ZAON Sci-Fi RPG: New setting; Exploring a curved % mechanic

[ZaonDude]

Hello,

Quick introductions and reference links are in order before I describe the system mechanic's parameters we're exploring:

[*]The ZAON RPG official website: //www.zaon.com
[*]Our dot-org customer-focused online discussion forums: //www.zaon.org
[*]Art Example #1: Imperial Hangar Bay with a Phalanx starfighter
[*]Art Example #2: Ranier Station in orbit of Earth
[*]Art Example #3: Skrow TradeCruiser operating in the Triangle
[*]MovieArt (with sound) Example: Imperial Customs Frigate lands on Eridon (Bink self-extracting .exe video 7mb {virus clean!})
[*]Current Galactic layout and political boundaries map: Forum Thread[/list:u]More examples are available at our web sites for those interested.


Down to business:

We initially debuted v6 of our game writeup with a new game mechanic based around a d12 action die and a d8 step penalty die. Since that time we've also been exploring WW-style d6 or d10 dice pool systems.

The problem is that recently I've noticed that a majority of players tend to associate or even prefer a percentile-style game system for sci-fi settings, citing largely 'that it has the correct feel for sci-fi'.

Therefore, I want to explore a percentile game mechanic as a last-minute system swap for ZAON before the corebook is published and thus 'locks us into' a mechanic throughout the entire game line.

First I'll outline the new parameters needed, and then I'll explain why those parameters are important to us. This is of course 'backwards' of normal system design (where normally the mechanic is devised first and then character stats/scores/figures are tailored to match and yield desirable odds), but because of the parameters approach we're taking it is necessary.

In the end, we're looking for help in constructing a replacement dice system that meets as many of the following parameters as possible:

[*]Very simplistic and universal dice mechanic, leaving complexity to a variety choice of stats and skills. That is, we desperately want to avoid 'special case rules' where any exception to the normal dice throw is needed. Simple is in order.

[*]Very fast resolution via a single throw of perhaps two or three dice.

[*]Character Stats must range from 1 to 9 (see below as for why), though 9 is not entirely a hard number it is the highest for the foreseeable future.

[*]Character Skills should range from as tight as 0 to 5 to as finely granular as 0 to 7 or so. I'd prefer a 0 to 6 range. Again, not a hard limit of 6, but near impossible to hold ratings higher than that.

[*]Character creation/improvement difficulty will scale almost exponentially as stats near 9 and skills near 6. Such numbers will be rare.

[*]Stat and Skill power (their effect on probability) should be exponential. For example, say a stat of 2 holds a value of 4%, 3=9%, 4=16%, 5=25%, 6=36%, 7=49%, 8=64%, 9=81%. This gives a '9' the power it needs in a roll. Again, see below for reasons why.

Incidentally, it was pointed out to my math-feeble mind that this particular scale could be achieved on 2d10 whereby both of the d10s must at least equal (but not exceed) the stat (both dice must roll equal to or under the stat score). This may give us the 'percentile feel' we're looking for!

[*]The average human stat must be a '4' and range from 2 to 6. However, using the above system this leaves our average Joe's chances at a base task equal to only 16% which is a problem of course. Some aliens will possess scores below or above this human range.

[*]A percentile or 2d10 pair mechanic should allow for us to use 'doubles' or some other mechanic to indicate a 'critical success' or 'fumble'. For example, an 11, 22, 33, 44, etc., is a critical-something (good or bad). Whether good or bad is determined by whether the roll succeeded or failed. Therefore, the higher chances you have on the %dice the lower your chances of fumbling and higher chances of amazing successes; and vice-versa. This is great.

[*]Degree of success should be determinable somehow. That way, the GM can require '2 successes' for one task, and '5 successes' for a very difficult task. Of course, the method of this doesn't need to be based on successes and so we are open to ideas of how to measure the degree of character success.

[*]Finally, we need provisions for outside influences. For example, a character with a certain stat and skill level may always hold a 50% shot at completing an average task and a 25% chance at a difficult one. But, if that character is wounded, distracted by noise/combat/explosions, or wearing a fumbly space-suit and/or dealing with zero-G, there needs to be a drop in the character's chances without slashing those odds to zero.

Using our existing Step Penalty mechanic, an additional die is thrown with the action dice and that die must roll over the number of step penalties assigned to the character. This, while tough for some players to grasp, is a beautiful multiplicative mechanic that modifies the character's overall probability without the possibility of slashing it to zero like a conventional subtraction modifier would. If this is not perfectly clear, please see this post for a detailed explanation and examples of its use: Step Penalty Description

Therefore, I'm game for keeping this mechanic and developing a better explanation and method of introduction to players so that the purpose and method of the mechanic is better welcomed.
[/list:u]

Now I'll address the reason why a few of the seemingly arbitrary parameters have been set.

I've been evaluating the psychological and emotional interpretations of game stats and dice systems. Specifically, how a stat range from 1 to 9 with a '9' being so very powerful tends to be recognized by players with more emotional impact than apparently free-form numbers such as how stat scores keep going higher and higher in d20 systems.

Though a bit tricky to research, the emotional 'woe' of seeing a Magic The Gathering critter card with a Power/Attack rating of 8 or 9 is greater by those polled than seeing a Yugoloth lord with a STR of 28. It appears to have to do with knowing of a soft boundary (like D&D's 18 attribute level) that makes an '18' a bit of a wow item. MTG does those more so with (last I checked) a 12 being the highest power/attack rating for any creature in that game.

So, seeing a creature (or later in the wargame a starship) with a stat/figure of a 7 or an 8 (or god forbid a 9) seems to carry more weight and 'whoa'.

And, because of this, it's important for that 8 or 9 to be very potent in probability. A 9 should be much more powerful over an 8 than how an 8 is over a 7, and so on.



So, in summary... My feeble mind cannot seem to wrap around the mathematical complexities of a full game system well enough for me to brainstorm or recognize patterns relevant to how dice work. I can only approach this from a marketing standpoint in how it will be perceived and how well/easy it is to learn and play.

If any of you have some ideas you'd like to share about how to devise actual dice mechanics that will accommodate these parameters, please speak up! :-)

--Justin Winters

[Andrew Martin]

[*]Art Example #1: Imperial Hangar Bay with a Phalanx starfighter
[*]Art Example #2: Ranier Station in orbit of Earth
[*]Art Example #3: Skrow TradeCruiser operating in the Triangle

I really like the pictures, Justin!


The requirements:

Very simplistic and universal dice mechanic,... Simple is in order.

Check!

Very fast resolution...

Check!

Character Stats must range from 1 to 9...

Check!

Character Skills should range from as tight as 0 to 5 to as finely granular as 0 to 7 or so.

50%

Character creation/improvement difficulty will scale almost exponentially as stats near 9 and skills near 6. Such numbers will be rare.

Check!

Stat and Skill power (their effect on probability) should be exponential.

Check!

The average human stat must be a '4' and range from 2 to 6.

Check!

...'critical success' or 'fumble'.

Check!

Degree of success...

Check!

...outside influences.

Check!

The dice/system mechanic I'm thinking of succeeds with virtually all your requirements.


The only trouble is that I've suggested using this mechanic before and you've ignored my suggestion. Perhaps you didn't realise that my suggestion will solve nearly all your problems? Are you prepared to look at it again with an open mind? I'm not being nasty, I'm just asking.

[ZaonDude]

Yes, Andrew... I'm approaching the game system with a much more open mind this time. I'm all ears. What have ya got?

Also, I'd very much welcome continued input from Jonathan Walton and Pete "Demonspahn" and others who helped offer ideas in the past that have now helped shape the gathering material for our Release Candidate 1 (version 8) being written now.

There is still time for additional 'myth' ideas for which can be woven into our new background setting material. :-)

[ADGBoss]

Ok I see a few things that bother me about the way your approaching the whole % idea.

Now I think its just fine keeping the numbers as single digit psychologically potent numbers.  I have no idea personally how accurate that theory is but it seems very reasonable.  However, I have never been sold on the X = % style Dice Mechanic.  Why?

Well, if I see a character with a Skill level of 9 I KNOW he needs a 81% chance or less to succeed or something in that area.  It allows the idea of the "min/max" to seep in. "Hmm ok he needs an 81 or less to hit me, ok, lets see I have this power up and this negative and..."  

Secondly it would suggest that anyone with stat 6 and skill 9 si on the same level.  This is not necassarily the case.  Two blakcbelts trained by the same master who can both run 3 miles a day and bench press 300lbs are the same stat wise (generally) but I doubt very much they would fight to a draw every single time.  Thats why we have variable systems.

One suggestion I have is to try this as an equation:

Base Skill % = [(Skill Stat Squared) x 2] + Training

Skill Stat is the Attribute or stat associated with that skill
Training is either a set % (5, 10, 7 whatever) or just the added Experience( i.e. 5exp =+2%) something along those lines.

Roll 2d10 as %

My point is that you should avoid number making more numbers as much as possible.  I have a Skill level of 5 which is 25% and I have a stat of 4 which is 16% and... I have always liked having skills as their own entity. That is once created or learned, they move up by themselves.  That is personal preference of course.

As always thanks for stopping by

sean

[Matt Wilson]

Hey ZaonDude, or uh, Justin:

Back up for a second and give a quick summary of what play is supposed to feel like. If I recall, it's somewhat (or maybe very) Star Wars. If that's the case, what I get out of Star Wars is that there are opponents that you can dispatch with a hand wave, and opponents that take a while.

You could factor that in by giving the mooks abysmally low percent ratings, but then they wouldn't be able to dispatch each other, and we all know that despite their inability to hit Han, Leia and Luke, stormtroopers are really precise shots.

IF you can figure out how to do that with percent dice, you rock the house, and I'll be playing that game like a madman.


-Matt, who loves his space opera.

[Walt Freitag]

Hi Justin,

Wow, tall order here. I peeked at the 24+-page thread on this topic you've got going in your own forum. There's no way I'm going to be able to read it all, but it helped to clarify some of the things you want. For example, I was going to ask whether a 9 opposed by an 8 should have a higher chance of success than a 3 opposed by a 2, but it's obvious from your posts over there that the answer is yes.

I'm going to show you a system based on exponentials, though I can't be certain that's what you really want. Your example is quadratic. But more to the point, an exponentially increasing chance of success cannot work without two artifacts: a point at which chance of success hits 100%, and low chances of success at the other end of the scale (which was also a problem with your quadratic idea).

My own house system is exponential, but it's based on an exponentially decaying chance of failure for the side with the advantage as the advantage increases. (And hence, equivalently, an exponentially decaying chance of success for the side with the disadvantage as the disadvantage increases.) This keeps all probabilities within the >0.0 - <1.0 range and eliminates all edge effects. It has some cool properties; you can read about it in this thread if you're interested. (Pay little or no attention to the details of the actual dicing mechanism described; the important thing is the underlying math and the functional implications in play. The dicing implementation can be modified in many possible ways.)

This system works with attributes (attributes just being another form of modifier, really) that are open-ended (rather than capped as you've asked for) but that still reflect an exponential scale. Each x-point increase in an attribute (or in any other modifying factor) represents a y-fold multiplication of effectiveness independent of all other factors. In the fine-grained version initially described on the linked thread, x is 4 and y is 2 so it takes an increment of 4 points to double the effectiveness. But this is easy to vary. You probably want something close to x=2 and y=3.33 so that you can make an alien ten times as effective (that is, ten times more likely to succeed at a challenging task and ten times less likely to fail at an easy task) by adding 4 points to its relevant attribute. Then you can make another alien ten times stronger still by adding another 4 points. Someone administers a drug that reduces the affected character's effectiveness by about half? Then take away one point. Did some complicating factor (such as rain during a rock climb) just make the task ten times more difficult? Difficulty is just another modifier and it's applied in exactly the same way. Subtract four points. Oh, wait, it's freezing rain? That makes the climb ten times more difficult still. Subtract four more points. Fortunately, the climber has six points of skill (representing about a thirtyfold improvement over raw attributes alone): add six points.

You seem to like multiplication mechanisms. This one multiplies by adding, because what you're adding (the modifiers) are logarithms.

Now all we need is an actual resolution method, a way of making a random roll with a success chance approximately equal to the inverse log (exponential) of the final net modifier score centered on 50%. Here's one:

Memorize this table (or reproduce it on every character sheet):

Code Select Expand

net modifier            –3  -2  -1   0   1   2   3
roll # or less on d10    1   2   3   5   7   8   9

If your net mod is less than –3, you can still try a roll. You must first roll a 1 on d10 to add 4 to  your net mod. Keep rolling until your net mod is in the range –3 to 0, then make the roll on the table. You fail if you fail any roll.

If your net mod is greater than 3, the GM can still require you to roll. Roll 1-9 on a d10; if you succeed, your action succeeds; if you fail, subtract 4 from your mod score and keep going. If your net mod reaches the range 0-3, make the final roll indicated by the table to succeed or fail.

This doesn't meet your specs for the attribute scales, but perhaps it might satisfy whatever deeper purpose you originally had for specifying such scales. (Though now that I think about it, there's no reason not to put a limit somehwere just for the heck of it, which is basically saying things like nothing in the universe is stronger than X. But it doesn't have the impact of, say, a 9 being vastly superior to an 8; it's superior by exactly the same multipicative factor as 2 is superior to 1 – which by the way is 1.778, the fourth root of 10. It is exponential, which is what you asked for.) If you're interested, we could go into details like how base difficulty is established, poosible alternate curves, cricial successes/failure, and degrees of success.

- Walt

[ZaonDude]

@Matt: The feel shouldn't be quite as 'heroic' as Star Wars. That is, ZAON is a bit more realistic in how characters dispatch bad guys... but not looking for a Saving Private Ryan realism level, either.

@Walt Freitag: ahhh.. finally, a math genius :-)  Are you on MSN Messenger or reachable by telephone? If so (and you're up for a chat) please private me with contact info.

@all: I'm also trying to be very open-minded this final time around and so am exploring ideas of using 3d10, 3d6, etc., with perhaps multiplying dice systems: e.g., a d10 'Difficulty Die' that you must succeed at versus a difficulty number issued by your GM else the whole attempt fails.. whereby the odds of this die is effectively multiplied by the odds of the other dice rolls involved when determining a final probability.

But, I'm ever becoming more determined to implement a very simple and straight-forward dice mechanic.

[Andrew Martin]

Yes, Andrew... I'm approaching the game system with a much more open mind this time. I'm all ears. What have ya got?

The system I'm thinking of is based on my Ratio system, described on my site and on the RPG.net Game Design Forum.

Basically, every player character descriptor is rated by a infinite precision percentile figure. An infinite precision percentile can look like: 99.99999%. Note that there's no limit to the number of decimal places! It's just like a real number.

When the outcome isn't obvious and there's a conflict between players, the players involved roll a D10 against the first digit on the most applicable character descriptor's percentile. Note that "0" on the D10 is literally zero, not ten! The number on the D10 roll shows how well the character did, and can be used as the number of concessions gained by the character. For example, my character's most applicable skill is 58%, I roll D10 and get 4. That's 4 concessions (good things) that I can narrate about my character's success.

If the number on the D10 is the same as the descriptor percentile's first digit, the player rolls again, against the next digit in the percentile. Keep repeating this process, until the D10 result is higher or lower than the selected percentile digit. Each re-roll adds "1" to the number of concessions or complications to be narrated.

If the number on the D10 is higher than the percentile's digit, the character has either failed or suffers from a number of complications equal to the difference between the percentile digit and the D10 roll. Complications are bad things that I narrate that affect my character, while still achieving success. Or the player can choose to fail. For example, my character is a excellent marine sniper with 99% skill and is shooting a Alien Bug lookout. My first D10 roll is "9", so I re-roll and get "9" again. Re-rolling again, against the last digit in my character's skill "0", I get "1"; with the two rerolls, that's "3" total. Thinking quickly, I narrate, "my shot strikes home, killing the bug lookout instantly. Unfortunately, a patrol of Bug guardians comes around the corner! I can see at least three of them!" The 3 extra bugs are the 3 extra complications for the character. Optionally, I could narrate, that my character misses the bug for now (failure, so delaying the mission.)

Character improvement is "linear" but gives diminishing returns. Once the first digit hits "9", improvement flows onto the next digit, and so on. For example, my marine sniper character intensively practises her sniping skill, and improves 4 points. Her current level of skill is 99%, so her new level of skill is: 99.4%. After another year of practising, she improves 12 points, the first five points improve her skill from 99.4% to 99.9%, the remaining 7 points improve her skill 99.90% to 99.97%.

50% is average for PCs and 40% is average for NPCs. For most NPCs, the GM can rate the NPC by just writing the tens digit, like

NPC Shopkeeper
All Attributes: 4
Bargain: 8
Stock level: 50%

Degree of success is simply the number on the D10.

Outside influences can either increase or decrease the skill level or can add or subtract to the success level.

Criticals and fumbles are present (they are simply high levels of concessions or complications) and scale with the character's descriptor value.

This version of my Ratio system seems to fit every requirement that Justin has described so far. It's also easily changeable from/to a concessions/complications narration system (from Chalk Outlines) or to/from a conventional RPG success/fail system + degrees of success system.

I hope that helps!

[simon_hibbs]

If the number on the D10 is the same as the descriptor percentile's first digit, the player rolls again, against the next digit in the percentile. Keep repeating this process, until the D10 result is higher or lower than the selected percentile digit. Each re-roll adds "1" to the number of concessions or complications to be narrated.

This leads to weird results. Compare two characters, one with 39.9% skill and another with 41.1% skill. The character with the higher skill succeeds more often, but is much more likely to suffer  'concessions' on a success. In fact the character with the higher skill is 9 times more likely to suffer a concession on success than the character with the lower skill. This seems a little weird.


Simon Hibbs

[Andrew Martin]

If the number on the D10 is the same as the descriptor percentile's first digit, the player rolls again, against the next digit in the percentile. Keep repeating this process, until the D10 result is higher or lower than the selected percentile digit. Each re-roll adds "1" to the number of concessions or complications to be narrated.

This leads to weird results. Compare two characters, one with 39.9% skill and another with 41.1% skill. The character with the higher skill succeeds more often, but is much more likely to suffer  'concessions' on a success. In fact the character with the higher skill is 9 times more likely to suffer a concession on success than the character with the lower skill. This seems a little weird.

That "problem" effectively disappears, if character descriptor values use a range of 10 - 90% in steps of 10%, then 91% to 99% in steps of 1% and so on. :) Like these steps: 50%, 60%, 70%, 80%, 90%, 91%, 92%, 93% and so on.

I also felt that it didn't matter too much as players have the right to have the character fail (or try again later) instead of taking large number of complications. And I envisage players getting their important character descriptor values in the 99% range very quickly.

Can anyone suggest another way around this problem that Simon has pointed out?

[simon_hibbs]

That "problem" effectively disappears, if character descriptor values use a range of 10 - 90% in steps of 10%, then 91% to 99% in steps of 1% and so on. :) Like these steps: 50%, 60%, 70%, 80%, 90%, 91%, 92%, 93% and so on.

Ok, I can see that working. How do you apply modifiers though?

Simon Hibbs

[Andrew Martin]

How do you apply modifiers though?

To greatly increase the effectiveness of a descriptor, one applies another digit in front of the descriptor value. For example, to hit the broadside of a barn, might be a 9 improvement, taking the character's normal skill of, say, 70% to 97.0%

To greatly decrease the effectiveness of a descriptor, shift the decimal point one or two places to the right. For example, to hit a nail head on a fence rail, might be a one decimal place modifier, taking the character's normal skill of, say, 70% to 7.0%

For moderate modifiers, one rolls 2, 3 or more D10 at once, and is forced to pick the best or worst result.

Smaller modifers can directly add or subtract from the descriptor value,  in the same sort of way that improvement is done. For example, a +10 bonus on the character's skill of 95% improves it to 99.6%. A -12 penalty on the character's skill of 99.6% modifies it to 93%.

A skill that helps another skill, could add it's result levels to the other's skills level, in the above way. For example, my character's Hate [Enemy] is 90%, and after rolling the D10, I get 6 result levels which can be added onto the character's combat skill descriptor value.

[simon_hibbs]

To greatly increase the effectiveness of a descriptor, one applies another digit in front of the descriptor value. For example, to hit the broadside of a barn, might be a 9 improvement, taking the character's normal skill of, say, 70% to 97.0%

But surely the modifier must depen on what the characetr's existing first digit is. For example if my ability is 80%, applying a modifier of 7 will make my ability worse, but if my ability is 60% it will make it better. So the same modifier will improve one characetr's success chance, but reduce anothers. So to know what modifier to apply, I have to first ask what the current skill is, and preform a calculation?

To greatly decrease the effectiveness of a descriptor, shift the decimal point one or two places to the right. For example, to hit a nail head on a fence rail, might be a one decimal place modifier, taking the character's normal skill of, say, 70% to 7.0%

For moderate modifiers, one rolls 2, 3 or more D10 at once, and is forced to pick the best or worst result.

So there are actualy three completely different, mutualy incompatible ways to modify an ability? Suppose in a given situation I get three modifiers, one of each type. How do I combine them?


Simon Hibbs

[Andrew Martin]

To greatly increase the effectiveness of a descriptor, one applies another digit in front of the descriptor value. For example, to hit the broadside of a barn, might be a 9 improvement, taking the character's normal skill of, say, 70% to 97.0%

But surely the modifier must depend on what the character's existing first digit is. For example if my ability is 80%, applying a modifier of 7 will make my ability worse, but if my ability is 60% it will make it better. So the same modifier will improve one character's success chance, but reduce anothers.

You're right. I initially thought of using just 9 or 0, then I stupidly extended this to the other digits from 1 to 8. This works best with just using either 0 or 9 as the first digit. As Simon points out, using digits from 1 - 8 causes problems.

So to know what modifier to apply, I have to first ask what the current skill is, and preform a calculation?

That sounds a bit too hard (though people could do that if they wanted to).

To greatly decrease the effectiveness of a descriptor, shift the decimal point one or two places to the right. For example, to hit a nail head on a fence rail, might be a one decimal place modifier, taking the character's normal skill of, say, 70% to 7.0%

For moderate modifiers, one rolls 2, 3 or more D10 at once, and is forced to pick the best or worst result.

So there are actually three completely different, mutually incompatible ways to modify an ability? Suppose in a given situation I get three modifiers, one of each type. How do I combine them?

At the moment, I'm just tossing things into a big pile. Like brainstorming. Editing it to cohere comes later. :)

And yes, one could combine several of these methods at once. Let's take the example of a character trying to hit the broadside of a barn with a firearm. It's very easy to do due to the huge target area, so the character gets to put "9" in front of their applicable skill. The player points out that the character is using a quality pistol, which was previously determined to give +10 steps to the user's skill. The character's Pistol skill is normally 70%, with the +10 steps for quality weapon, that makes 98%. With the modifier for the hitting the broadside of the barn, the value becomes: 99.8%. The player decides to fire twice without bothering to aim, and so rolls two D10 instead of the normal one D10, picking the worst of the two rolls.

That's just an example. One really needs to work through vehicles, space craft, AIs, Expert systems, tools, settings and so on, and decide what penalties and bonuses they give to characters.

Anything else I've missed? :)