I always think about stats because of multiplayer competitions rather than single player games<since I believe single player games shouldn't be worried about balance, but rather about fun as it would be to the player's benefit>, since almost every game I have ever played allowed for some form of playing against others in order to test and compare your skills against others in the game.

I myself am working on a technical document for a game design I've been refining for the last few years. I believe this iteration of the design is the best so far. I've been waiting for years for this type of game to be made, but even now in 2010, no such game has appeared despite the fact that it doesn't require the best graphics or anything that elaborate. So I think I will take it upon myself to see this game come to fruition as I keep seeing games stick to the same systems of play, even after 10 years. Sure, its tried and true, but doesn't anybody experiment anymore? If technology can evolve, then game design should be able to as well.

Portugaz D Ace: It's a good idea to simplify the number of individual stats while making them do different actions which is what you did. It's easier to understand for most people and makes it more appealing and user friendly. Your stats seem balanced.

However, I personally am trying to do something different, due to this bug I have about how something with low agility and high mobility would be impossible define if say, the speed stat defined both agility and speed at the same time.<Like an old steam train monster, which has incredible speed yet almost no dodge rate.>

I think I will eventually read that story since I heard good things about it. I'm only afraid of spoilers if I go on there, although I managed to read the description of Soul Resonance. It sounds like how a weapon works fundamentally; i.e. how a sword focuses the inertia of the swing unto the blade, or how the sun's light is focused through the lens of a magnifying glass and turned into burning ray.

You probably don't have to worry about trying to balance the stats since it is single player, although after reading the description of soul resonance, it seems to be a technique rather than something like a stat, so the technique/skill itself should be based off stats that have to do with focus and trying to get in synch. Focus might be involved with the ACCURACY stat and maybe the MAGIC stat comes into play. So Soul Resonance would be a technique/skill that uses those stats in its formula.

EDIT:

Vaneger: Doing it that way is easier to understand for most people, but I believe that a intricate system would make it easier to make attacks that would auto-balance themselves if made within a heavily organized and defined system. The simpler paper system is actually much harder to balance since simple actions are too centralized and vague and eventually some effects carry more power than others.

An intricate, defined system is what I am working towards, so that when a skeleton of it is made, any skill made in this system will not be greater or lesser than any other skill. If an attack has low power, it will triumph by having the lowest energy drain or the fastest attack velocity.

This won't be a game where an attack that has high power, the fastest speed, the fastest firing rate, and low energy drain will exist<its okay to have that in a single player game, but put that in a multiplayer game, and everybody and their girlfriend will be using it, and such games lack any diversity.

This system will help promote diversity in not only attacks, but style of play and player characters. Too many games that aren't balanced end up only have a small portion of their characters end up in competitive games.

For example, in the current Pokemon generation, only a combination of the 52 of the 500* pokemon that exist are used the most<due to them having an edge over all the others.> Or take SMASH brothers, the majority of the cast went unused since a handful of characters had the upper hand in terms of attacks and defense.

This happens when somebody tries to manually work at and guess at the balance. If a system is made beforehand, with formulas that will help keep things in check, then things will become easier to balance. For example, a formula that is able to calculate the cost of an attack based on its attack power, speed, size, and other variables. If a person tries to guess the cost of an attack, then some of the attacks will be better than others, and this will stick out in competitive play when players find that using so and so attack gives them 20% more attack power at no disadvantage in any stat than if they used the other attack of the same type.

These kind of formulas can be used to help determine how to balance the game, and prevent broken attacks from appearing or being made.

Also, I've noticed in games that are not played often, the things that are game breaking are usually not found or not obvious, and sometimes somebody might win a tournament because they had a slight statistical advantage that nobody noticed due to the infrequency of multiplayer at the time<only later to find out that such a set was broken.> While in games that have a high frequency of play by thousands of players, the stuff that's imbalanced rears its ugly head and begins to dominate the metagame. For example, in some regulations of the AC4 series, an imbalanced part will be found, and everybody playing to win will begin using it. It then ends in a single designing dominating everygame, and then the other hundreds of parts go unused due to this part having an advantage over all the others.<For example, the dominating part in the early regulations were a type of missile that had tracking great enough to hit fast Mecha while having the firepower to quickly take out somebody of average defense.>

Making a system that determines the balance based on the stats inputted into the formula will go far in a game.

To design a system using simple means is actually the hard way in my opinion, because the more skills you add, the more the inequality arises. If the attacks are input into a uniform system, it will also make it easier to make changes to attacks, and if you change something in the system, it will affect all attacks uniformly since they will all be weighed by the same scale instead of having to fix each attack individually if there is some kind of imbalance that appears in high level play.<High level play being between players who have analyzed the metagame and tested their attacks many times and practiced often as I have observed players doing.>

The other way to balance things is just to give everybody a "broken" attacks and moves, so that everybody has a chance of winning<although even then, those who analyze everything will still manage to find that there are tiers to this "brokeness".

If I finish designing this system, things like tiers and imbalances should be diminished to a low percentage. I think its not impossible, I just think it hasn't happened yet because a lot of the focus in games goes into other things. Perhaps, there hasn't been time to analyze such an approach to a game. In order to maximize the ideal of a "fair fight" in a game that has diversity while keeping it entertaining, a system of this nature is required I believe.

[Edited by - Fulgent on June 20, 2010 6:52:24 PM]

Here is the plan I came up with:

The large weapon fighters rely on Strength and Accuracy to fight, while range-based fighters focus on Agility and Accuracy. The large weapon fighters are weak against range attacks giving them a reason to boost Agility so that they can dodge. Range fighters are weak against close-quarters combat, giving them a reason to boost Strength. In any sense, they are both weak against Magic attacks, so they have to boost Magic to increase Resistance. There will be times where you will be fighting an enemy that you cannot beat with normal attacks no matter how much you buff your character. This is where the Soul Wavelength stat comes in. With Soul Resonance, the character is able to deal a fatal or near-fatal blow. The higher the Soul Wavelength stat, the more attack power available with Soul Resonance. There will also be times that players will have to find certain characters without knowing what the character looks like, but they know what the characters soul wavelength is. They will need Soul Perception to find those characters.

The way I have this game set up is so that players will at some point have to rely on stats other than the required stats for their class. What do you think?

You could do something a bit more adaptive.

For instance if you train your body to do something it will adapt itself automatically and sometimes in surprising ways. so basically to simulate this you have the players start with full control of their stats, then as stats start to reach certain thresholds they can "choose" fewer stats themselves with the rest being decided by the game.

So essentially someone who has reletively balanced stats choose where 90% of their stats went, where as someone who heavily specialized in one stat only got to choose where 30% of their stats went. The missing percents on both characters were chosen by the game.

Another thing you're going to need to be able to do is make it so that a balanced character has access to 90% of the abilities in the game where as a specialized character only has access to 10%.

_

Now in a competitive environment this would mean that when someone sees some tell to specialization they will be able to predict the weakness of someone who is heavily invested in one stat. By being heavily invested they didn't get to choose 70% of their stats, and can use very few abilities. Meaning that they can account for most of the advantage of being heavily invested by hitting your weaknesses(if they're unspecialized), if two specialized people fight it will just be a quick battle of which ever force has the intrinsic advantage.

However as pretty as that sounds on paper it still doesn't work so nicely because most people aren't hardcore enough to plot weakness/resistance graphs and actually use them. Which means that you're probably going to end up very very frustrated as the developer having to choose for every issue if you're going to balance for the hardcore or the core player. Since what is balanced for the hardcore player or the core player tends to be very different.

_

I also think that a game that offers heavy advantages to "not" specializing, players will still specialize because that what they have been trained to do. In nearly all games specialized forces win, and that is an extra hard habit to break because it works in the real world with instinctual reinforcement.

This game is only going to be a single player game for right now. However, over time I may adapt it to be multiplayer. But since I don't have a stable financial foundation, single player only.

Quote:

Original post by Fulgent If you wanted to be able to use, say a MAGIC attack, then it would use the stats of ATT+INT. If you wanted to use a skill that uses projectiles, the formula would use INT+SPD to determine it's attack power

I think the danger here is having a single statistic that affects multiple outcome. While it adds depth and complexity it makes balancing a whole lot harder for the devs. In the above example, if projectiles are overpowered I might try to adjust INT but then that will inadvertently affect magic attacks too.

hi, my opinion is that stats will never manage to be balanced no matter what.

a specialised glass cannon character with 0 dodge, 1 hp (a lv1 rat can 1-shot it), 0 crowd control resistance (everyone can eternally stun-lock it), 0 defense, is gonna always dominate the field, and manage to 1-shot everything.
Why ? because it sacrificed 8000 stat attributes, and now it is running with 8000 strength.

A balanced character is gonna have 2500 defense, 2500 strength, (2500 crowd control or 2500 healing power).

class rogue:
Fight starts : stealth, ambush and 1 shot my enemy.
class paladin :
fight starts : bubble, autoattack and 1 shot enemy.
class warrior : charge = 3 second stun, autoattack = 1 shot enemy.

What you are missing is that the other side has no way to protect itself from getting 1-shotted or stunlocked to death, just because your character is a glass cannon.

If you tell me that a balanced character is gonna survive all this damage, then you are lying throught your teeth. Its impossible.

So what ? he is easy to beat. he has 1 hp, you have the chance to win him.

No you dont, because they can find and hire tank fotresses (all points in defense), and put them in front of them.

So in a 5vs5 fight: your team is balanced, enemy team is 3 glass cannons and 2 tank fotresses :
you search for glasscannons to attack them before they kill you in 1 hit.
you lost 3 party members.
You click on the glass cannon to attack. The tank fotress accidentally got in front of your mouse pointer, mouse - ray tracing missed and instead of attacking the glass cannon you attacked the huge Tauren tank fotress.
You use crowd control dealing 0 dmg but disabling your opponent, but whatever you do its futule its 2vs5, enjoy getting 1 shotted in next second.

Statistics : Battle duration : 3 seconds. Total damage dealt : 0.


Healers are useless in this type of setting, they are glass cannon foods, they never manage to heal anything since everything is dead before fight starts.
its what wow was in burning crusade, and before burning crusade, and still is in wrath in a lesser extend (berserking pve mages 1 shot you when you have 1200 resilience).

This happens because offensive stats are as powerfull as defensive stats.


Now lets say the nerf hammer lowers offensive stats power by reducing total damage dealt by 70%, and buffs defensive stats.

A glass cannon deals 30%-50% dmg of a balanced characters hp.
A balanced character when he attacks will deal 5% of a balanced characters hp, or 0.5% of tank fotress hp.

You now buffed tank fotresses, they now run on field, grab the flag, and go to their base even if a whole army is following them and attacking them.
If they are healers nearby its is even worse, their hp pool never decreases beyond 100%.
Fights against tank fotresses last 5-10 minutes, just to kill 1 player, which destroys gameplay.


Conclusion, no matter what you do 1 stat is gonna be stronger than another and the only way to balance it is to put stat caps, e.g you cant put more than 100% of your level into strength. So the player is forced to put stats into stamina, defense, and other stats.

If anyone beliefs he found a stat system not subject to this, then he is welcome to post his stat formulas, and see how easy it is to get exploited.

Portugaz: I think soul perception should affect some offensive techniques. In this way, it won't be completely specialized for passive roles.

lithos: I think the thresholds are neccessary for games that use less stats. Although in a accurate physics engine, those thresholds seem to be naturally reached from what I observed. <In this paticular mech game, a booster that has double the output of another booster and has about maybe 3 times the drain, so while they may go at 1000kph, they will only do so in bursts, while the mech going at 500kph will be able to do so for several times the duration of the faster mech.

Under my system, in regards to that part about making 90% available to generalist types while specialized only get 10%, I won't have to make such a limitation. Due to the system, a specialized will only be able to use attacks that pertain to their best stats. So if they tried to use any other attacks/techniques/abilities, it would be ineffecient and drain them of valuable energy that would have gone to their specialized techniques. I've experienced this personally. As I have extremely strong legs, but weak arms, and my stamina is so and so. For some reason, my legs can help me travel several miles and not tax my lungs. While if I start punching the air randomly, I'll wear my self out in only a few minutes, so basically my arms low stat value drains my energy pool quickly due to how inefficient it is. While no matter how much I strain my legs, it would take a very long time before I tire myself out. That is why I think form is important and why I was leaning more towards battle simulator than RPG, as it would be easier to regulate glass cannons in a simulator. Anyways, for the same cost of lung power, my legs seem to do exponentially more than my arms did on the same amount.

Another example is that, if my legs are what controls my mobility, and it happens to be stronger than my arms, then I'll want to depend on my legs for both mobility and attack. This is balanced out in that I can't dodge while attacking, but my arms can attack at the same time I use my legs, so I can dodge and attack if use both arms and legs, which is what inspired me to want to use other stats for attack. The speed stat can be used in attack, the cost should be that dodge is unusable during the attack<or your attack has to be interrupted in order to dodge.>

<And if you want to get more technical about it, I have observed that both the arms and legs would have their own independent "limit" or "bar", so there would be two energy bars at work. The bar that limits your arm/leg, and the lung bar. So the arm/leg can independently run out of their own energy<which you will observe when they start to burn> and you may still have lung energy left to use the other. I've managed to make my arms reach their limits while my lung still had some energy left. On the other hand, I have managed to run out my lung power while my legs still had tons of energy left. All three seem to have a 'regain' stat, where the bar is always trying to regenerate, allowing you to use that stat in a few minutes. I like to observe things that way so that I may find a way to apply it in game design.

Loom_weaver: I didn't mean to add INT in both examples, but thanks for pointing that out, as it proves my other point of why I dislike stat systems that only use a handful of stats, even though it is user-friendly, problems like those arise all the time. <I meant ACC+SPD for projectiles, although then I put INT without thinking of the previous since the point of the example was that attacks should use multiple stats while trying to make sense in function of the stats they use. Something with ACC+SPD would then be able to hit anything and dodge anything, which is why there needs to be more stats to allow for more diversity in function and ability. Otherwise, something that moves fast but isn't agile wouldn't be able to define itself in such a system. Anyways, I used those paticular stats because everybody would be familiar with them. I'm aware of the dangers of a stat dominating over others if it is able to affect multiple, important aspects of the game at the same time, although this should be obvious from my previous posts.>

Titan_Fantasy: This could be solved by having a stat cap, although in my system, the animals form acts as the limiter for the highest possible stats. Animals that try to attempt to use an attack that exceeds the limitations of their forms would end up hurting themselves in attempting such a technique.<Their stats would not be able to handle the extreme force of such an attack.>
Tank fortresses would be limited by their weight<if I go with the battle simulator version> and they would either be immobile with slow turning, or when they move<possible with the roll technique and if their round> they won't be able to attack while rolling to position themselves at a better spot. This is just some stuff from my own document, which I have up to this point, been reluctant to reveal in detail just yet.

I like your example, as many present games have this kind of problem. Although, in a battle simulator, it would be difficult to position the glass cannons behind the fortresses as they might hit their own for friendly fire, and without legs for stability or any kind of accuracy, they would have to use splash damage type attacks in order to deal damage to the other team. And that kind of attack would spare even a generalist from the full fury of a specialized stat. And if the other team has anything fast, they would run past the fortresses and to the glass cannons, and if the glass cannons have horrible accuracy, they won't be able to hit at those melee ranges, and if they use their splash damage attack, they'll end up defeating themeselves since they have no defense to resist the power of their attack while the enemy unit might survive unless they were in the middle of the explosion.

Anyways, I think that problem in your example applies more to RPGs than to a simulator. Or perhaps since that problem exists in that game you mentioned, are there any "skills" that basically act like an attack nullifier? What I mean is a "defensive" attack. Maybe when you see the person casting that single shot attack, you use a skill that "nullifies", or "shields" you from the attack, for equal cost and equal power. If no such skills exist, I think that might be the problem rather than the stats being too equal. Then if such skills exists, they should be as widely usable as attacks are, and not be specialized to a single type. Or they should at least allow people to use their own attacks to defend them from such a strong skill.

I actually am reluctant to post my stat formulas, I apologize. It's just that I have this strange feeling like a lurker might go "whoa, those are good, lets apply it to our own game" and before I am able to make my own implement it in my game, somebody with the man-power and resources will have already have used the system themeselves. I truly believe when I finish this system, it will be extremely balanced, although not perfectly, but there won't be things like 'single hit kill' teams and other imbalances that stem from the problems of using limited stats and limited types of skills/techniques.

I sort of want to pitch the game first to a few companies or at least start it up myself before revealing the system. If there is a working demo of it made, then I would gladly share it for analysis and testing. The purpose of this engine is to maximize enjoyment while lessening any factors that would diminish personal entertainment. I think it will be a new gaming experience for those that play. I always get excited just looking at the document and just thinking about how it would be like if it were already made.

[Edited by - Fulgent on June 25, 2010 2:01:45 PM]