Gameplay design is HARD!

Its good when something completely new comes up, but in my opinion for most people its better to take something that has been down before and try and add something new too it or change it in some way. As oppossed to trying to radically change something, change things radually so you can see what works and what doesn't.

In the end of course it comes done to the individiual person and what they want to do.

As for the help wanted post, you'll probably get some response but most of them probably will be people wasting your time, intentially or not. But if your careful you'l probably end up with someone that can help you.

It's actually a lot easier than you think to come up with a compliant game. TO give you an example, take 2-player Doom. Actions are move (change location statistics), and shoot (change your ammo statistic, and in certain circumstances the enemy's health statistic) etc.

Now you could play Doom turn-based, and just using numbers, but I can pretty much guarantee you that it wouldn't be fun. What made it fun was the stuff you claim to have covered already - visual feedback, story and setting.

So you need to be a bit more detailed in your requirements. I'm thinking you want to restrict the relationships between those statistics. But statistics are rarely fun in themselves.

Quote:

Original post by Argus2 It's actually a lot easier than you think to come up with a compliant game. TO give you an example, take 2-player Doom. Actions are move (change location statistics), and shoot (change your ammo statistic, and in certain circumstances the enemy's health statistic) etc.

I appreciate your detailed response, but your design above is NOT compliant with the specifications.

My original post said: "There are two entities, each with a slew of statistics. Every turn, all the player can do (other than quit the game :) is choose an action (from an available menu of actions) for each entity."

In other words, It's a one-player game, and the one player controls BOTH entities' statistics. The bulk of the gameplay mechanic design work is to arrange the statistic balancing mechanics such that the game IS fun!

It's a lot harder than it looks! :)

Quote:

Original post by Argus2 Now you could play Doom turn-based, and just using numbers, but I can pretty much guarantee you that it wouldn't be fun. What made it fun was the stuff you claim to have covered already - visual feedback, story and setting.

Turn-based Doom isn't compliant with the design specifications. A dual-city sim-city (or two-avatar SIMSs) might be ... the trick is to balance the statistic relationship to make the gameplay fun. This is the hard part of the gameplay mechanic design! And this is what I'm asking for.

Quote:

Original post by Argus2 So you need to be a bit more detailed in your requirements. I'm thinking you want to restrict the relationships between those statistics. But statistics are rarely fun in themselves.

Once again, I appreciate the response, but I believe my requirements are quite sufficient. The restrictions between the relationships are part of the design I want someone else to do. I can handle the rest of the program's presentation myself.

Feel free to brainstorm additional design ideas!

[Edited by - GemuhDesayinah on November 15, 2005 5:59:50 AM]

Quote:

Original post by GemuhDesayinah In other words, It's a one-player game, and the one player controls BOTH entities' statistics. The goal of the game is to arrange the statistic balancing mechanics such that the game IS fun!

You seem to be neglecting the fact that you've already designed the gameplay mechanics yourself. What the statistics and actions are is highly irrelevant; the most important aspects of the gameplay have already been covered (one player, two entities, statistic and turn-based).

Your mechanics are boring and not fun, any way you spin it. Imagine playing a Final Fantasy game where you are given control of the enemies too; it lacks conflict, it lacks challenge, it lacks strategy. All you do is click and watch numbers go up and down until they end up the way you want.

Click button statistic goes up click button statistic goes down.

I'm quite confident you could come up with a very complex puzzle-game by say, limiting the amount of turns the player has, where the gameplay consists of the player figuring out the correct pattern of actions to end up with the desired statistics.

But that doesn't make it fun, and the reason is because all your game can possibly amount to is figuring out the pattern of buttons to press and watching numbers. Why even have two entities? It all becomes one, continuous pattern of actions anyways. There's no dynamic to it.

Click button statistic goes up click button statistic goes down.

Fun gameplay mechanics require a fun concept all the way down to very basics of the game design. The foundation of your game is not a fun design, and that is why you are having trouble making a fun game off of it. A single change, such as a)adding an AI entity, b)making it real-time, or c)not basing it on statistics, would allow for a wealth of options.

As you toss more and more limits onto your gameplay design, at some point it becomes your fault for not giving anyone anything to work with.

Click button statistic goes up click button statistic goes down.

By now you're quite tired of seeing that line. You're probably already skipping over it each time you see it. Perhaps if you look at your own game design, you will see how quickly it will become just as boring as me saying:

Click button statistic goes up click button statistic goes down.

I think the problem is that the input options are too limited. Most games have something more than just a choice of menu options. Usually, there's some geometric constraints (Tetris, Bejeweled) or some evolution of the world (Lemmings, Bust-a-move) that provide more of a "hook" for the player's imagination to make the game fun.

Sorry, I didn't read your specifications correctly, as you pointed out. However, I think my version of Doom does become compliant if one person controls both players. Or at least with minimal adjustment.

Note that any statistic that isn't related to either entity can still be modelled as a statistic in both entities that stays constant relative to those entities, so the designer does have a bit of leeway here.

In any case, I think you're missing a big part of game design. There are countless mathematical models you can use to make a game's solution non-trivial, but the fun part is often elsewhere - eg. sensory feedback (eg. FPS death animations). Imagine a game like Scrabble except that the accepted "word" list was just a bunch of random strings of letters. The mathematical model is equivalent to Scrabble's, but the game itself would not be fun at all.

And as with LockePick, I think I'd want to know why you feel the need to restrict the game to 2 entities. ie. what do you gain by restricting the game to 2 entities, rather than 3 for instance, or n entities? The only sensible interpretation I can come up with is:

The game consists of a set of statistics such that every statistic s1, has a paired statistic s2. An operation o1 can affect any statistics, but there must exist a paired operation o2 that has an equivalent effect on the pairs of those statistics.

If this is the design, then the designer needs to know whether the initial value for any statistic required to be the same as that of its pair. And what makes you think this class of "symmetrical" puzzle will be any more fun than other classes of puzzle?

Oooh I made the game you're talking about... no wait that was my expenses spreadsheet at work.

How about a hotseat game then where the players take it in turns to control the entities. Each entity has 16 statistics representing positions for various pieces. There is a virtual 8x8 board which the pieces sit on and the actions move the pieces around. If you land on another piece then you take it off the board (position -1 and can't move it anymore). Each player has a piece that must always be in a place where it can't be took. Damn I'm describing Chess :P

I see your point though but I think those constraints are far too restricting. I wouldn't give a world famour painter one colour and then tie his hands behind his back and expect a masterpiece.

First of all, let me start off by thanking the responders for their well-written feedback. However ...

Quote:

Original post by LockePick You seem to be neglecting the fact that you've already designed the gameplay mechanics yourself. What the statistics and actions are is highly irrelevant; the most important aspects of the gameplay have already been covered (one player, two entities, statistic and turn-based).

I haven't really designed much: all I listed was a vague, loose overview for a simulation strategy game. The devil is in the details, so to speak. "What actions affect which statistics, and when" is the crux of what will make the game fun - or not.

Quote:

Original post by LockePick Your mechanics are boring and not fun, any way you spin it. Imagine playing a Final Fantasy game where you are given control of the enemies too; it lacks conflict, it lacks challenge, it lacks strategy. All you do is click and watch numbers go up and down until they end up the way you want.

I respectfully disagree. The SIMS, Harvest Moon and numerous commercially successful simulation games are very fun games that use approximately the same gameplay architecture (the number of entities differ). A panoply of emergent strategies exist depending on the game's goals and the detailed gameplay mechanics. Time/Number of turn limitations, random events and the addition/alteration of existent rules based on certain statistic values can also add additional strategic elements.

Quote:

Original post by LockePick Click button statistic goes up click button statistic goes down.

Commercially successful simulation strategy games are usually a LOT more complicated than that. Raising one stat above a certain level may start to automatically decrease another stat which you want to keep high. Or it may trigger a rule change that affects your entire strategy. Or it may require resources (resources are a stat!) that one doesn't have at the time, thus necessitating either a delay in one's plans, or the alteration of a different stat. Or it may spawn an random event that requires a different strategy to 'solve' in order to raise one's statistics to the level one wants. Or it can 'spawn' (or 'kill') an event that affects certain stats at regular time intervals, etc. Some statistics may have a rate of change dependent upon the value of other statistics. Some statistics may be easy to raise, but may require constant management to keep high in order to prevent attenuation. Some stats require the user to use actions in order to increase (hereby consuming turns), whereas they may increase automatically once a certain stat rises above a certain level. Increases in some stats force decreases in other stats. Some stats may be affected by the sum of other stats, or the difference BETWEEN certain stats, etc.

Quote:

Original post by LockePick ... all your game can possibly amount to is figuring out the pattern of buttons to press and watching numbers.

I disagree. That situation can be ameliorated (or in some cases, even eliminated) by either (1)adding some amount of randomization, (2) creating such a large number of variables that finding such a pattern would be INCREDIBLY difficult, or (3) Creating a whole slew of different gameplay settings (of the statistics, actions and goals, each requiring a different strategy to win with (e.g. you beat that game as a Knight? Good - now try it as an archer!))

Quote:

Original post by LockePick Why even have two entities? It all becomes one, continuous pattern of actions anyways.

Good point! You can change the number of entities if you please. My mind was thinking of gameplay options involving symmetry, but even those can be modelled with only one entity.

Quote:

Original post by LockePick Fun gameplay mechanics require a fun concept all the way down to very basics of the game design ... The foundation of your game is not a fun design, and that is why you are having trouble making a fun game off of it. A single change, such as a)adding an AI entity, b)making it real-time, or c)not basing it on statistics, would allow for a wealth of options.

The foundation I described is the basis on which all successful turn-based strategic simulations are based. No AI or real time options are necessary.

Quote:

Original post by hplus0603 Usually, there's ... some evolution of the world (Lemmings, Bust-a-move) that provide more of a "hook" for the player's imagination to make the game fun.

The effect upon gameplay by evolution of the world can be modelled as a change in the statistic alteration rules. I agree that geometric (or linear i.e. location) constraints can also add a new layer of gameplay, but I'd like to create a fun game without adding that additional level of complexity.

Quote:

Original post by argus2 There are countless mathematical models you can use to make the game's solution non-trivial,

Could you please name some? I'm very interested in this. Seriously!!!

Quote:

Original post by argus2 ... but the fun part is often elsewhere - eg. sensory feedback (eg. FPS death animations).

Not exactly. I acknowledge the need for it, but no matter how viscerally pleasing sensory feedback is, the game isn't fun unless the underlying gameplay mechanic is.

I already have the sensory feedback potion of the game design covered: I don't need any help with that potion: what I'm looking for now is someone to put down the detailed core gameplay mechanic.

Quote:

Original post by argus2 Note that any statistic that isn't related to either entity can still be modelled as a statistic in both entities that stays constant relative to those entities, so the designer does have a bit of leeway here.

That is correct. Thanks for pointing that out.

Quote:

Original post by Metorical ... Damn I'm describing Chess :P

Heh :) The gameplay you described involves geometric/location constraints which (although they can be modelled with statistics) I'd prefer not to complicate the game design with. And Chess is a two player game that doesn't make sense to be played by only one player.

Some more gameplay statistic ideas:
For example, let's say the goal of the game is to MATCH certain disparate statistics: does one raise stat A, lower stat B, or alter both at the same time? And if the latter, how much time spent on each? 75% time raising A while 25% of the time lowering B?

Two oozes, say red and blue.

Apply heat:
red ooze will solidify and slick.
blue ooze will solidify and sticky.

Apply electrical current:
red ooze will melt and defy gravity.
blue ooze will become elastic.

Add other reactions here!

Now you put them in a small room with some obstacles and you control the ooze probe which applies the head, electrical current, etc to specific areas of the ooze to get it to react and move. Apply electrical current to the blue ooze to create a rubber band, then apply heat to the red ooze to solidify so it can be shoot across the room.

Quote:

Oooh I made the game you're talking about... no wait that was my expenses spreadsheet at work.

Hay I played that game the other day.... oh wait, it was an CRPG.

Quote:

I see your point though but I think those constraints are far too restricting. I wouldn't give a world famour painter one colour and then tie his hands behind his back and expect a masterpiece.

There are quite a few competitions where they do just this, like Iron Chef. A bit off topic but the scene I liked in the movie Apollo 13 is when a scientist walked into a room of other scientists, dumps a bunch of stuff on the table and said "we have to use this to fit this square filter into this round hole" (paraphrased of course).

Quote:

I think the problem is that the input options are too limited. Most games have something more than just a choice of menu options.

I agree, you have to have more than just the menu options, even if it is only to drag the items around the screen, or tell the unit where to go or what to attack, or, wait, isn't that a RTS?

Here's one. I generally find this kind of game fun until my brain melts which is quite quickly:

You've got a pane of glass in 3D space, where each side of the pane is divided up into a grid. Every square on one side of the glass corresponds directly to a square on the other side (i.e. they are in the same place). Within each square you either have empty space, a wall block, a red door, or a blue door. Corresponding grid squares on each side of the glass can have different pieces in. The net effect is of two different 'mazes' on either side of a pane of glass with red doors and blue doors in.

Designate red and blue start squares and red and blue finish squares such that there is exactly one start and one finish on each side of the board. Place playing pieces in the start squares, with colours corresponding to the colour of the start square (i.e. red piece starts in red start, blue piece starts in blue start).

On each turn, the player can move one of the playing pieces into an adjacent square on the same side of the glass that is empty, or into a 'door' square of the same colour as the playing piece. Alternatively, if both playing pieces are in corresponding empty squares, the player can choose to 'switch' them, turning the red piece blue and the blue piece red.

The game is won when both playing pieces are in their correctly-coloured finish squares.

This design doesn't meet your second requirement, but it would be relatively simple to decrease a statistic on each piece each turn automatically, and place powerups around the maze that restore this statistic to maximum. If this statistic reaches zero the game is lost.

I just looked at it, so my idea is pretty raw. Develop it if you want, ingore it if you like.

Those two entities will fight each other. The players goal is to not to keep them balanced, but to keep them unbalanced for as long as possible. That means keeping the advantage of one entity as high as possible, but not letting it win. The advantage can also be transfered to the other entity. Basicly - keeping the battle exciting without a winner for as long as possible.

Since your game is turn based, the main enemy will be the random events. You can't transfer the advantage all at once, so you have to plan few turns ahead when so you can help the losing side before he loses and the battle is over. The random events can interupt your planning, and if you won't be carefull, they will make you lose.

Since you want an abstract idea, and you won't even tell use the settings of the game, that is all I can give you right now.

I might not be reading this right, but essentially if you add a map, you've got every map-based strategy game in existence covered by your description.

So you don't have a map, and the game is simply exploring combinations of variables until you reach a desired goal by manipulating menu-based choices? (This thread is all so vague it's mind-boggling)

THe first thing that comes to mind is the "lemonade stand" game. You know, you determine how many cups of lemonade to make, how to mix the lemons and sugar, what price to charge, etc. You saisfy consumer expectation to gain repeat business to make more money, etc., etc. Maybe you're competing with other stands, and by lowering your price you take market share from them. The goal could be scoring as high as possible (ie, making money) or maybe accomplishing certain objectives, like grabbing 50% market share. Maybe it rains and decreases the sales of everyone; maybe everyone decides they want sno-cones instead.

Anyway, a business simulation is the first thing that comes to my mind, if I'm reading your description correctly. But you can do anything with this, political sims, high school popularity sims, etc., etc....

Quote:

Original post by abstractimmersion So you don't have a map, and the game is simply exploring combinations of variables until you reach a desired goal by manipulating menu-based choices? (This thread is all so vague it's mind-boggling)

Almost, but not quite. See - a well designed game would provide some logic behind the choices rather than a random exploration of variable combinations. Players should see the correlation between their choices and the effect, and should use logic to figure out what combinations would best lead them to their goals (whatever their goals are).

Quote:

Original post by abstractimmersion THe first thing that comes to mind is the "lemonade stand" game ...

Quote:

Original post by abstractimmersion Anyway, a business simulation is the first thing that comes to my mind, if I'm reading your description correctly. But you can do anything with this, political sims, high school popularity sims, etc., etc....

You are correct: those games satisfy the criteria! Now can you come up with a FUN detailed gameplay mechanic for one (without ripping another game off)? What statistics are changed with what actions, and by what amounts? What is the goal of the game? etc.

but... but... didn't you specify, from the beginning that it was up to the player to define his own goals in teh game? So basically, what you are asking, is a sandbox design that will allow to do anything, with no graphics so that you don't have to bother creating them, as little text as one dares, so that you don't have to write them, and to make it FUN? Is this a trick bargain with the devil, or are you trying to make fun of us? No story, no point, no graphics, no mechanics. So can you, at least, point us in the direction of what you would like?

I had imagined something, but it requires graphics. And it has predefined goals.

Let's say you have a two-sided panel, and on both visible sides, you find a maze made of sliders affecting both sides. The goal of the game is to move the sliders in the right sequence so that both beads reach the bottom at the same time. It requires a little story-making, but basically, this is a puzzle game. Maybe you can add some more twists, like holes in the panel, so that the beads can change sides, and have to land on the proper color? Or sometimes make the beads of different sizes, so that one bead can go through both-sizes holes, but the other one cannot go through the smaller ones? maybe have the sliders form a grid, so that you can move the beads sometimes up, by using the right slider at the right place? I dunno. This kind of "affecting some statistics and playing both sides" just tells me that.

Maybe I am crazy and shouldn't be allowed near board games anymore...

Original post by Dragoncar
Its good when something completely new comes up, but in my opinion for most people its better to take something that has been down before and try and add something new too it or change it in some way. As oppossed to trying to radically change something, change things radually so you can see what works and what doesn't.quote]

I like both just as much, we always have our MMO's, FPSs, Sports, any game improved and it normally works, when one that makes the genre better it sells good.

However new gameplay experiences are needed to keep people ontersted, the RTSgenre has really slowed down, at the moment I'd say it would be pretty hard to improve without making it hybrid. Games like Katamari Damacy show you that you can create something new, and people will want to play it. Wario ware aswell, a modern game that uses two buttons and a D-pad with over 200 minigames is absurd yet works.

It's good that we have both improvements and new things, which iswhat we can expect from the next generation. Nintendo if it pulls it off, will bring many new things, while PS3 and xbox will improve (I hope) many things.

No, fournicolas, he didn't say he will have no graphics. What he said was that there will be no part of the game that depends on location, ie, on a grid or map.

to the thread starter -- play some lemonade stand games and some sims. there are plenty of games out there like this, you just need to find what you're interested in making and tweak the variables.

Quote:

Original post by Fournicolas but... but... didn't you specify, from the beginning that it was up to the player to define his own goals in teh game?

No I did not.

It is up to the game designer to set the goals. The game designer has the OPTION of making it a sandbox game design, but it isn't necessary.

My original post stated:
"Declare the goal of the game, define what actions are available and which statistics they affect and by how much (and - if applicable - what statistics determine the magnitude of the change if any at all), and declare the total number of statistics available (the fewer the better). And make the game fun - it should be very challenging (but not frustrating) for a player to achieve the goal!"

These are all the things the game designer does to make the game fun. All the player does is select an option each turn.

Quote:

Original post by Fournicolas Maybe I am crazy and shouldn't be allowed near board games anymore...

You aren't! Yours (as well as some of the other ideas on this thread) are pretty sharp and original! Neat work!