Im working on a multiplayer battle game, where players have server-saved characters that they can battle out with other players via turn-based battle system. The problem I am foreseeing as I lay out the specifics to the battle system are some of the same issues that Im sure many designers face/have faced in the past. With an MMORPG, battles are usually not "Turn Based", so much as a "Real Time" battle engine with time delays based on different moves. With my design, its a legitimate turn based, ie. I attack, you attack, I attack, you attack, etc. Although I have not finalized the algorithims, or really even started on them, the basic will be something along the lines of, I have weapon "X", does 10-20 dmg. So attacking will generate an attack between 10-20. This is going to introduce a certain amount of luck determining the overall outcome of the battle, and not so much strategy. To counter this effect, Im goin to adopt the Xenogears/Xenosaga point system, where the command DEFEND may spend 1 point a turn, and ATTACK spending 2 points. At the beggining of every turn, the player will recieve 2 points to spend. This way, the decision to DEFEND will save the player an extra point, and at the beggining of the next turn, there will be 3 points available to him, leaving the option of doing a "3-point" action. I believe this adds *some* strategy to the battle. In order to add more strategy, there will be a large assortment of items/armor/weapons that change the status of your character, ie. immune to specific status ailments, absorb different elements, etc. In order to incorporate these feature into each players strategy, your opponents most recent battle history is viewable, so that you can cater the items/equip you bring to battle according to what your opponent has done in the past. However, I just dont see this being enough strategy. Do any of you have some input on what other features should be available in a game of this type? There is no adventuring, battleing NPCs, etc. Its purely a multiplayer battle engine. We have plans for 3+ player battles, in which case, strategy would be dicteted by the teams of players, however, ina one-on-one battle of this type, do we as game designers dictate strategy? Do we know previous to releasing the best strategies to use and why? We obviously have insight, but do we dictate it based on our balancing tests results and statistics? I just dont see this working as well as I hope, if after balancing the game and looking past the small strategy factors I mentioned above, my game has become just a random roll of the die.. Any suggestions?

I recommend using the arena in the fight. Instead of just standing toe-to-toe and exchanging blows, let them move around, get behind or on top of things, and otherwise interact with their surroundings. Take a look at something like Tactics Ogre to get an idea for how high ground or distance can affect combat.

A tile-based system would be sufficient, but don't forget things like using cover effectively, range of sight, zones of cover or concealment, and various types of attack/defense. I'm thinking of gunfights, here, but if you're thinking swords and sorcery, you might do it differently.

EDIT: Think about excluding levelling systems. If it's to be a duel system, then levelling will just lead to invincible characters. Nobody could touch the top guy, and he'd have no challenges. Instead, maybe a customization method, based on the assignation of a finite number of points, would be better. It would also be conducive to teamwork and specialization.

Or, if you do include levelling, make sure that XP is gained based on the difficulty of the fight. If a level 60 player wants to get any decent XP, he'll have to fight two or three lvl 40 opponents. If he wins, he gets some decent XP, and if he loses all of his enemies get a ton of XP.

Better still, dish out XP in real-time, like Tactics Ogre. A maneuver would award XP based on difficulty, success, and effect. That way, you can improve even if you don't win the fight, and the newb will learn a lot from fighting his betters. Of course, if you get steamrolled before you get a chance to attack, you won't gain anything, which will compel players to be reasonable about their fights. No getting maxxed out by having six archdemons crush your skull three times a day.

In a nutshell, any turn-based strategy games that do not involve luck relies on the a system based on prediction and deception. In other words it is about guessing what your opponent is trying to do, stopping your opponent, carrying out your moves while hidden your objectives.

The primitive form of turn-based 'battle' is chess. The rules are simple and deterministic. In addition to that the designer of chess did not need to know all the strategies involve. However, chess is able to provide a platform for strategies. Why?

1) Variation of intention
In a chess game, when I move my knight, you do not necessary know what I will do afterwards. While you don't know exactly what I am trying to do, you know that that move is part of a strategy. Is there variation in the type of targets that I can attack? How many ways are there to win (how many ways are there to defeat the opponent)? In what ways is each move a meaningful clue for the opponent? Referring to you point system, how many things can a player do by saving points? Does the actions of saving points provide meaning clues for the opponent, or does it regress into a game of guessing?

2) Delay between cause and effect
In chess, this refers to the setup required for a checkmate. As long as the opponent can predict the effect earlier enough, there are ways to counter the strategy. In your system, this refers to your point system, that there is a delay before you can perform the multi-point actions. Is there enough delay between the causes and effect for the opponent to react? Are there ways to counter the strategies in your system?

3) Incorporation of enemy moves in strategies
In chess, enemy moves are considered and incorporated in strategies (i.e. I knew you would do that therefore I did this beforehand, so that when you got there I could ...) To what degree do you provide strategies in these forms? How do enemy moves make certain strategies preferable and applicable? How do enemy moves make certain strategies inapplicable?

4) Deterministic outcome
Due to the risk already involved in planning and execution of strategies, the harder the strategy is to setup, the more the outcome should be deterministic. In other words, if I take several turns to setup a super attack under your scrutiny, that attack shouldn't miss. And when that attack is execute, I should know very well what state you will be in. You can deduce that every element involved in setting up should be more or less deterministic. In general, these opportunities are one-time only. These are not games in which a player will be satisfied with "well, out of 100 times that I use this strategy, 80 of the final moves deliver the desired effect" The more indeterministic the outcome, the more you will drive the player away from risk taking and strategy, and relies on 'fail-safe' modes such as relying on damage-over-time, and playing too defensively.


Putting them together:

Elemental seals combat system (Turn-based):

In this combat system, you can move around, attack directly, cast spells, and all the usual things. However, whenever you cast a spell (spells are all elemental in this example), a symbol (or a colored dot) is created on the shared grid representing the environmental affinity to the elements. As the battle goes on, certain patterns on the grid can be harnessed to allow certain higher power spells, summons, and whatnots. After the pattern is used it is discharged and disappears from the gird, to allow patterns being discharged in sequence.

Thanks for your feedback folks. Estok, those are great concepts, and certainly applicable to my design. I appreciate you laying it out in such a format, I have began thinking about my design in a different way, and its has helped me very much. Thanks!