Should I just use Random Encounters?

[hian]

Hian - I'm curious, what's your opinion on RNG in games?  i.e. hit / evade / crit chance

That's a tough question.


I can't really speak for other people, and I think how people feel about a games math is kinda hard to


deduce on the fly.


I'd make the argument from a indie dev perspective, that you should keep your math simple, with low digits and


as few factors going into calculations as possible, for the sake of making your game easy to balance, and


easy to test-play.


The bigger the number, and the larger the amount of factors, the more issues you're bound to get, and the more


time will be demanded of you doing battle balancing.


This is time and energy consuming and is likely to put you on a path of getting tired of your project.


In my games I usually set all base hits on attacks and skills to 100%, while


keeping the base evasion of all enemies on 0 (or 1 if it's necessary to make the base formula work), and let the stats


take care of the rest.


I try to keep misses and crits to a bare minimum for ordinary attacks and skills in order to make it easier for the player


to make their own calculations and feel like the battles conform to their expectations and deductions


rather than blow up in their faces because half of their attacks suddenly miss,


or because the enemy suddenly gets 4 crits in a row.


That's not fun, nor interesting - it's just frustrating.


If I am going to lose or win a fight - it's going to be because I deserved it on the merit of my battle-plans,


and my level, not on the basis of some fluke where the computer threw a dice that randomly put me at a


disadvantage/advantage.


I only make large hit/crit discrepancies in very specific circumstances, such as giving high evasion to specific


enemies like small creatures, or flying creatures, and make up for that by providing a party-member that


has stats/weapons to deal specifically with such enemies.


Similarly, I only give increases in crit-chance to the party, never to the enemy.


And when I allow for party-members to have increases in crit-chance,


it's always balanced against a reduction in hit percentage, or with a weapon/skill that comes with a cool-down.


My personal design philosophy, based on experience, is that people very seldom find loses due to flukes to be


anything other than frustrating, and therefore one should generally try to avoid


math that allows for flukes to happen in your game.


Making sure that misses/evasions and crits don't happen except in very specific instances when you want them


to happen, is an important part in making your game-play predictable, and that's important because predictability


is the fountain of game-strategy.


Imagine how games like Chess would be ruined entirely if you started introducing chance elements like those


we often find in RPGs.


"Oh no, today your Bishop can only move half of it's intended distance"


"Why?"


"Because I threw a dice, and got a 3 instead of a 4 or above."


"Eeeh, ok..."


But that's just my opinion. A lot of people love games of chance. That's why slot-machines and betting is


so popular after all.


Again, it depends on what kind of person you are, and what kinda people you're making your game for.

 

[Iavra]

Personally, i would completely remove hit/evade and implement an ATB/PTB to give "agile" characters another advantage. It's easier to balance and feels better in my experience.


Crits, on the other hand, can feel nice when they happen, so they can be left in if they are carefully balanced.


Also, change blind to make every physical attack miss.

 

[headdie]

Badly balanced RNG is one of the biggest evils you can put onto a game because then the game stops being about player skill and about the number generator which is only fun to slot machine gamblers.

Subtle and well balanced RNG on the other hand adds excitement to the game, having 50 hp left and knowing your opponents next attack is 50 +/- 5% could be the difference between win or loss, it adds excitement but 5% also means that the battle mostly came down to player skill in making the right call on attacks.

@Iavra

Thing is, there is a chance you might hit, end of the day blind dosnt stop the attack from happening which means you need to take into account that a hit can happen and that damage might be done when it does which is why a modifier on chance to hit is more believable than preventing the hit at all.

 

[Iavra]

If i want blind to make every attack miss, i can implement it this way.

 

[headdie]

Indeed you can

 

[dungeon diver]

I think part of what makes misses on damaging attacks and skills so frustrating is that they're all-or-nothing. I think I'd prefer grazes to misses, where it deals a fraction of the usual damage but it's still damage.

Condition etc. skills are a little different, I'd say whether or not their hit chance is frustrating depends on how much of an advantage they give you. i.e. a powerful poison effect that takes 25% health every turn doesn't cause as much frustration and feeling of usefulness when it misses because you buy into that gamble knowing it will pay off over time if it hits.

 

[headdie]

Information like that certainly helps remove frustration because like you say it is a decision the player is making with the important information

 

[bgillisp]

One thing I've noticed in the past when this comes up is people don't like RNG affecting what the player can do, but then get excited when they do doge an attack that is meant to hit them. Way I look at it is, make it the same to both the enemy and player. If the player can miss 25% of the time as the default, so should the enemy. Otherwise, one side is just cheating.

The biggest problem with the RNG I've seen in games is most times developers will gimp what the player can do, but forget to match with the enemy. So they will give the player a 75% chance to hit, but the enemy is 95% (both as defaults. Obviously, there might be exceptions for unique monsters).

As for my game, what I've done is I left the RNG in for attack, as it is a free skill. However, magical attacks (which cost MP) almost never miss. I say almost as there is one skill which blocks all attacks for one turn, but you can plan around that. So now it is a choice between take a chance on a miss with a skill that costs nothing, or burn MP and have a (almost) guaranteed hit?

As for CRT, I have one mage who can buff it, and another who can negate it. So you can cast that spell which drops enemies CRT chance to nothing for a couple turns, but some enemies (and actors) can boost their CRT chance too (watch out for those!). By this, I hope to give the player a chance to handle the RNG in the game and make it manageable.

I do hope to write a script at some point to overcome how variance works. For instance, remember the Bless or Curse spells from HOMM? Cast those and you do the maximum possible damage (in the case of Bless?). Don't like the RNG? Just cast bless. It's cheap too.

 

[jonthefox]

Way I look at it is, make it the same to both enemy and player. If the player can miss 25% of the time as the default, so should the enemy. Otherwise, one side is just cheating.

Funny you mentioned this, because I feel the same way--when I balance my game, I set up an excel spreadsheet with both players and enemies, and I derive all of the stats based on my intended power levels.  However, this always poses a problem for me when it comes to this issue.  I know as a player how good it feels when you get a critical hit, especially if you have the option of following a class/skill/item path that allows you to increase your critical hit chance or damage or states that occur when you land a critical hit, for example.

Yet, on the other hand, let's say you're venturing out to a dungeon you're not ready for, or you're on the right quest/dungeon but for whatever reason you're just underleveled.  Now let's say you get into a fight with one of the tougher enemy troops in this dungeon, and they crit you...welp, the player is probably going to feel not just frustrated that he died, but also because the enemy "got a lucky crit" - even if he was legitimately treading too deeply for his power level, it's going to feel like he just died because of RNG.

So that makes me entertain the idea, maybe I give players hit chance and crit chance, and I give enemies eva chance and cev chance.  This way, I can sort of balance out their power levels without one side having a totally unfair mechanical advantage over the other--but, while a rare miss might be annoying for the player 1) it's unlikely that this would cause the player to end up dying because of it 2) having some decent eva chance would be a way to distinguish rare but supposed-to-be-powerful enemies 3) the player still has a way to combat this through increasing his hit chance (items/skills) 4) the player enjoys the benefit of uber critz.

What do you / others think?

 

[Another Fen]

It depends on the situation I think. When an action has a 90% chance to hit, I'd probably expect that action to hit and would feel relieved/betrayed if it doesn't (depending on the party in whose favor that miss was in). If the user is blinded and has a hit chance of 10% instead, it would be the other way round.

Usually when dealing with a RNG based battle system, you can take the risks of a worst case scenario into account. If you still decide to use your supporters turn to buff the parties ATK instead of healing the damaged actor from half health and that actor gets critted to death as a consequence... Well, you still could have prevented that unlikely scenario.

However, misses can be a problem because when they occur your action does completely nothing. You can't really plan with a worst case scenario because that would be an infinite combat. Especially when you have to rely on certain actions you have no "guaranteed hit" alternative for, a miss can be devasting while a non-critical action has at least some effect.

This could be a reason why failable healing spells are quite unpopular.

(Have to continue later)

 

[bgillisp]

@Anotherfen: Why? 90% is a miss 1/10 of the time. The player should expect 1 miss in 10 with that number, which means in an average fight, if you hit attack all the time you will probably miss one of those attacks. Personally I don't think you should feel betrayed, but should expect that 1 or so miss to occur about every battle with that number you listed.

@jonthefox: Might work, as long as it is explained well to the player. Myself, I've been running with normal attacks having a 75% chance to hit, but if you up your DEX (I replaced LUK with DEX) your to hit increases. Also there are accessories and spells which boost it. Since the cap is 100%, this means if you wish to use attack a lot, you will want to boost your to hit. If you wish to ignore it or use it as a last resort, then ignore that stat and focus on the others.

Personally, I think some players just want a free skill that always works. Attack costs 0 MP (in most systems) so shouldn't it have a disadvantage too, to balance it out? If we make it always hit and always do the same damage as a spell, then why would I ever cast?

 

[Another Fen]

"Betrayed" might be a bit rough to describe it though... "Disappointed" would have suited better.

Usually I don't really count the hit/miss ratio of an attack, so 90% hit chance would probably just fall into the "most likely to succeed" category. Of course misses do still happen now and then but a particular use of that attack can be usually expected to succeed.

 

[bgillisp]

Fair enough. Maybe I'm also weird because I grew up on the old D and D gold box games, where a to hit rate of 50% or lower was normal. The only thing is, the monsters weren't (usually) these HP bloated beasts we see today that are 5 - 30x the HP of the average player, instead they had HP near yours, and as a player you had to increase your to hit to be able to defeat them. And, if you got your armor class high enough, the enemy would miss you often, making them easier to handle.

 

[hian]

@Anotherfen: Why? 90% is a miss 1/10 of the time. The player should expect 1 miss in 10 with that number, which means in an average fight, if you hit attack all the time you will probably miss one of those attacks. Personally I don't think you should feel betrayed, but should expect that 1 or so miss to occur about every battle with that number you listed.

 
The problem is, of course, that each calculation is done in vacuum, so it's not like if you miss once,


you're guaranteed a hit next time around.


Even if the chance is low, it's still chance, which means that you might actually end up missing 5 times


in a row, even with a 90-95% hit-rate.


Such misses might ruin the player's entire game-plan and get their character(s) killed. That will feel like the


system is stacked against you, even if, mathematically speaking, it really isn't.


That's why players will get frustrated - because, to be fair, they don't understand probability.


When they see the number 90%, they assume, like you, that they're basically bound to hit 9/10 times


however that's often only going to be true if you hit a 100, or maybe even a 1000 times.


With that ratio(90%), and a small sample selection, it is however quite possible to get


misses several times in a row.


Another thing that messes up the mechanics is "shadow maths" I.E factors that go into the hit/miss


calculation that isn't disclosed in in-game text.


RPG-maker games often suffer from this issue,


such as when you're using Yanfly's parameter scripts that change


up the hit equation -


Put the hit rate to 100% in a character-class, and remove


the evasion parameter from an enemy entirely, set up a test fight and notice that you'll still


be missing quite a lot.


In fact, I've just experimented with the base parameters and noticed that even against an enemy with


no EVA at all, you still need somewhere around 115% Hit (make to ex-parameters for hit, one being 100%


and the other being 15%) in order for you character to hit consistently and only miss something like


once in a while instead of several times during a relatively brief encounter.


That's extremely counter-intuitive, because one would assume that having 100% hit-rate against a


0% evasion enemy would mean that you hit 100/100 times.

Personally, I think some players just want a free skill that always works.


Attack costs 0 MP (in most systems) so shouldn't it have a disadvantage too, to balance it out?


If we make it always hit and always do the same damage as a spell, then why would I ever cast?

Why wouldn't players want a free skill that always work? And why would you want to take


away what the player wants, especially something as integral to game-play as a


consistently functioning base-attack?


The balance of attacks is that they generally don't do anything close to as much damage


as skills and magic, except perhaps in the end-game. Some enemies might not even respond much


to physical attacks (such the Flans in Final Fantasy)


If the attack is likely to miss as well, you've just lampooned the player's only consistent albeit


limit measure of dealing damage consider that MP etc. is a limited resource.


You cast spells and use skills because if you're fighting an enemy with a lot of HP and strong attacks,


you're party would be dead long before they could finish the enemy off with regular attacks.


If your attacks are likely to miss a lot, the question instead becomes, why would you use or rely


on attacks? If they're unpredictable they can't be counted on, and won't be useful as a prominent


feature in the battle-plan of the player.


I can only speak for myself personally - while I think unpredictability is important for making


a game exciting to play, I think the unpredictability of not knowing what my enemy is going


to be throwing my way, is more than enough, because it's still something you can prepare for


with proper strategy.


However, I shouldn't also have to worry about whether or


not my skills are even going to work once I decide the use them - because that's not something I


can prepare for, unless you provide me with certain skills that don't miss that I can use


as a back-up.

 

[bgillisp]

One thing to keep in mind is the default in RPGMaker is odd. The system first computes if you hit, then it computes if you evade. So if you have 90% to hit and 5% evade, you really only have a 0.90 * 0.95 chance to hit, due to that.

I actually rewrote the way to hit and evade was computed as I didn't like that myself. Something all of us will want to remember though if we stay with the default though.

As to why you use attack (or a skill) even if it can miss, there are two reasons:

1: Last resort. Maybe you are out of MP, and/or are desperate.

2: Maybe it has a chance of inflicting a status aliment that if it hits, can really turn the battle?

And that brings me to my point on RNG. I don't think anyone here has made a game that has all their status aliments always work. Which means, if your game has status aliments, you have RNG in your game. That is, unless all your status aliments always hit.

 

[hian]

One thing to keep in mind is the default in RPGMaker is odd. The system first computes if you hit,


then it computes if you evade. So if you have 90% to hit and 5% evade, you really only have a 0.90 * 0.95 chance to hit,


due to that.


I actually rewrote the way to hit and evade was computed as I didn't like that myself.


Something all of us will want to remember though if we stay with the default though.

It's also worth remembering that whatever equation you do end up opting for, the player usually won't know what it is -


unless you find a way of telling the player through your game.


Furthermore, the more complex the equation is, the more of a bother it's going to be implementing that info


into the game (such as in the equip menu etc.) and the more bother it's going to be for the player to do the math


on his or her own.


Herein lies the problem - because if it isn't readily apparent to the player what their chance of hitting is,


and the hit-ratio is set to a percentage that allows for a lot of variation given the small sample range of


attacks in a brief encounter, the attack becomes a real "hit and miss"(sorry for the pun) option.

And that brings me to my point on RNG.


I don't think anyone here has made a game that has all their status aliments always work.


Which means, if your game has status aliments, you have RNG in your game. That is, unless all your status aliments always hit.

Nobody has said RNG is all-out bad, but that it necessitates good balancing, and that large discrepancies on the hit-rate


of your base-attack is bad.


It does not follow that that having RNG at all, such as in the case of status effects,


somehow means you might as well have high hit-rate dips on your base-attack.


These two game-play elements play very different roles in a game.


Status effects are commonly additions to the overall strategical nature of combat, while the attack is the base-line.


It's supposed to be the stable, root element of your offense. It doesn't have to be, but if it isn't, at least you should


provide an alternative that fulfills a similar role.


Also, as a side note - I have both made and played games with 100% hit-rate status effects.


In these games, status effects can only be stopped through the use of preemptive measures, such


as barrier-type spells, or equipping certain items/weapons/armor.

As to why you use attack (or a skill) even if it can miss, there are two reasons:


1: Last resort. Maybe you are out of MP, and/or are desperate.


2: Maybe it has a chance of inflicting a status aliment that if it hits, can really turn the battle?

And both of those are gambles. Unless you want your entire battle system to revolve around chance,


it needs certain (at least one) stable elements. Stability is the root of strategy.


Now, I haven't said that attacks should never miss - I am simply saying that your attacks should hit reasonably often,


so that the player can comfortably rely on them as stable offensive moves.


If the player can't, and has no other option either - the player has no reliable strategy to rely on in combat.


Each move is a potential failure on the alter of chance - which the player may or may not enjoy.


Personally - if I wanted that level of chance, I would be playing slots, not RPGs.

 

[bgillisp]

From what I can see in your post, it comes down to this for you: Player control. If you give the player a way to control the to hit/evade rates/critical rates, etc, then it can work. D and D used a system that had a (by us) low to hit on average (often it was 50% or lower for opponents of equal level) but players loved it! Why? Because they knew the numbers and could control it, so it didn't feel like as much of a gamble.

Where I think many people go wrong is they hard code a low to hit, and leave it fixed, so it never can go up, and the actor never improves over time. So the entire game, you will always have a 80% (or whatever number you hard code) to hit, and you will always miss 20% of the time. However, if the player knows about it, and how it works, then it can work. It's harder to do in RPGMaker, as you will need to edit the default system to do something like "Boost to hit by 5% per level up" (which was what D and D commonly did, but it was called attack bonus or THAC0 depending on your version). Then, if you miss often, go get another level or two. Or, cast Bless (which made your to Hit go up). Being hit too often? Cast Curse, which lowers enemy To Hit.

I think in the end, here is where we will agree. If you use the default system, then you need a high to hit. It just doesn't work well otherwise (Personally, I'd recommend a 100% to hit, and only use evade on nimble enemies on the default system. Then you only miss if Blinded). However, if your monsters depend on having low HP (or even same HP as the player) and defending well, then a lower to hit and/or evade can work, because then the system depends on finding ways to score reliable hits instead (or hoping for a miracle). But, if you use those HP bloated monsters that exist in 99% of the games today, a low to hit will be a turn off, and probably make the fight impossible.

It all depends on the system being used, more than anything. I think we are trying to look at it as a universal thing, and no matter what it won't work that way. Everything has a place.

Back to the OP's question: I think it is going to come down to your battle system. Maybe study up how some games implemented this in their battles? D and D is a popular one from the 80's and 90's. FF is another, though their numbers are not as public as D and D's are. Wizardry is another one, though the last official Wizardry game came out in 2001. Or, can you tell us more on what you are thinking for your battle system? Maybe we will have (differing) opinions on what you should do?

 

[Another Fen]

You probably wouldn't gamble on a 50/50 chance to lose the entire fight while having a success rate of 90% or even 95% can make a miss occur much more unexpected. I usually don't mind random misses, but a miss can still hurt if you finally collected enough TP for a powerful attack, got an active ATK buff and a rare "guaranteed critical" turn flag just to miss the 90% success chance, even though it didn't really change the outcome of the fight.

Usually, I think it hurts less if you actually have the chance to gamble or not. If you have the choice between perhaps finishing off an opponent or certainly restoring your party and wait for a better opportunity, you are at least to some degree at fault when you try the former and fail. If healing the party would have also had a similar failure chance you simply couldn't have done better.

Actually I don't think giving enemies a ton of HP requires a high hit chance. After all, in theory the more often you have to hit an opponent to finish it off, the less impact each particular attack usually has. I'd probably be more annoyed if I miss a foe several times that would normally be dead after one or two attacks (unless I am at fault of course).

In games where you achieve incredible high hit counts with most attacks for example, having a certain rate of misses should not really matter.

 

[bgillisp]

Yeah, 50/50 only works in a few cases honestly, and it really depends on 1: what the skill/attack is, and 2: your battle system. My D and D example for instance, there is no DEF stat, so you either dodge the attack, or take the full damage from it (well, ok, damage was in a range like 4 - 12 damage, but that was randomly determined). So in that case, the system was dependent on you being able to dodge damage totally, and reduce everyone else's chance to hit you and/or increase your chance to hit.

Of course, this does get annoying when your Level 1 Wizard loses a battle to Mr. Whiskers the house cat because you can't hit the broad side of a barn.