Sign in to follow this  
Wavinator

Thoughts on Tactical Caps vs. Unbounded Skills

Recommended Posts

I'm thinking about using a combat and skill resolution system which is both simultaneously simple AND tactically complex. It would work by comparing skills against each other (say in hand to hand fighting) or against some arbitrary challenge rating (say the difficulty in defusing a bomb) in order to get some percentage. This percentage would then be tested against a random "roll" to determine success or failure. That's standard stuff but the wrinkle I'd like to introduce is the idea that some situations or abilities cap skills. The rationale is that no matter how good you are, some challenges are such that nobody can be godly facing them. For example, let's use stealth: Let's say that you have to cross a room with a single guard. Your Stealth Level is 100, the guard's Detection Level is 100. Normally your chances would be 50% to cross undetected. But in this example let's further say that you've triggered an alarm and that this incurs an automatic Stealth Cap, regardless of your abilities. In this situation, the Stealth Cap is generated by a module in the base called Second Sight (a special sensor and AI system located somewhere on the level). Second Sight, when activated, brings all Stealth Levels down to 50. This means that no matter how godly your Stealth Level is, unless Second Sight is disabled your Stealth Level can never be higher than 50 (giving you a 33% chance to cross undetected). Caps can come from abilities, gear or tactical situations and, in turn, can be blown automatically by the same. In the stealth situation, for instance, you might use a Tachyon Jammer which disrupts the signals Second Sight relies on, returning your Stealth Level to 100. Levels would be the general case in most situations, but caps would provide wrinkles. Dogs might cap your Stealth Level, for instance, because they use scent to track you. Or maybe you're a great noble among your own people but scum to the bigoted invaders who have conquered you; so your Reputation Level is whatever it is among your kind, but among the invaders it's capped. The goal with all of this is to provide challenges regardless of your skill levels, to simplify the skill level itself so that you know your general case effectiveness, and to allow for procedurally generated levels where the difficulties can be either generated randomly or keyed off of your character / party details. Thoughts? Is this too arbitrary? Would it invalidate the whole point of leveling if you found some situations brought you down to the effective abilities of a early level character?

Share this post


Link to post
Share on other sites
Maybe instead of capping skills due to environmental factors and such, you could apply percentage changes to them. Rather than limit skill to a max of 50 for the Second Sight, it just reduces skill rolls by a percentage. This seems to make more sense to me, since someone with a 50 Stealth might only have a 33% chance of making it through normally, but the Second Sight thing doesn't affect them at all because it is a cap not a reduction.

Also, it would allow you to increase skills as well (you could wear your semi-invisibility belt module for a 10% boost to Stealth).

Or maybe a hybrid system, where there is a percentage reduction / increase but also a hard maximum / minimum (10% off Reputation with invaders, with a hard cap of 70).

Share this post


Link to post
Share on other sites
I don't love caps - they lead to situations where a higher skill doesn't offer any advantage. Some math trick whereby increasing skill offers diminishing returns could achieve the same result (challenges against which no skill level can do better than a desired percentage of success).

For instance: to roll a skill check take the first ten skill points and roll once floating_point_random (10). Roll the second ten skill points twice and add the lowest result to the total. Roll the third ten points three times and so on. For a 64 skill roll fp_random (4) seven times and add the smallest result to the total. Compare to the challenge difficulty (which itself may be randomized depending on the challenge - a d5+9 challenge would require a level 10 skill level to have any chance to succeed at all - a d15 challenge could be passed by any skill levels with a bit of luck but would require a high skill to succeed consistently) to get success or failure.

[Edited by - Diodor on March 19, 2010 2:08:58 AM]

Share this post


Link to post
Share on other sites
Quote:
Original post by Wavinator
Let's say that you have to cross a room with a single guard. Your Stealth Level is 100, the guard's Detection Level is 100. Normally your chances would be 50% to cross undetected.

But in this example let's further say that you've triggered an alarm and that this incurs an automatic Stealth Cap, regardless of your abilities. In this situation, the Stealth Cap is generated by a module in the base called Second Sight (a special sensor and AI system located somewhere on the level). Second Sight, when activated, brings all Stealth Levels down to 50. This means that no matter how godly your Stealth Level is, unless Second Sight is disabled your Stealth Level can never be higher than 50 (giving you a 33% chance to cross undetected).


You're cooking the numbers too much. When an alarm goes off, the infiltrator's ability to sneak increases because he has a loud noise to mask his footsteps. The purpose of the alarm is to make the guard alert. Instead of zoning out (detection level reduced to 50 due to the amount of time the ability has been in use), the guard is now actively scanning the room for anything out of place.

Share this post


Link to post
Share on other sites
paper & pen RGP shadowrun has interesting way of solving auto success / failure.

playerA [skill level]d6 vs playerB [skill level]d6

exeptions:
(rule 1.) on roll of six dice is re-rolled and added to result.
Rule 2. is removed from re-rolls but rule 1. is still active.

(rule 2.) for each roll of one, higest dice is removed from roll with this dice.


This system makes possible to success in impossible tasks, but it all so makes
bossible to fail in trivial tasks.

/Tyrian

Share this post


Link to post
Share on other sites
Just to be clear, do you mean to say that a guy with 40 skill in sneaking would not be adversely affected by the alarm, and a guy with 55 and a guy with 100 would both be dropped down to 50 by it?

Introducing an arbitrary cap does feel phony. Your example makes me think of a Splinter Cell game or something, but from your other posts I'm guessing you're dealing with a more abstract interface than that, maybe something as simple as a minimap with a room highlighted and "Hangar A: One Guard (Patrolling), Active Second Sight, Exits to Courtyard, Comm Center, Storage Locker A" in a "Current Location" window somewhere.

I'd say a system of buffs and debuffs would be more intuitive to players, so they could see that the room has the "Second Sight" buff on it, which penalizes sneak by 50% and reduces the response delay if the guard calls for backup by 25%, and then they can choose to deploy one of their Tachyon Jammers to debuff the room, or just hope their skill check goes well, or go out the way they came in and try to find the Second Sight console and disable it, or say, "Fuck all this tiptoeing, let's rock and roll," and get out their bazooka and start making their own doors to the data storage wing.

The end result would be about the same--reduced success rate and a series of measures and countermeasures that can be played--but the whole thing might be easier for players to wrap their heads around if it's percentages rather than hard restrictions. Plus it allows a higher skill to confer some benefit regardless of the situation. Even if the benefit is negligible, you don't feel like the skillpoints you allocated were totally wasted.

I think just making a hard cap on total skill level might be the ticket for your overarching goal to prevent godhood. After all, if the skill maxes out at 100, then even a Solid Snake would have to think twice about going into a facility that has measures in place to make sneaking hard. He'd need not just stealth, but a large number of potentially expensive gadgets and/or support personnel to have a good shot at success.

That way, every mission objective has a profile of strengths and weaknesses, "moves" in its repertoire that give it a personal character and flavor. Limiting the number of gadgets you can carry can further enrich the preparation and intel-gathering phase of the operation, so you might spend three months in-game getting a mole inside to learn about the base's abilities and floor plan, plotting the route and setting up contingencies for when things go pear-shaped and kitting out your operative/squad to set up diversions, sneak or fight through defenses, hack computers, all that fancy stuff. You could even have the facilities change up their defenses, or have scheduled maintenance periods for certain key systems, so it would be like planning the perfect bank heist.

Share this post


Link to post
Share on other sites
Quote:
Original post by Iron Chef Carnage
Just to be clear, do you mean to say that a guy with 40 skill in sneaking would not be adversely affected by the alarm, and a guy with 55 and a guy with 100 would both be dropped down to 50 by it?


Yeah that's it.

Quote:

Introducing an arbitrary cap does feel phony. Your example makes me think of a Splinter Cell game or something, but from your other posts I'm guessing you're dealing with a more abstract interface than that, maybe something as simple as a minimap with a room highlighted and "Hangar A: One Guard (Patrolling), Active Second Sight, Exits to Courtyard, Comm Center, Storage Locker A" in a "Current Location" window somewhere.


Right, with some mechanism of showing your options and their chances for succeeding.

Quote:

I'd say a system of buffs and debuffs would be more intuitive to players, so they could see that the room has the "Second Sight" buff on it, which penalizes sneak by 50% and reduces the response delay if the guard calls for backup by 25%, and then they can choose to deploy one of their Tachyon Jammers to debuff the room, or just hope their skill check goes well, or go out the way they came in and try to find the Second Sight console and disable it, or say, "Fuck all this tiptoeing, let's rock and roll," and get out their bazooka and start making their own doors to the data storage wing.

The end result would be about the same--reduced success rate and a series of measures and countermeasures that can be played--but the whole thing might be easier for players to wrap their heads around if it's percentages rather than hard restrictions. Plus it allows a higher skill to confer some benefit regardless of the situation. Even if the benefit is negligible, you don't feel like the skillpoints you allocated were totally wasted.


The only problem I see with applying percentages as buffs or debuffs is that it's harder to bring a disparity involving a big number and a small number into line. If the numbers are uncapped, which is more threatening, to know that your 1000 points will be brought down to 10 or that you'll be cut to 50%?


Quote:

I think just making a hard cap on total skill level might be the ticket for your overarching goal to prevent godhood.


My goal in making skills uncapped was to allow for a sense of nearly unlimited growth and to allow groups to combine skills points toward a single objective. Let's say, for instance, that you've got a door with a shield which can be dispersed provided you can overcome the difficulty level. I was thinking that there would nominally be natural caps (100 sounds good), but that you could put together a team and achieve more from the sum of your skills than you could alone. So if the door was something like 500, procedurally determined based on stats for the area you're in, a maxed team could put 500 against it, giving them a 50/50 chance.

Uncapped skills allow me to do other things like pepper the game world with special locations in the game that break your natural skill point caps and any outside imposed caps (so maybe you find that mystic monastery that can teach you to walk on the rice paper without leaving a single step). I can also throw highly arbitrary barriers in the game world and tie them to time, technology level or team sizes: Maybe the "Entropy Tombs" have some superscience barrier that amounts to a 10,000 difficulty total. A single character maxed has not a chance, but rich entrepreneur or faction leader with a shipload of experts might make it. If I allow you to gradually reduce difficulty in the same way a treasure hunter works to gather clues then I can dole out points here and there, allowing you as the player to determine if 100 here and 50 there is really worth enough to bring down 10k. Or maybe you can go trekking for a phenomenal gadget, like that neural accelerator that will give you 1 turn of 100x skill for the small cost of 50 years of your life.

Quote:

You could even have the facilities change up their defenses, or have scheduled maintenance periods for certain key systems, so it would be like planning the perfect bank heist.


This is cool. I'd like to have this approach be viable, but I also like the idea of not only leveling up but "leveling out." Being a faction leader that can leverage a small army against a problem would be cool as well. (Side note: Without something like this there's never any real way of providing really viable non-combat paths through a game; if you focus everything on non-combat skills and find yourself in a violent situation, this group notion is one potential way to automatically protect you)

In terms of getting the gist of your chances I take your point, but I'm already amenable to having to provide quick calculations to show you the chances of what you can do (and maybe the ability to modify them so you can see what it would take to have this or that chance).

Share this post


Link to post
Share on other sites
I think I'm getting a better idea of where you're coming from. It's beginning to seem like the stats don't necessarily pertain directly to the individual character performing the task, but instead to the situation, or to a relationship between (in this case) the ninja and the room. I dunno how to feel about it, it's still a bit hazy in my mind, so all I can say is that as you explore the idea, you should be thinking of ways to make it intelligible to the player. One of my pet peeves is when a game requires me to interface with a game dynamic, but offers me only the most general and uninformative description of it, like the squad morale system in Ogre Battle or the zodiac compatibility chart in Final Fantasy Tactics.

Share this post


Link to post
Share on other sites
Quote:
Original post by Wavinator
Let's say, for instance, that you've got a door with a shield which can be dispersed provided you can overcome the difficulty level. I was thinking that there would nominally be natural caps (100 sounds good), but that you could put together a team and achieve more from the sum of your skills than you could alone. So if the door was something like 500, procedurally determined based on stats for the area you're in, a maxed team could put 500 against it, giving them a 50/50 chance.

What if a single team member with a stat of 100 attempted it? Would his chance be 1/6? Could he just attempt it 6 times in a row?

I wonder if the system needs a kind of "natural 1 always fails" and "natural 20 always succeeds" element added in. A threshold that needs to be overcome before even having a remote chance of success.

The door's threshold could be 250, meaning that a combined stat of 249 would have 0% chance of forcing it open. At 500, the chance becomes 50%. Maybe at 1000, the chance becomes 100% (automatic).

I also imagine there may need to be some kind of diminishing returns on combining stats. A coach drawn by four horses doesn't go twice as fast as a coach drawn by two horses.

Share this post


Link to post
Share on other sites
Quote:
Original post by Silvermyst
Quote:
Original post by Wavinator
Let's say, for instance, that you've got a door with a shield which can be dispersed provided you can overcome the difficulty level. I was thinking that there would nominally be natural caps (100 sounds good), but that you could put together a team and achieve more from the sum of your skills than you could alone. So if the door was something like 500, procedurally determined based on stats for the area you're in, a maxed team could put 500 against it, giving them a 50/50 chance.

What if a single team member with a stat of 100 attempted it? Would his chance be 1/6? Could he just attempt it 6 times in a row?


Yes if nothing stopped them, which I *think* I'm actually okay with because of what I'd like to do in terms of how these attempts would work. I imagine that in order to get unlimited opportunities to do something you've somehow isolated the challenge so that you're not interfered with by anything and you have all the resources you need. Hacking a door, for instance, should consume something which you could potentially run out of.

In terms of combat, I'd like *most* locations to not run out of challengers (particularly escalating challengers). If you neutralize all the guards in a base, for instance, and the base can call for backup then the militia shows up. Take them out and the military arrives. After them come the elite guard, etc. etc.

How long it takes to make an attempt and how time and reinforcements arrive will all need to be designed to make this even relevant, though. But I have no problem with you taking an age to break through a vault only to exit and find yourself surrounded by several counties worth of law enforcement. :)


Quote:

I wonder if the system needs a kind of "natural 1 always fails" and "natural 20 always succeeds" element added in. A threshold that needs to be overcome before even having a remote chance of success.

The door's threshold could be 250, meaning that a combined stat of 249 would have 0% chance of forcing it open. At 500, the chance becomes 50%. Maybe at 1000, the chance becomes 100% (automatic).


Thinking about this maybe adding a threshold along with caps *might* make it more interesting (though responses so far don't favor caps at all).

Auto success I want to be a function of actually tabulating attempts, rather than a chance roll. So I'll work out "the chance of getting at least a roll of X" over successive "rolls" and present that as the number of attempts needed. (I want to be able to make time passage variable, with this being the equivalent of D&D's 'you wait for 20 turns and automatically succeed' but with a bit more variety)


Quote:

I also imagine there may need to be some kind of diminishing returns on combining stats. A coach drawn by four horses doesn't go twice as fast as a coach drawn by two horses.


I didn't mention this but I've got this. Successive characters get tallied at fractions of their skill naturally in what amounts to a curve of diminishing returns. The amplitude of the curve, however, is determined by skills and abilities (leadership, for example).

Btw, I also throw in a separate "roll" to determine degrees of success and failure (with critical success/failure included) which determine things like resource use/wastage or detrimental side affects. So that's another way to limit things.

Share this post


Link to post
Share on other sites
Quote:
Original post by Wavinator
For example, let's use stealth:

Let's say that you have to cross a room with a single guard. Your Stealth Level is 100, the guard's Detection Level is 100. Normally your chances would be 50% to cross undetected.

But in this example let's further say that you've triggered an alarm and that this incurs an automatic Stealth Cap, regardless of your abilities. In this situation, the Stealth Cap is generated by a module in the base called Second Sight (a special sensor and AI system located somewhere on the level). Second Sight, when activated, brings all Stealth Levels down to 50. This means that no matter how godly your Stealth Level is, unless Second Sight is disabled your Stealth Level can never be higher than 50 (giving you a 33% chance to cross undetected).
[/i]


When I read this example, what comes to my mind instantly was: Why artificially restrict skill levels?

Why not instead have the triggering of the alarm resulting in more guards spawn, greater guard sight radius (more alert guards), activation of dormant security devices etc? It achieves the effect of raising the difficulty in the event of an alarm trigger.

Share this post


Link to post
Share on other sites
Quote:
Original post by Girsanov
Why not instead have the triggering of the alarm resulting in more guards spawn, greater guard sight radius (more alert guards), activation of dormant security devices etc? It achieves the effect of raising the difficulty in the event of an alarm trigger.


This would probably be the more straightforward way of doing it, but how do you feel about having a dozen or so stealth scores? The cap idea is meant to sidestep the problem of having a separate category for several different classes of threat. Rather than saying "here's your infrared stealth level, your ultrasonic stealth level, your stealth level against pheromone detection, your stealth level against psionic dogs from Rigel which just happen to also be able to detect pheromones too, etc., etc." you get one cap against one linear skill.

(Heh, the fight for caps isn't going well :P)

Share this post


Link to post
Share on other sites
I think caps can have a place, but you have to be careful where you use them. An alarm makes guards alert, but they don't make a character less stealthy. However, you can only hide so well under a spotlight. But they aren't the same thing, so players have the option to avoid searchlights so the guards don't find them.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this