Currently working on a project, and i'm really enthusiastic about it, and it seems like really neat stuff to me, but i was talking to several others about it recently and got the feeling that they were completely loosing what i was trying to illustrate to them, not because they are stupid, but because looking back on the discussion, there is just too much stuff being said. The 'complete status buildup' of a typical player controlled piece in my new project consists of up to 20 pages of numbers and various listings, everything from armor integrity to energy draw to weapon condition to a breakdown of dozens of components [you play as an android, or rather each of your characters is potentially an android, and the androids are the most statistically complex, so they are being used for this discussion(future rpg)]. While i really like the idea i have, and like the way its headed, i fear that the effect of changes to a character are being lost in the sea of numbers. Its time for a major decision, and it needs to be made before i proceed. Do I... A: Stick with the complex system, letting it be a project that'll only be enjoyed by those who are willing to wade through pages of text, and those who actually want to learn how to use the characters. B: Merge inter-related statistics together [instead of weapon heat, weapon damage, likelyhood of weapon jam, and a bunch of other things, narrow it all down to 'weapon condition', 'armor condition', 'sensor condition' ect], and keep a lot of the complicated game mechanics behind the scenes, exposing only the more generalized forms to the player. C: Gut the current mass of status screens, and attempt to actually simplify the equations, and reduce the volume of numbers that go into each little thing. something else?.... feel like i'm presenting the player with just too much stuff All these numbers also make it so that a player would absolutely be required to sit down with a manual before even playing, and would make those who just like to 'jump right in', get creamed by an AI [that i can assure you, will take full advantage of the complexity]

You could go with a combination of A and B. Have a menu with 'weapon condition' and when the player what they can click on it and expand the menu and get more detailed information. In that way you can make it pretty easy to start with the game and still be able to tweak for the players who played it a little longer.

Or, prehaps, you could keep all the information, but not show it to the player at all.

I'd go with A or C. If there are stats that won't really have an impact on the players' choices, I'd get rid of them (eg. if a weapon overheating and a weapon jamming causes pretty much the same thing you don't need a stat for each, but if overheating causes the weapon to blow up while jamming just causes it to stop working, the player would probably like to know which is likely to happen with that weapon he's considering).

Dunno enough about the game to know if this'd work, but you could make sure, and make it obvious, that lots of the stats won't really need to be considered early in the game, and then "introduce" more stats as the game progresses. Weapons available to low-level characters could have values just about equal to each other for most of the stats, so that the player needs to compare just, say, their speed and damage, while later in the game he needs to consider stuff like overheating/jamming/whatnot. Same thing can be applied to other kinds of equipment and abilities/skills/whatnot etc.

@Ezbez: Not sure i follow, not show any statistics at all? or only certain statistics?

@Gnarf: Your describing the difference between over-heating and jamming is exactly along the lines of what i'm doing. Weapons that overheat to a certain degree have their ammunition supplies explode or their energy cores fail [in the example of bullet/missle based and energy based respectively], and keeping your weapon at high heat [or other kinds of things, like having your weapon take physical damage] cause it to loose other kinds of performance, like increase chance of jamming [as the internal parts get worn and react to the temp changes] or outright chance of failure in the case of energy weapons.

After so much use, and so much wear,a weapon becomes so undependable that it's best off being dropped by the side of the road :P

The current point of the complexity is to allow a player to exploit subtle weaknesses if they are prepared to do so, and to make certain that no single configuration makes a player highly resistant against everything. Using the stuff discussed above, if your enemy with insane amounts of armor has a particularly vicious machine gun like weapon, with very low heat build up, but low heat tolerance, and is making your life a huge pain with this weapon, it might be the best course of action to use heat-raising weapons specifically targetting that weapon, to force it to fail, explode, and kill the the enemy, since then at least you'll be fighting a heavily armored, highly damaged [or even destroyed completely], unarmed enemy :P


The problem is that i'm not sure the game retains much playability to those who want to just casually play, since it requires this degree of observation. Thats why i'm considering dropping a bunch of the stats, or consolodating them into more general things.

@Fyber: Interesting idea, and definately worth consideration. Catagorizing the attributes with more generalized states that give a sort of overview, to allow for a smoother learning curve [and one that only requires expanding on an area if the player chooses to delve deeper.]

Oops that was me...

Was actually thinking along those lines somewhat, as far as ramping up game complexity, but instead in the form of weapons/armor/attributes/ect that all have low performance, but no drawback, and forcing the player to specialize as they move into the more powerful equipment. I agree though about the problem of finding motivation if i change the project to something i'm not satisfied with.... that has ground more than a couple of my projects to a quick halt.


maybe the volume of statistics won't be SO bad if they are introduced slowly.

Too bad ya can't just trust users to dedicate an hour or two to reading a detailed manual :P

As for reading the manual, I know that after putting a few hours into alpha centauri I did actually read through the whole manual, although I didn't learn as much as I learned from many lost games :) Maybe what you are making is more of a strategy game than an rpg, and if you bill it as such, the people who will play it will be more interested in all the statistics. Although there are a lot of rpg nuts who like statistics as well. As long as you don't have a bunch of redundant statistics and its fun to play, I wont complain. Also keep in mind how difficult balancing a complex system like this could be. OMG. You will be playtesting the thing long after all of the technical issues have been completed. Maybe you could do a board game version that you can play test to find out which things dont really feel right from a gameplay perspective. Sometimes something sounds totally fricking awesome in your head and on a design document, but really just doesn't work at all in a game (for who knows what reason).

I think largely you need to decide who your target audience is; there are some who love pages and pages of numbers, and some who hate them. However, one thing I must disagree on across the board is forcing them to read a manual. This is unnecessary in even the most complex game; you should have all the info and explanations available in game with tooltips and in-game help. This is doubly important in very complex games, especially online ones that will be patched fairly often, since the information in the manual will quickly become outdated anyway.

I don't think there's a simple answer for how to make a great learning curve for a complex game; a lot of it comes from getting other players to play and see how they adapt, and changing it based on what confuses them. In general, you should try to organize everything so that the most important "summary" information is easily available in a big colorful display on the main character sheet, with the option of drilling down into specifics; that way you can cater to the people who like to micromanage and those that just want to go out and shoot things.

For example, on a large list of weapons, you could have the computer automatically look at the player's stats, figure out which weapon would more or less do the best damage per round for them, and display them in a sorted list with nice green arrows representing weapons better than their current one, and red arrows for weapons worse than their current one. Weapons should have non-stat-based descriptions too, like, "This laser gun is good for androids that focus on speed and stealth" and "This laser rifle is good for androids that prefer massive firepower over speed." This should be enough to let players figure out what they want to do. In addition, you could then click on each gun and choose "Extended Info..." to get the huge twenty page list of numbers for those who really want to delve into it, so those that are willing to put the time into plugging away with their calculator before going into battle still have an advantage over those who just buy the reccommended gun, but not an overwhelming advantage.

How far along are you in development?

I'm asking because, if I were in your situation, I'd do my best to pump out a quick and dirty proof-of-concept demo that would utilize as many of the gameplay complexities as possible. It wouldn't even need working AI, just enough functionality for you to see how all of these pages of statistics will actually affect gameplay. Without knowing every last detail of your design it's hard to give specific advice, so having something you can actually play around with might highlight areas where you really are just adding complexity for complexity's sake and other areas where it has a meaningful impact on gameplay.

That said, my solution would be a combination of B and C. If you really do have pages of statistics then it's probably a good idea to trim the fat any way you can, because even extremely dedicated players are likely to suffer from severe apathy when confronted with such a huge volume of numbers and outright ignore a good portion of their options.

Descriptive text should be used to replace statistics wherever possible as well, especially for an RPG. If most weapons have an "average" reliability, then there's no reason to display a 5% jam chance (or whatever) on every weapon's statistics page; that information is going to be ignored anyway and most players would probably assume that most weapons fall into the average. Likewise, there's no reason to list the actual numbers for the few weapons that are more or less reliable, just make a note of their unusually high (or low) reliability in the descriptive text. This kind of information has always struck me as a nice compromise, since "flavor" text doesn't produce the same kind of information overload that pages of statistics tends to.

Finally, don't forget the importance of good interface design. If you can convey information with a picture of a gun and informative text or stats displayed when mousing over different components then that's an infinitely superior method to scrolling through pages of numbers. The more information you can display graphically the less complex and intimidating your game is going to seem.