[Ur] Thoughts on Initiative

[ODDin]

Ok, as my last post was pretty much ignored, I'll narrow it down to just one issue.

I've been working on my combat system, mainly on the subject of Initiative, and had the following general idea. There is a concept of Time Units (TUs). Roughly, 50 TUs is a single second, but that may change, and this scale is completely ignored in the examples that follow.
Now, at the beginning of combat, initiative is decided upon (the general range is 1-100). Surprise is accounted here, and simply adds to the initiative (I don't yet know how). Then, the difference in the Initiative scores is the number of TUs that pass before the other people become aware of the action.
For instance, John has an initiative of 80, Scott of 60 and Jill of 40.
John begins his action. 20 TUs after he started his action, Scott is ready to react and start his action (80-60). So, if making a strike with a sword takes only 10 TUs, John can cut Scott's head off before Scott can attempt to dodge. If, however, 20 TUs is only enough to lift the sword up into the air, then Scott can begin dodging, or may attempt to stab John with his knife, or whatever. Jill needs to wait for 40 TUs to pass before she can react, of course.
Now, this all seems nice, but the problem arises when people need to react to more than one action, and when people high in the initiative need to react to people low in the initative.
For instance: when John is hodling his sword high in the air, Scott pulls a knife out of his belt and attempts to stab John before he hits with the sword. How many TUs pass before John can react to that new situation (dodge the knife, cut at Scott's hand instead of his head, etc)? And how much time has Jill to react to Scott's action? Would it be 20 TUs (60-40), or perhaps more, since it's a thing of reaction time; dhe needs to wait 40 TUs to react to John's actions, so she should probably wait around the same number of TUs before she can react to a different action.

Essentially, I envision combat without rounds, but rather built on these reaction times. For instance:
John does something, Scott has 20 TUs before he can react, Jill has 40 TUs before she can react.
Now Scott does something. John has 15 TUs before he can react (until then he continues his previous actions unabated), Jill has 20 TUs before she can react to that action (that is, she will be able to react to both actions at simultaneously).
Now John does something again, and you have x, y, z TUs to react.
Etc...

However, I am really lost on how to decide on these reaction times, especially when someone with a higher initative needs to react to someone with a lower initative (yet there is obviously some reaction time needed, nobody reacts in zero time). Any ideas are welcome.

Various actions will, of course, take different numbers of TUs. In 20 TUs one may be able to dodge aside a deadly blow or utter a quick protective incantation, but not draw the sword that's on his back or open his backpack and pull out a magic ball.

Another note: Don't be concerned with the amount of bokkeeping required (to keep track of all the actions one needs to react to and the different numbers of TUs he needs to wait for each action), as I'm intending it to be a system used alongside a computer program, and I generally strive for maximum realism, whatever the cost.


Michael Pevzner

[J. Scott Timmerman]

Aloha Michael,

So will you have the computer equipment set up so all the players have a convenient interface, or just so that it's not a problem for the GM to handle everything?  In any case, if you're letting a computer handle all the numbers, why bother creating a new unit at all?  You could use milliseconds, but I suggest just using seconds, in a floating-point variable.  It will keep things simpler for you when you work with data you'll find elsewhere, and show your data on the screen in a more coherent fashion to the players and/or GM.

I just spent a minute looking at Wikipedia and found a couple articles.  The one on Reaction time isn't very detailed.  There was a link to their article on Mental Chronometry.  At the bottom of that page are a list of magazine articles you might want to look up on the subject.

My suggestion would be, if you can find any random samples of data in any of these articles, would be to set up your character statistics and the program to decide character's reaction time based on their statistic.  Seems simple enough.  But I'm sure you also want to account for the difference between waiting for something to happen and reacting to it, reacting to the unexpected, or reacting to stimulus that is hard to comprehend.  As far as whether readily available data is available on that, I wouldn't count on it.  From the Wikipedia articles, it looks like it's mostly reacting to flashing lights or sounds.

But I hope that's at least a starting point.  I know it might take no small amount of work doing it this way, but at least you'd be basing it on scientific findings.

-Jason T.

[ODDin]

Hello there,

Regarding the computer: I'm intending the interface to be available for the GM only. First of all, it is technically very tricky to have an interface for everybody. I mean, it's reasonable to ask of a gaming group to have a computer conveniently available to the GM, but how can you ask everyone to have access to the computer? Even if it's a laptop, it would need to be passed over all the time, and that's rather inconvenient. Second, it only helps the realism. Just as in real life you don't really know that you have a Strength of 78, you shouldn't know it in the game either. You know you're pretty strong, you see what you manage to do and what you don't manage to do, but you don't know the exact number. Same goes for wounds or other things like that.

Also, I'll probably agree with you and switch to ordinary units of time. Milliseconds, probably.

As for the matter at hand. What you said actually gives me a very good idea, and in fact, it makes my current job a lot easier. You see, the core system is just a toolbox with the use of which worlds can be created. So, all I need to do is to provide the rules for how the already calculated reaction time needs to be used - the job of actually calculating it is a matter for future designers. After all, perhaps it works differently for Elves, Humans, Orcs and whoever.
Granted, after I finish with the core system I will start working on the real thing for an ordinary human (for a start), but eh...

So now I think it will work something like that:
John does something. A check is made for how long it takes him to perform the action. Then, other characters need to check for perception through whatever means (spot, listen, smell, whatever) against whatever John is using to hide his actions (sneaking up behind them, for instance). If nobody senses anything, John may proceed until another check is granted - this is his "surprise round", so to speak. When somebody does notice the action, he checks for his reaction time to such actions (possibly also accounting for mental stress, fatigue, fear etc.), and that's the time until he gets to do something.
Same goes for every action, although it's normally easier to notice an action when one's in the middle of combat, so the element of surprise is largely reduced.
Again, there is no concept of rounds or turns, only the time it takes to perform specific actions and the time one has before he can react to it.
Of course, although the system itself won't provide the specific way to calculate the reaction times and methods of pereption, it will be more elaborated than this. This is only the general concept.

So, what do you think?


Michael Pevzner

[J. Scott Timmerman]

Interesting, Michael.

First of all, it is technically very tricky to have an interface for everybody.

I'm pretty sure there are interfaces out there, sold commercially for this.  At least, I've seen game-show style USB buzzers and stuff that can accomodate large numbers of players.  I was actually thinking you might just get some old 9-pin (Atari, Sega Genesis, MS-DOS) game controllers and an older computer; stuff you can get at some garage sale or pawn shop, pretty cheap.  If you hooked the hardware up right, everyone could have an interface.  And then you'd just have to make sure everyone had a clear view of the screen, whether by projection, hooking it up to a TV, or whatever.  But that's just an idea.

Again, there is no concept of rounds or turns, only the time it takes to perform specific actions and the time one has before he can react to it.

Of course, you have to think about communication, especially if the GM is running everything.  Do the Players just call out their actions, and the GM puts them into a buffer until their character's can actually act?  Or does the GM wait for each moment any character senses something, announce it, and then wait until each character is able to act, announce that, get feedback, and then type it in?

Since you're dealing with very minute amounts of reaction time and everything, handling the communication of the actions throughout this process is going to possibly make each in-game second take several minutes at least.  Is that what you want?

One way or another, whether the players are entering the information or the GM is, you're going to have to deal with all possibilities that can affect the computer calculations.  I suppose it might be easier to handle this with a keyboard than with a game controller.  So the game controllers idea above isn't perfect by any means.

It seems cool yet daunting as a task, to set up the system to be able to handle inhuman cognition.  I like that things like precognition (like spider sense) would have effects that make more sense than what they do in other systems.

But System isn't the only part of a game.  From what you're telling me, you won't actually have a "Game" available, just a "system."  It appears that, instead of character creation, you'll have some sort of "race creation" rules that are required first.  You have presented a vague idea of "situation."  It seems like the situations that would make your game shine is heavy combat/survivalist style stuff.  Hunt and be hunted is the idea I get.

So yeah, if you can program it such that the computer reduces everything (both math and otherwise) down to a reasonable handling time, and if you could get maybe a module which has enough interesting non-system bits to make a game, I'd certainly be willing to try the game out. 

-Jason T.

[David C]

I'm telling you my experience, in case it might help you.

I first made a system in high school that used something similar to what you are describing. My idea came from a game called "gunbound" where different actions took longer to do, and the next person's turn was whoever had accrued the fewest "TUs." In my system (which I was using for a gurps home rules set) actions took 1-5 "TUs."  Each time a player did something, they declared how many TUs they took (based on how long the rules say it takes), and I made that many Xs in their column on gridpaper. I than just looked to see who's column was shortest, and than it was their turn. If people had the same columns, they went simultaneously. For example, Fred runs forward (1 TU) and Sarah shoots her gun (3 TU). Sarah misses fred.
Fred  Sarah
X ____X
______X
______X

It is Fred's turn since he has the shortest column. He decides to bite Sarah's arm. Fred bites Sarah's arm.
Fred Sarah
X___X
X___X
X___X
Sarah and Fred go at the same time. etc. etc.

It was interesting, but definately a gamist (or is it simulationist?) solution, and not one that many people are able to or willing to do.

[ODDin]

I'm pretty sure there are interfaces out there, sold commercially for this.  At least, I've seen game-show style USB buzzers and stuff that can accomodate large numbers of players.  I was actually thinking you might just get some old 9-pin (Atari, Sega Genesis, MS-DOS) game controllers and an older computer; stuff you can get at some garage sale or pawn shop, pretty cheap.  If you hooked the hardware up right, everyone could have an interface.  And then you'd just have to make sure everyone had a clear view of the screen, whether by projection, hooking it up to a TV, or whatever.  But that's just an idea.

The thing is, it's a game that is supposed to be for everybody, not just for my group and me. While it's reasonable to say "to play this game, the GM needs to have access to a computer at the gaming table", it is much less reasonable to ask for what you're proposing here. It might be pretty cheap, but consider that it's actually becoming an added price to playing the game. Don't know about you, but I wouldn't buy the rulebooks of a game that starts with a 20 page instruction on how to create some weird homemade interface out of old computers and game controllers... Or worse yet, demanding me to buy some interface like that (which would cost a lot more, and won't likely be used for anything other than this game).

Of course, you have to think about communication, especially if the GM is running everything.  Do the Players just call out their actions, and the GM puts them into a buffer until their character's can actually act?  Or does the GM wait for each moment any character senses something, announce it, and then wait until each character is able to act, announce that, get feedback, and then type it in?

Since you're dealing with very minute amounts of reaction time and everything, handling the communication of the actions throughout this process is going to possibly make each in-game second take several minutes at least.  Is that what you want?

I would go for the last option. The GM tells each character what he senses. Then, when the time comes for the first character to react, the GM asks him what he does. Then, other characters react to the event, or to what the previous character did, etc. The GM would have an interface that tells him who is the next character to address, and to what is it that he is now able to react to. If two characters need to act at the same time, the GM decides who to ask first, but the order of him asking won't actually affect the game. Naturally, the players need to remember what they know or don't know, but it's not that hard. There shouldn't be really many actions that characters can't yet react to at any given moment, if they are not of some weird race that has a reaction time of two minutes...

Now, as for the time it takes - I sure hope the computer program (which I myself will be writing) will be efficient and convenient enough to speed things up and allow the GM to concentrate on what is important, and not be bogged down with all that's going on. Other than that, from my experience with D&D, rounds (6 seconds) tended to take quite a lot more than 6 seconds...

But System isn't the only part of a game.  From what you're telling me, you won't actually have a "Game" available, just a "system."  It appears that, instead of character creation, you'll have some sort of "race creation" rules that are required first.  You have presented a vague idea of "situation."  It seems like the situations that would make your game shine is heavy combat/survivalist style stuff.  Hunt and be hunted is the idea I get.

I began working on the game with the idea of creating the most realistic game possible. Eventually, I decided that I can't cover all setting and genres. Instead of taking the GURPS approach of "provide the designers with everything and let them weed out what they don't need", I decided on "provide the designers with an empty field and let them plant what they do need". Thus, there are very complicated rules on creating lists of abilities and skills, functions for calculating reaction time, perception, body models for analysing wounds and damage etc. Naturally, the "System" itself is completely unplayable. It's a lot of work and playtesting for each race created. This is why when I finish with the core system, I will start working on my first race - a human! After a lot more work, there will be something playable. Preferably in a world with no animals.
Of course, I'm not intending to stop working on it. I will continue creating more and more races, for the enjoyment of me and other players. After all, what are supplements for?

As for combat/survivalist - well, the combat part is what I'm working on at the moment, so that's why I'm asking about it, but I hope the system will work fine enough in other situations too. The section dealing with abilities and skills goes into much effort to provide with accurate rules on improvement and experience, and how abilities affect each other. From looking at it, the system seems pretty much obssessed with character development.


@ David C:
Well, your idea does resemble what I've got a lot. My system is highly simulationist, yes, in the sense that it strives for maximum realism. I know that not everybody really wants or needs an extremely realistic system. Not everybody is a simulationist, after all. However, I hope there are those to whom this system will appeal.

[J. Scott Timmerman]

Now that I see your goals for publishing, I agree that you shouldn't want to make the costs too high for potential players.  If this niche exists, it's right now just a bunch of people who already use a computer to help with non-computerized games.  Having an actual computerized game might be worth a try.

I've also worked with having actions with different durations, and reaction time.  But both you and David C are going very different directions than I have with my system.  David's actually seems a lot like Exalted 2nd Edition.  Maybe there are people on here that share your vision and can give you better advice than I can.

If one of your goals is to get the game at least as fast as D&D, I'm not saying you can't do that.  It's probably possible.  It should be even easier to get it faster than GURPS.  The survivalist ideas came from your ideas about reacting to ambushes etc.  I just imagined travelling through wilderness, sneaking up on prey or being snuck up on.

-Jason T.

[Adam Riemenschneider]

Um, wow. This is so close to what I've already cooked up it's kinda scary.

What I went with in Factions is an Initiative check before each action. This measured against a chart gave the amount of lag (in tenths of a second) before the action began. Then the action took place (standard time was 10 units, or 1 second). There are ways to modify the length of an action (declared Rushed, Quickly, etc) that increase difficulty. You declare your action, roll for Init, wait out your lag, then the action happens.

Meanwhile, the Ref just calls out phases. When he gets to the phase where your action "finishes," you roll for hit and whatnot. Then you declare, roll Init, and continue.

It drags if combats take more than a handful of seconds, which is why I went with both Phased and Non-Phased combat rules. Can use Non-Phased when it's a bigger fight involving more folks, and Phased for very high-intensity moments.

We've never needed a computer in playtesting, though I'd scratch down the next phase when characters were going to act. Worked out fine.