News:

Forum changes: Editing of posts has been turned off until further notice.

Main Menu

Experience Point Simulations

Started by Mark Stahl, June 08, 2005, 09:34:34 PM

Previous topic - Next topic

Mark Stahl

After reading the 'Indie Game Design' notes concerning posting policy, I decided to err on the side of caution, since there is no specific game to which this applies.

As I was working on character creation, the usual questions popped into my head.

What level of game play (as a general rule) should each game be run/start at?
How long would the average game last (not hours in a session, but weeks or years in a story)?
Given a certain length of play time how powerful should the average character be?

Normally this wouldn't bother me too much, but I have bad memories of playing single character on-and-off in a live-action Vampire game. It became painful obvious, after the character reached its third year, that WW's live-action system was not built for long running games. Or maybe the experience points given where just too much week-per-week. Nonetheless, whether good or bad, balancing the distribution of experience is important, and in regards to live-action Vampire, not handled well.

So I am here to make a small recommendation. Make an experience point simulation.  If you come from a programming background, there are a wide array of tools one could use for such a project. Any general purpose programming language would do.

If you do not know any programming languages (HTML, in this instance, does not count) then I recommend a paper and pencil method, or hell, you could probably use Excel.

All in all, it is hard to divine the future and predict how GM's will run their games, or how fast a character will advance week-per-week, but you can make informed decisions. Currently I working on a few python scripts for my own project. Although my project modifies Fate's beautiful mechanics, I am sure it would make a decent starting point for other systems. I will post these files online when I feel they have reached a point of proper representation, and post the links to these files on the board as well as my development blog.

Thank for taking the time to read, and hopefully consider.

MJ Stahl

Andrew Morris

Quote from: mjstahlWhat level of game play (as a general rule) should each game be run/start at?
How long would the average game last (not hours in a session, but weeks or years in a story)?
Given a certain length of play time how powerful should the average character be?
Well, that depends on what kind of game you are looking for. There's too many undefined variables here to give a single answer or "general rule."

As to creating a simulation, well, okay, if one of your main design goals is having characters that advance in such a fashion. Again, though, I can't imagine an answer that would address the many options out there. Heck, some games don't even have any experience/advancement rules, and are just fine without them.

Trying to create something like this is similar to saying, "How fast should a car go?" Without any context or stated preferences, this is impossible to answer, other than by saying, "As fast as you want it to go."
Download: Unistat

FzGhouL

I start my characters off at level 0, a totally un-unique template, then when they get the hang out it, things start.

I think experience should be based on challange rather than repitition. If a character goes through a trial that is difficult, they should learn more than a trial that is easy. This is 100% subjective, and you cannot program subjectivity.

On the other hand, you could have a tertiary slope on a series, if you want some sort of "objective" Exp curve that come in RPGs, where you have a series as follows

Sum 1 to n, an^3+bn^2+cn+d, where a,b,c,d are increases in slope as n rises. It will grow quickly.

Or, you could have a natural function modeled by polynomials as

Sum 1 to n, (x^n)/n! where x is the curve. This growth will approach e^x and n equal to infinity, so you can set the maximum from the slope.

To program this, I'll show you in pseudo code;

Define n as integer (level)
Define x as floating point (Exp)
Define a variable to store the value of  experience needed, like y.

loop
 y = y + function you model.
 loop until x = y
    x = x + experience earned.

Etc etc.

Subjectively rewarding is much better though.

komradebob

MJ:
QuoteNormally this wouldn't bother me too much, but I have bad memories of playing single character on-and-off in a live-action Vampire game. It became painful obvious, after the character reached its third year, that WW's live-action system was not built for long running games. Or maybe the experience points given where just too much week-per-week. Nonetheless, whether good or bad, balancing the distribution of experience is important, and in regards to live-action Vampire, not handled well.

If you're looking for suggestions to fix that, I've got a ton. All of them are likely to get you burned at the stake by the other participants, however, and not your character either.

The second part? No idea.

Robert
Robert Earley-Clark

currently developing:The Village Game:Family storytelling with toys

Ron Edwards

Hello,

MJ, a lot of what you're talking about is consistent with a long-standing point of mine: that the heart of any role-playing system is its reward procedures. This is actually bigger than the limited issue of "how characters improve," although that may be included.

Have you checked out any of my essays in the Articles section? If not, see "GNS and other matters of role-playing theory," especially the stuff on Currency. What you're describing sounds a lot like "running the numbers" in the economic sense, regarding the Currency of a given game system.

Best,
Ron

Warren

Someone (I wish I could remember who) once said "Experience is what you get when you don't get what you want". I really like that, and I likes the way DitV represents that. I know it's not a very d20 way of doing things, but I would have thought that you get some experience points for each roll you fail, or each time you got hit (maybe). If I recall correctly, this is like the BRP system of having to fail a roll of a skill to get an improvement in that skill, which had a side-effect of slowing down effective advancement the more skilled your character became.

The problem with this could be that the player got bored of having a relatively static character, and would want to create a new one; the problem with this being that unless the whole group wished to do likewise, the new character would need to be "pumped-up" to a similar level on generation so that the player could still step on up alongside the rest of the party. This, of course, meant that the player was still stuck with a static character, essentially.

Troy_Costisick

Heya MJ,

QuoteSo I am here to make a small recommendation. Make an experience point simulation. If you come from a programming background, there are a wide array of tools one could use for such a project. Any general purpose programming language would do.

That's a fine suggestion and it may very well be worth the time of those using that reward mechanic to do such a simulation.

But the real question should be, "What do you want the characters to be capable of doing?"  The numbers are important, but the meaning behind the numbers is what really affects play.  Levels, whether they take 100 XP to reach or 10,000 XP to reach, are nearly as important as what that stage of development grants the character and the player.

Peace,

-Troy

M. J. Young

Boy, was I thrown when I started seeing posts addressed to "MJ"; and I'd just commented elsewhere about how useful it was to have a moniker no one else seemed to use. Well, welcome to the Forge, MJ; you've got a good name.

I think what MJ is saying, really, is when doing your game design, if you expect player characters to improve over time, figure out how fast you expect them to do this, and make sure whatever method you have to drive these improvements actually keeps to the pace you want.

Most people do that by playtesting, but you'd have to playtest over the long haul to really see whether you get that outcome (and you'd need to have several different groups run by several different referees, to see how different it was between them). Using mathematical modeling to get those answers makes good sense, if that's what you're doing.

My own answer was to put significant control of these variables in the hands of the players, who can increase whichever abilities they wish to the degree that they're willing to kill time away from the adventures bettering themselves. This tends to create characters who reach a solid level of ability and then plateau, because the players are content with what they can do and put themselves to doing it.

But there are other effective ways, and it's important to understand how yours works.

--M. J. Young

Mike Holmes

Yep, long term playtesting is problematic. For example, I used to work on the demo team for the game Europa (WWII in Europe on the batallion scale). The game takes a dedicated group of several people meeting regularly about as long to play as the war actually took to fight - several years. So how do you playtest such a thing?

Well, you don't, most likely, is the ugly secret. In fact, most such games never get played to completion much, if ever. What you do as the designer is to play some of the game, and extrapolate from the play you saw, how long things will take. Perhaps in three months of play, three of four characters managed to level up once. From which you interpret that the standard rate of advancement is one level per X sessions.

This is, of course, highly prone to error. In fact, it turns out that Europa just falls apart after 1943. Literally, the game is broken as written. But nobody had the time to play the game out for four years to find this out. It was only after many years of the game being on the market that somebody good actually played the game out to that point, and discovered the flaws.

So, consider this: while I have a deep sympathy for long term games, and the feelings of accomplishment that they can deliver, instead of working on the durability of a game over the long haul, find ways to condense those feelings of accomplishment into shorter time periods. I mean long-term games are just prone to problems, good rules or no. Somebody will move away, get married, have children, get sick, get a new job with tough hours, find a new hobby (that takes less time), or just become disinterested over time. Yes, some few long-term games survive. But that's despite the things making them unlikely, not because they're really all that especially rewarding for their length. I truely believe that.

So, instead of making play last as long as possible, with X rewards in the process, find a way to make those X rewards happen as fast as possible. You'll get the same overall reward, with a more sustainable play model.

Mike
Member of Indie Netgaming
-Get your indie game fix online.