easier to hit the more attempts are made (unless the target is becoming fatigued), since a stationary target isn't really stationary as I gather. The animations and nature of the pieces on the board might not show it but I assume they're dodging and weaving and constantly moving even though they're standing on the same grid cell, otherwise it makes no sense that anyone would ever miss. So I assume there's a lot of things being simulated behind the scenes, so to speak, in those. In general I think it's much more challenging to make melee combat nearly as interesting as ranged in such turn-based tactical games, since they tend to diminish the importance of positioning, line of sight, terrain, cover, etc.
« Last Edit: March 23, 2019, 11:23:04 pm by old_school_gamer » Logged
old_school_gamer
Apprentice

Posts: 50



View Profile
« Reply #54 on: March 23, 2019, 05:04:24 pm »

Invisible Inc. is a tactical puzzle in my opinion, but the the risk vs. reward management is very important and i have lost many games, because i was too greedy Grin

I have to give it a try. I developed some strong thoughts in favor of the RNG somewhat in response to reading blogs from game designers who thought it was horrible, to which I half-agreed initially, tried their games completely absent RNG, and felt like it wasn't my cup of tea... then I swung around in the opposite direction adoring the most chaotic use of RNG like Blood Bowl and X-Com (though always with some underlying structure and skill factor involved for those who can navigate probabilities), and came to realize that most of my favorites tended to use the RNG quite heavily (though in more interesting ways than variable damage and hit/miss).

Mostly I've come to develop the odd thought that the RNG is a way to expand on player freedom -- there's more freedom to navigate choices when they aren't black-and-white "right vs wrong" but instead "risky vs safe". As a most simplistic example, take a lock which requires 7 lockpicking skill to open. That's black and white. Either your character has the adequate skill to access that content or he doesn't (say he has lockpicking at 5) and it's impossible for him to access unless there's another way to get there. Meanwhile if we throw in the RNG, we might make it so the lock is difficult and you need lockpicking 7 to have a very good chance of opening it, but lockpicking 5 might suffice to have a coin toss chance of opening it still but at the higher risk of the lockpick breaking (and perhaps lockpicks are very expensive). Now it becomes possible for that player with moderate lockpicking skill to open the lock, but he's taking a bigger risk than someone with higher lockpicking skill... and now it doesn't feel so black and white, and I feel like player freedom is enhanced and not quite as constrained to doing things such a certain way or building such a specific build as a result.

I like thinking about things in terms of risk vs. reward that way. Feels closer to my daily life where nothing I do is guaranteed to succeed, not even typing this response (I might fall out of my chair, knock myself on the head and die at any moment). So I like thinking about risk vs. reward constantly and estimating, if not being given, the probability of succeeding vs. failing at something, as well as trying to arrive at backup plans for when actions don't succeed.
« Last Edit: March 23, 2019, 05:13:57 pm by old_school_gamer » Logged
Wizard1200
Colonist
*
Posts: 743


View Profile
« Reply #55 on: March 24, 2019, 06:23:20 am »

I never played Hard West but from what I've gathered of videos watching it, it has a system like that to discourage players from taking cover in one spot indefinitely. It makes a whole lot of sense in that one that ducking under one piece of cover indefinitely would eventually make you more and more susceptible to enemy fire due to your stationary nature, and instead encouraging the player to take cover temporarily in one spot and then move to another in an intuitive way.

A character could be encouraged to move if he gets an evasion bonus of 1.5 % per tile of movement. With the current movement cost of 1 AP per tile a character could get an evasion bonus of up to 30 % if he has a Dexterity of 10 and uses 20 AP for his movement.
« Last Edit: March 24, 2019, 06:25:57 am by Wizard1200 » Logged
old_school_gamer
Apprentice

Posts: 50



View Profile
« Reply #56 on: March 25, 2019, 02:19:48 am »

A character could be encouraged to move if he gets an evasion bonus of 1.5 % per tile of movement. With the current movement cost of 1 AP per tile a character could get an evasion bonus of up to 30 % if he has a Dexterity of 10 and uses 20 AP for his movement.

I love that idea and it makes a whole lot of sense to me for ranged exchanges, perhaps combined with an evasion penalty that accumulates to some extent if the player is stationary to address the frustrations of those who miss many times in a row.

The stumbling block to me is opportunity attacks. I think opportunity attacks make a whole lot of sense in games like these to counter those who might otherwise kite with ranged and max AP indefinitely if not for it. When we want to start making melee vs ranged more viable, it starts to become desirable to have more ways for the melee unit to be able to pin the target to one cell on the grid with heavy risks for moving in those cases. One possible way to counter that might be to make ranged attacks and reloading really costly in terms of AP but more devastating in terms of the damage they deliver; that might prevent the kiting strategy without involving an opportunity attack system.

So it's like on the one hand we want to discourage stationary exchanges, but on the other we want to encourage it, and it might be difficult to find a happy medium there. The reason I think they effectively do those evasion penalties for being stationary for long periods in Hard West is because it is predominantly about ranged exchanges. It's not the type of game that's trying to make charging at someone with a machete who is firing at you a viable way to fight (it's generally completely suicidal).

In general I think trying to make melee viable vs ranged is tricky like that. Line of sight might help sometimes since it could allow a melee unit to sneak up to a ranged unit by staying out of their line of sight. A party/squad-based system can also make that a bit more viable than ones that make playing solo viable.
« Last Edit: March 26, 2019, 06:25:57 am by old_school_gamer » Logged
Fed
*
Posts: 162


View Profile
« Reply #57 on: April 05, 2019, 02:42:29 am »

I think it would be a fair model - to improve your current THC on the same target, on a miss, basing on your base THC.
When skilled people miss, it's often not a complete waste. The next try does not come as if there was nothing before.
They probe the defence, adjust their aim, guide their target in a aposition where it's harder to defend.
Logged
dogwaffler
Novice

Posts: 26


View Profile
« Reply #58 on: April 10, 2019, 03:21:11 am »

in reply to OP:

I think that there is a market for unforgiving seeds, but they tend to require constant moderation.
Logged
Kitsune
Neophyte

Posts: 1


View Profile
« Reply #59 on: June 14, 2019, 02:26:09 pm »


For the first question, it depends on how combat plays out in the demo.

For the second question, what probably shouldn't occur is letting the player feel cheated. It may be a psychological trick, but if players feel that they are being cheated by the system, then they may be less likely to participate.
Logged
Pages: 1 2 3 [4] 5   Go Up
Print
Jump to: