Okay, so here's the gist of what I wanted to say in my first post that had an epic meltdown. I think that RPGs could be potentially rekindled with an idea of using Randomization of switches during the gameplay of an RPG. The idea came to me when i first heard and start learning about switches a very long time ago and it just hit me again today. Doesn't randomization of switches (more than one preferably) remind you of a Goosebumps book from when you were a child? It did for me when I first started working with them. "If you want to ride the roller coaster, turn to page 88. If you wish to exit the theme park and go for a walk with your friends, turn to page 11." This is a perfect example of what RPGs and other games need in their gameplay to make them more innovative and keep starcraft from getting dull.

What I'm refering to is a variety of switches as conditions based on what has happened in the game. Lets take this to a mathematical level of probability and statistics. Lets say that I want to have a random number of quests that a user has the chance to complete at the beginning of an RPG. I would like to use 3 random switches as the conditions of the quest being chosen by SC at the beginning of the map, so we have Random1, Random2 and Random3 as switches now. Each switch has a value of either on or off, one or zero (as does binary code). This means that we need to calculate the total possible outcomes of the Random(#) switches when we randomize them at the start of our RPG.

In order to calculate the total possible outcomes, we need to use a simple formula known as the "Multiplication Rule". This rule should be applied whenever you are creating a game that has a variety of randomized switches as conditions so that you can control the outcome of your switches having their different orders set off. To further explain this idea, I'll just go directly to the calculation process. So, we have 3 switches that can each carry 2 conditions (on or off). According to the Multiplication Rule, the total possible outcomes is calculated as:

n1 x n2 x n3 = total possible outcomes

where each n(#) represents a specific variable that has a possible outcome of the integer that it holds. So therefore, each n that we have should have the value of 2, because each one of our three switches have 2 possible conditions. If you calculate the total outcomes now, you should get:

2 x 2 x 2 = 8

Lo and Behold, we have 8 possible quests for our heroes to start off with when they play our RPG (only one should be chosen by your triggers). When you put this into perspective, it makes the gameplay completely random (assuming that SC does actually randomize switches) if you keep adding additional randomized quests after the first one that is chosen. This would keep an RPG or any other game that you can apply this to very fresh and original with each time it is played.

Now, I'm sure you realize that this has been used before in many games such as Random Item Defence, but I have never really seen its full use in a really good RPG. The possibilities are endless with this application. Gamers could play a single RPG over and over and over again, having a new experience every time, if randomization permitts. Keeping starcraft as innovative as possible should be key in the design of any game. Sure fancy text and EUDS are nice, but how about try something that makes the user want to play your game 1,000,000 times over because it's different every time?

An example of what I mean is shown below for the trigger and how it should look if you attempt to implement this into any game. First of all you need to randomize your triggers for the initial set of quests which is simple.

TRIGGER: Randomize Beginning of Game
PLAYERS:
-Current Player
CONDITIONS:
-Elapsed game time is exactly 3 seconds
ACTIONS:
-Randomize Random1
-Randomize Random2
-Randomize Random3

Now we need to set the conditions for if each one of the possible 8 conditions are met.

TRIGGER: Randomized Quest1
PLAYERS:
-Current Player
CONDITIONS:
-Random1 is Cleared
-Random2 is Cleared
-Random3 is Cleared
ACTIONS:
-Comment: Quest 1 is Set
-Set Quest1 (triggering quest one to display and set up)


With the simple application of changing the switches from "Cleared" to "Set" you can create the 8 possiblities or outcomes for the 8 different quests that were mentioned at the top of this post. Remember, you must realize that you do not need to use exactly 3 sets of switches for this. You may use the multiplication rule with only 1 switch which would simply be

n1 = total 2 = total

Or you could use two switches to make a possible 4 outcomes as shown:

n1 x n2 = total 2 x 2 = total = 4 possible outcomes

The only thing you need to remember is that the switches depend on the number 2 because they have an ability of 2 outcomes themselves (set or cleared).
If we added this to an RPG they would truly be "adventurous" and be played much more often than the average anyday "classic" RPG because they have more outcomes. Think if we added an addition 2 quests after the first quest was complete from the possible 8, and an additional 2 after that. We would use the Multiplication Rule again and simple calculate how many different ways they can play the game.

n1 x n2 x n3 = total 8 x 2 x 2 = total = 32 posbbile ways to complete 3 quests

Now you should see what i mean. This has many applications besides RPGs as well, we could use this idea in defence games and create shear chaos for certain times that the game is played, and easy times for the others. I challenge you to use this idea and create a game that is not only innovative, but random as well.

If you would like to read up on the Multiplication Rule, I have included a PDF file that talks about probability (nPr, nCr, Multiplication Rule) from when I took it in university. Enjoy, and happy gaming!

Attachments:

Post has been edited 3 time(s), last time on Jan 23 2010, 10:41 pm by JaFF. Reason: Final fix, hopefully

String limits will laugh at your attempt to make enough quests to make this plausible.

Here's my old quest ideas thread for Random Quest RPG, I would ignore most of the first town ideas, however, the ones for the second town are fairly fun to play.

Some ones I added but aren't in that post:

Push cattle to another area: I had 2 critters, that would be ordered toward their destination whenever you walked near them, and after a little while would be ordered to themselves, so they begin wandering randomly if you walk away.

Dragging an anvil to Blacksmith: Kinda boring, but once you picked it up, it made you walk really slow.

I think this would be easier to do, and take less triggers, with single switch randomization.

The RPG I'm currently working on uses randomization at some points for dialogue, placement of an objective (only for side-missions), and a bunch of other things. There are a couple of maps already that almost completely randomize the gameplay - for example Majesty RPG 2 by PCFredZ.

Also, your ideas heavily remind me of my random quest generator idea.

Your ideas are interesting. I like how you have some more simple plans for the quests of your game, makes it look a bit realistic in todays world (odd jobs). If you follow my calculations posted here, you could have 9 larger quests, and 2 quests here and there afterwards for each large quest. The calculation would appear as:

n1 = 9, n2 = 2, n3 = 3 => 9 x 2 x 2 = 36 total possible quests.

The user on the other hand would only complete:

9 x 3 = 27 quests per game (assuming they finish the game up to the end)

This would leave 9 quests incomplete.



Do you have a post of your random quest generator? I wouldn't mand taking a look at it if you did, I think this topic is very interesting, but I would also like to see how random SC really is. I've played many games that claim to use random triggers, but some switches and settings (set/cleared) are favoured over others most times. Does anyone have any posts on how random SC really is?

Post has been edited 1 time(s), last time on Jan 25 2010, 4:13 am by atto. Reason: added another quote

Yeah, if you had clicked on my link you would have had access to a much more powerful and easier to implement system for randomizing DCs.

Here, take a look at this post, and maybe the entire topic http://www.staredit.net/195978/

If you'd like, I can explain how to use this. Basically what it does, is it uses switches to randomly pick any amount of DCs (within mapping limits), if the parameter of the DCs is anywhere from 0 to 2^(x-1) DCs, with any outcome having equal odds of any outcome. Read that stuff and useth it!

Post has been edited 1 time(s), last time on Jan 25 2010, 5:40 pm by CecilSunkure. Reason: Tucking Fypo

Well, I read the entire thread. And wow..... many gamers have little/no interest in taking any computer courses whatsoever. I couldn't believe how many users posted questions about binary, probability, and what is happening.

I only intend to criticize constructively when I say that you have no proof that SC generates evenly distributed random numbres. No calculations or background knowledge of probability are shown in your posts. You merely posted a map that will do so for us. In my eyes, this is not very convincing because you only "reassure" that it is truthful and does indeed work. I know that my words sound harsh, so please take what I just said with a grain of salt.

I do however, find both examples of using death counts very intriguing. Although my method for generating a "random event" in starcraft becomes useless with the limitations of only 255 switches, I believe it is still much easier to understand than your use of DCs (Death Counts) to other users.

Post has been edited 1 time(s), last time on Jan 26 2010, 11:41 pm by atto. Reason: very wrong on comments :(

If you are doing this for some sort of class, then you could argue and take into account human free will. If humans have free will, then randomality can be achieved if it is based around acting from initiation via human choice.

There is no theorem that states that 'everything in life is uniform'. As your number of trials goes to infinity, your estimates of parameters of the parent distribution tend to their actual values. If you run the Bernoulli randomizer I originally posted many many times with the same P value, you will see that it is indeed consistent. If you wish to put it to a very harsh test, set P to, say, 15 and set the number of trials to 100,000. Leave the map running for about 2 hours and 20 minutes and see what the results are. What's the limit on displayed resources?

Yes thanks for catching me on this Jaff. I have no idea what was going through my head when I posted my previous comments. I will get back to you on the whole "random events" as they go to infinity. I know there is a distribution chart that can identify an event of random occurances, I just need to recollect my memory on this subject. The chart should show the distribution curve if the number of random events occuring is sufficiently large.