I was considering last night games such as diablo, maplestory, etc, in which you level up your champion over time, and assign skill points each time you level. My dilemma is this:

Do I, as a designer, want to make enemies such that there is an optimal formula for killing them? Consider this simple example. I have 2 attributes, A and B. Both attributes have 1 point in them. Monster1 requires 1 of A and 2 of B to kill. Monster2 requires 2 of A and 3 of B to kill. Monster 3 requires 2 of B and 4 of A to kill. Assuming you can level 1 skill point per level-up, the optimal fighting requirement is Kill Monster1 (I have 1 point in A, level-up attribute A). Kill Monster2 (I'd have 2 points in A, level-up attribute B), Kill Monster3 (I'd have 2 points in B, so I could kill him). Any other path would require you to grind out an extra level before you could kill the next monster.

So, how is the above any different from some complex system of HP, Mana, Dexterity, Agility, Wisdom, Endurance, etc that allows you to kill a monster if you have the right combination of attributes, but requires extra grinding if you couldn't find the magical formula?

And even more importantly, how obvious do you make clues as to what attributes a player needs to level? Could you come right out and state their weakness? Do you reveal it after xxx encounters? Do you let users 'see' the enemies armor, and deduce the weakness after doing some basic math? Or, do you not reveal it at all, and let the user figure it out through trial and error?

If there is only one right choice, why allow choice? Just give the players the skills he needs. This way you don't frustrate the player. The goal with having different skills and allowing the player to choose point allocation is to allow for different styles of play. High def for tanks lots of agility for ranged attacks etc. Therefore you should be able to be effective with a wide verity of point allocations maybe not all possible combinations but a wide selection.

I guess I think of point-allocation like a maze. There is the 'fastest' path, but there are a lot of other options that could get you there, although it would take much longer. And the 'best' players are the ones that can choose the fastest paths. But, perhaps there is a compromise. I could perhaps allow for 3 paths of equal play. Anything on these 3 paths will be nominal. Anything outside of these 3 would result in slower progress (allowing others to overtake you).

That would only create cookie-cutter builds and stigma against those who do not follow the builds. Players will soon find optimal paths and communicate them to other players. If your business is selling respec items, then this would be good(for your pockets).

If I allow 2 optimal paths, but a multitude of choices, I don't think that will create a cookie cutter build. For example:

You have 3 attributes. A, B, C. You have 1 skill point. You can place this into any attribute. Place it into a skill, and I will tell you if you can defeat the LvL 1 monster.

If you succeed, you will level up, and you can add another skill point. Do this, and I'll let you fight the LvL 2 monster.

If you succeed vs. the lvl 2 monster, you will level up, and you can add another skill point. Do this, and I'll let you fight the lvl 3 monster.

This will go on until lvl 10. The person that can reach lvl 10 in the least number of turns is the winner. Each attack cost a turn. Fail to defeat the monster, and you lose 3 turns (1 for attacking, 2 to revive).

You may spend 2 turns attacking the previous level monster for a level up.

I welcome all challengers... I will update the original thread with the current progress.

[Edited by - Cygnus_X on September 16, 2010 11:06:14 AM]

First of all.. it's usually called a character. Not a champion.

Secondly, attributes don't necessarily determine whether you're CAPABLE of killing a monster. They just determine how fast/efficiently you kill it.

And to make everything said so far moot, a simple solution is that all stats have gradually diminishing returns (at least geometrically).

It's fairly easy to balance things that way so that after putting points into one attribute for awhile, you'll want to put them into another to gain more ability.

Or give each attribute increasing returns so they will want to specialize in one.

Attributes in many games determine if you can defeat a monster. Example:

Maple Story. You literally cannot hit the the enemy if he's too many levels ahead of you.

Final Fantasy. I dare you to try to defeat one of the ultimate weapons at level 1.

League of legends. Take your level 1 champion (yes, champion) up against the barron. You will lose.

Secret of Mana. Again, level or die. There are some bosses you simply cannot defeat if your level is too low.

Zelda. Ever try to take on gannon with 3 heart containers and no shield? I'd like to see someone attempt this.

Diablo. You can't defeat Diablo at level 1 without a hack. There is just no way.

WoW - Again, level = power = ability to move forward.

WC3 - Same theme.


Any game that requires you to level to become more powerful (or obtain items that provide more powerful attributes), has a point at which it is not possible to defeat the enemy without gaining at least one more level or one more item.

So why drag it out? The game mechanic I have listed above simplifies the problem. Either you level and get that 1 more stat thats needed to move on, or you die, and waste precious time. There is no geometric math. No calculus. This is the most basic combat system possible. This can be beaten in 10 moves.

It sounds like your system is less about building the players desired character build and more about guessing which stat combination is need to win each encounter.

In which case you need to provide the player enough information to figure out what the correct combination is. After all if I load my points into stat A and easily win the first three battles then I’d assume that build is an effective one. But then battle 4 might require 6 points in stat B and I loose every time until I can get that amount of points, after first figuring out that I need to up my B stat to 6.

It sounds like that would get old really fast and most people will probably just end dividing points evenly across all stats as that it least likely build to fail.

TechnoGoth - good post. Thanks.

First, I have designed this so you can brute force through the game with all A, B or C attributes.

But, your commment is kind of the whole point to this thread. And this is a bit of theorycraft, so I'm still working on it. But my thought here is, do you 'really' know how much improving a skill by 1 point will help? You may see a difference. Damage received reduced by 1. Damage inflicted increased by 1. But if there were an 'optimal' path, would you ever know it? And isn't that kind of what a player wants? To think or to know that they can out-smart another player by picking a combination that works best? And is that what we, as designers, want to do? Certainly a game in which all paths are equal isn't much fun. Or is it? Perhaps all paths in my scenario above lead to victory?