next »

[JasonPickering]

Hey guys. just wondering how people feel about chance in RPG combat.

Chance in RPG combat dates back to D&D when dice were used to decide outcomes. well I have been thinking about that a lot and came up with a simple game mechanic I think.



an enemy has a given number. you Roll a die and must get a number equal then or larger to theirs. if you win they take a hit if you lose you take a hit. now you can spend mana to help yourself out. so in the above picture one sword is active that means whatever the player rolls is +1 (ignore the number above in the picture). Mana is a limited resource however and must be saved so you could click all 5 swords but do you really need it. I want the player to have some control of combat, but still have to deal with chance. this sound good? any better or different ideas? how do you feel about chance in combat?

[namragog]

Hey guys. just wondering how people feel about chance in RPG combat.

Chance in RPG combat dates back to D&D when dice were used to decide outcomes. well I have been thinking about that a lot and came up with a simple game mechanic I think.



an enemy has a given number. you Roll a die and must get a number equal then or larger to theirs. if you win they take a hit if you lose you take a hit. now you can spend mana to help yourself out. so in the above picture one sword is active that means whatever the player rolls is +1 (ignore the number above in the picture). Mana is a limited resource however and must be saved so you could click all 5 swords but do you really need it. I want the player to have some control of combat, but still have to deal with chance. this sound good? any better or different ideas? how do you feel about chance in combat?

Here is an equation from pokemon:
((2A/5+2)*B*C)/D)/50)+2)*X)*Y/10)*Z)/255

A = attacker's Level
B = attacker's Attack or Special
C = attack Power
D = defender's Defense or Special
X = same-Type attack bonus (1 or 1.5)
Y = Type modifiers (40, 20, 10, 5, 2.5, or 0)
Z = a random number between 217 and 255

I see a bit of chance there 

I'm pretty sure chance has been used in every RPG ever.

Miss a swing? Chance.
Miscast a spell? Chance.
Block an attack? Chance.
Swing you mithril godlaced hammer, dealing 4d17 + 29 bludgeoning and electricity damage? Chance.

And also... I have seen your game mechanic in about 5 other games to date. 

[Sankar]

I believe chance is something that the player will usually try to avoid.

Most pen-and-paper RPGs have chance only in the "critical failure" mechanic, all other things only depend on chance if the player has not invested enough points in it. So it's more like a choice, you can choose to improve a simple stat and be somewhat a "master" of it, or share these points between other more stats and need to depend on luck.

So, I think chance in the combats is OK, but you should allow the player to minimize it.

[Ranordine]

Any introduction of chance into combat will necessarily make tactics less important.  If you want to have combat that is tactically demanding, try to minimize coin-flipping. Giving the player a good idea of the consequences of his actions before he takes one is also important.

A good example of a jRPG combat system where combat involves little chance is Paper Mario.  While the system is simple (i.e. only one character, no action points, no interrupts, etc.), I am unaware of any other game that borrows its use of low numbers and predictability.  I, for one, would be much more interested in many jRPGs if their combat was more like PM's and less like craps.

[JasonPickering]

originally I was envisioning something closer to paper mario, but turn based. where as Paper Mario is turn based its has you always involved, with the attack minigames and dodge button presses.

My goal was to add choice into the game and not make it a strict math problem. I.E. this guy has a 3, I need 3 swords. which is why I added the dice, giving the problem of the player only having partial information. I guess my original idea was you need to hit the enemy or they hit you, but I need to think of a better way to decide if you hit. I wanted something more akin to a puzzle where the person would have to weigh the pros and cons before acting.

[SirNiko]

I think this is something that might work. It's hard to say without knowing what you'd add to it.

It's nice that you make it really clear what the chances are, and the player can see the rolls. In a Tabletop RPG game, this turns it into a puzzle of statistics where the player can know the odds and then use various mechanics to shift odds sufficiently in their favor.

What does the enemy do? Is the player simply completing against himself to not die?

This method, coupled with copious player-controlled mechanics like consumables to alter odds, enemy types that demand different mechanics, and player skills you can equip or use in combat to turn around an unlucky streak, or capitalize on some luck.

Also, if you're going to use luck, it's wise to draw things out so that the player has lots of rolls to make over the course of a session. If the session is short, the rolls will tend to vary more. If the session is long, the player will roll many, many times and luck will even out over time. That makes it a little easier to cope with the possibility of the player having a lucky streak, or an unlucky one.

I'm thinking it could work out. Make sure the player has several encounters in a session so that luck balances out. Make it so the player has skills that let them manipulate their odds a bit, like potions to improve the next roll, or a skill to let them reroll a bad roll. You might want to check out combat in 4th Edition D&D as well, it shows a lot of ways to make RPG attacks that utilize luck in different ways.

[JasonPickering]

well right now the player will go through a series of "combats". they will fight the monsters then be brought to a screen to heal and stuff. it will be like

2 -> 3 -> 4 -> Boss

each arrow is where they would have the chance to heal and possibly buy more items. the enemies do not attack per say. they have a number and you must get that number or better. there will be different kinds of monsters though. like a ghost who flips between a low and high number each turn. also I though about just giving the player 3 moves pokemon style and letting them choose them, maybe +1, +2 and +4. these would have a limited number of uses, although items might let them reset their counter. I guess now I am wondering if I can do away with the dice altogether somehow. its quite a challenge I dont want the player to feel like luck screwed them, but I dont want every battle to have a clear cut answer to it. I think the smart thing I should do now is take the dice mechanic out, and shift it more towards the enemy. That way it doesnt seem like the game screwed you over, but the enemy just outsmarted you.

Edit: so I actually had an idea for this after mentioning the attacks, maybe I could give the player a number, and the monster a number. each would have two attacks to choose from that add to the number. the player attacks and the enemy attacks, after that both numbers are tallied and the highest one does damage. that way players have some information, but are not being screwed over by dice, but actually competing against an AI. 

[dustin]

I like chance but only if it's super clear.  Like a dice is rolled so I know my odds and can figure out what I want to do. In most games I just find it annoying because I have no idea how the chance is worked in.

[antymattar]

I like games that have mechanics that are intertwined and yet verry different. The basic attack is a standard mechanic. Mana is a limited resource mechanic etc. Fun and all that.

[JasonPickering]

so I agreed with what Dustin said about chance being super clear, and I remembered an RPG called Mouse Guard, that used dice as a success/failure determiner. 1-3 fail and 4-6 success. so I started looking into different ways to use this in game and came across some ideas of weighted dice. so perhaps a green dice has a 4 in 6 chance of being a success, a yellow is 3 in 6 and a red is 2 in 6.

so an example

player  - 2 green 1 yellow
monster - 1 green 3 yellow

now we can see what the chances the player has and what the monster has. so I think the real strategy comes in before the roll. the player can look over the situation and then use different items to help this roll, maybe downgrade an enemy die, add a dice to themselves.

[sublinimal]

I have a pretty clear stance on it, believing that the amount of randomness in combat is inversely proportional to fun, as a rule of thumb. Combat feels balanced, fair and meaningful when all variables come together in a way resembling rock-paper-scissors mechanics. Noise subtracts from the possibility of that happening and discredits player interaction - which I believe is the heart of video gaming. There are no technical limitations today that would force one to add a dice element to combat either.

Basically, I can't see any upsides to using chance over choice.

[SirNiko]

In a strategy game, chance introduces an element of uncertainty. Is 90% certainty enough, or do you spend the money to reach 99%? Where is the break-even cost statistically? If this is the last turn of the game, does that break-even value change? Probability based decision making a whole new set of decisions that are neither better nor worse than strict cause / effect decisions, and attract a different group of people.

Probability and chance can easily be used incorrectly and suck, but that shouldn't be confused with probability and chance being naturally bad on their own.

[dustin]

so I agreed with what Dustin said about chance being super clear, and I remembered an RPG called Mouse Guard, that used dice as a success/failure determiner. 1-3 fail and 4-6 success. so I started looking into different ways to use this in game and came across some ideas of weighted dice. so perhaps a green dice has a 4 in 6 chance of being a success, a yellow is 3 in 6 and a red is 2 in 6.

so an example

player  - 2 green 1 yellow
monster - 1 green 3 yellow

now we can see what the chances the player has and what the monster has. so I think the real strategy comes in before the roll. the player can look over the situation and then use different items to help this roll, maybe downgrade an enemy die, add a dice to themselves.

YeahI think this would be great fun if there were items to reroll, improve rolls etc.  It also sounds like it would make for very quick combat which I like with all the strategy coming in equiping charectors correctly etc.


I strongly disagree that randomness decreases fun and don't believe that choice and chance are enemies.  Here if combat had no die rolls your character would just have a number which would be either higher or lower then the enemies.  This makes it obvious what items you need to use removing choice (if I'm a +2 and the enemy is a +3 then of course I should use my potion of +2 there is no choice there). But if die rolls are involved there is a choice (I'm a +2 compared to the enemies +3 but maybe I'll get a better die roll then the enemy so I have to decide if I should risk it in order to save my potion for a harder enemy or use it now).

The use of dice (as opposed to standard computer random numbers) I think helps because everyone intuitively understands the statistics of dice.  If I say you have to roll a 6 to hit that's much more clear to someone then saying (as in pokemon) ((2A/5+2)*B*C)/D)/50)+2)*X)*Y/10)*Z)/255 is the damage you deal.

I do like some rock, paper, scissors combat styles but those seem to have chance in them also in that you won't know what your opponent will pick.  You can have a good idea of what they'll pick but I don't think I ever feel 100% sure therefore the choice feels like it has some element of chance in it.

[JasonPickering]

yeah basically I am using chance to make an incomplete information problem. the player must make a choice without having all the info needed. Even if I switched to a Rock Paper Scissor motif, the chance is still there with the enemy randomly picking one of 3 elements, unless I start adding tells and then this becomes a puzzle game as opposed to a combat game.

now I could remove some chance by giving each monster a target number to reach

example: you have a green dice and a yellow dice you need one success
vs.
example: you have a green dice and a yellow dice the enemy has 1 green dice.

now this seems much more like combat as the enemy is participating now. I want the player to have to make some choices about what they do. ultimately all combat will come down to who has the best chance.

on a side note I always wanted to make an InfoGraphic RPG that was very old school like wizard, but all info was available to the player, so if a player moved over to like heavy attack, it would show a 73% chance of dealing 15 - 18 damage, with a 12% to inflict stun.

[Destral]

As has been said, chance is used to add uncertainty to combat. If attacks always hit, and hits always do the same amount of damage, combat is reduced to a predictable mathematical formula (my attack, defense and HP vs the enemy's attack, defense and HP). Fun/excitement is added when something unexpected happens (you critically hit a stronger monster, increasing your chances of winning, or a weaker monster hits you, or w/e).

There are ways and ways of doing chance: you can have to-hit rolls, but then if chance to miss is too high combat becomes tedious. You can randomise damage, making the race to 0 HP less predictable. You can add critcal hits, but eliminate misses. It depends on how you want the combat to feel.

One possibility you might consider is, at the beginning of the player's turn, rolling two dice, then allowing the player to choose which one they want to use to attack, and which to defend. Then do the same for the monster. If the monster is dumb, make their choice a pure 50% chance of either. If the monster is more defensive oriented, weight the rolls so the chances are 70/30 that it will choose the higher roll for defense.

So, for example:

Player encounters Monster.

Start of combat, roll 2 dice for player, and 2 dice for monster.
Player gets to attack, chooses which of the two dice they are going to use to attack. Monster chooses which die they will use to defend. Add the chose die score to the player's attack and monster's defense, calculate damage.
If monster survives, it uses it's other die to attack, and the player uses their remaining die to defend. Calculate damage.
Reroll die. Player chooses which one to use.

Continue until end of combat.

Not sure if that helps any.

[Ranordine]

As has been said, chance is used to add uncertainty to combat. If attacks always hit, and hits always do the same amount of damage, combat is reduced to a predictable mathematical formula (my attack, defense and HP vs the enemy's attack, defense and HP). Fun/excitement is added when something unexpected happens (you critically hit a stronger monster, increasing your chances of winning, or a weaker monster hits you, or w/e).

If taking chance out of the equation reduces your combat model to a predictable mathematical formula, then your combat model is poorly designed (unless you want it to be formulaic/quick to get through). Adding probabilities only turns it into a (predictably) unpredictable mathematical formula.  Dying just because the computer rolled low numbers for you when you made the best possible combat decisions may be acceptable for some, but it's hard to argue that it's good game design. Satisfying challenge comes from giving your character interesting (which implies a degree of predictability) choices.  Randomization is not the only way to spice up your combat. Consider adding stances, a movement grid, interrupts, ways to force your opponent into situations (or vice versa--think of chess), and widely varying enemies. If this is outside the scope of your game and you treat combat more as a locked door mechanic (with strength/equipment being the key), then randomization just encourages the player to reload until they win.

[JasonPickering]

Ranordine: well some of those ideas dont work in the confines of the game I have built so far. I cant really do movement or interrupts. The stance idea is kind of neat and I could combine that with something like what Destral said. where each stance has an attack and a defense value and then certain stances get bonuses against other ones. Could you elaborate more on the different "Situations" you mentioned. I also plan on having a pretty big variety of enemies, but those will be created once I have a specific combat system working and I see how I can change it.

[gimymblert]

@ranordine
Risk reward (uncertainty assessment) is the foundation of good game design, without uncertainty no risk reward is possible, random != probabilities.

@jason
It's been a while you are trying to get that battle system, and everytimes I read your post it look like random guess, I don't understand where you are getting with those guess

[JasonPickering]

yeah I mostly have the random in there, because I want some chance in combat without just giving a miss or a crit and the player not understanding why they got the hit.

I have been looking around for different ideas now and looking into Rock Paper Scissors. The double blind situation would give the chance I kind of want, and would make the player feel like they have an opponent instead of just numbers. I think that the hardest part of this idea would be enemy AI. I don't want the Enemy to just randomly pick one of 3 choices, so I think that would be my biggest hurdle without falling back on the idea of telegraphing moves, like oh enemy has red eyes now he's going to attack, but I think I have an idea of how I can make enemy AI. I think I might actually use this, but the problem is building it up so I have written down an entire list of ideas and now I am working on building the cycle.

I guess the main problem I have is I want combat that forces the player to make a choice. as a kid I played all the way through pokemon red and the combat just didn't work for me. tackle, tackle, tackle, tackle, tackle, enemy dead. while the computer is using things like leer, tail whip. they are affecting my stats, but not enough for it to make an actual difference in my opinion. damage would look like this. 5 5 5 4 3, while theirs looks like this 5 5 0 0 5. even though they lowered my attack, the pause of lowering it didn't actually help them any. and I feel it usually ends up the same in Rock Paper Scissors. the player will choose rock, but they dont know why they are choosing Rock. its not because they think that the enemy is going to choose scissor, they are just hoping they do. so that's what I am mostly fighting now.

How do you make a player need to think instead of just select attack?

[gimymblert]

You need to pause and think anout the dynamics first, you are stuck with mechanics and they are pointless because they do not support a dynamics, trust me RPS won't save you if you don't know what you want.

You also need to study a bit more feedback loop or else your system will only be an attrition system which you obviously don't want, attrition can be interesting if you design situation for like I had outline in your devlog once. You do so by balancing short term decision with medium term decision about potential and risk.

A battle is really a management system with resource in and out, you may need to look here to find bbasic idea of dynamics: http://www.jorisdormans.nl/machinations/wiki/index.php?title=Pattern_Library those are basic brick of every game dynamics.
look here for a little tut about feedback loop
http://www.jorisdormans.nl/machinations/wiki/index.php?title=Tutorial_1
Try the interactive schema "A Lock and Key Mechanism with Feedback 3" there:
http://www.jorisdormans.nl/machinations/wiki/index.php?title=Mission_Mechanics
See how a simple system have decision without ai with the use of feedback, if you had a pressure mechanics (health) it became a true challenge. You need to balance fighting, harvesting and upgrading (buff in a rpg) while enemy build up his strength, the feedback economy create an interesting situation.


Don't think in gterm of AI, but in term of behavior, set a goal and class enemy by type of challenge you want to provide, instead of looking for ai, you should look for cool moment for the player. Give advance enemy "mode" with different stance they may switch randomly with each stance having distinct strategy (and tougher one will have opposite counter strategy like actions that removed energy now give energy against them, if these action last multiple round, it became a risk reward). Thing like tail whip in pokemon can work great if their impact are significant and can reverse the attrition lead through dynamics only if there is counter move. Look for enemy faqs to seek interesting encounter strategy to replicate into your battle system.

Battle system is a fallacy, you only generalise an interface base on what type of encounter and fun you want the player to have, define this before digging too much into a battle system, under the fancy hood, most battle system are the same and use similar dynamics. Better, totally different game use the same dynamics under the hood. Battle system are only the dress up and most of times most mechanics are there to give superficial variety.

You have an interesting game, I would be sad to see it scrap like the previous one for the exact same reason.