# Don't halve skills

#### burdock

##### Mongoose
why cant increased chance at criticals be the advantage of +100% (non-halved skills) chraracters? Really simple. And perhaps introducing "special success" at 20% of ability

I'm all for this!
But, we would have to add in critical rules for opposed rolls (by the book can't critical in opposed tests).

atgxtg said:
I'm all for this!
But, we would have to add in critical rules for opposed rolls (by the book can't critical in opposed tests).

I don't see why any extra rules are required. There are already rules for the effects of critical successes with skills. Just apply those.

Simon Hibbs

The key problem here is that it's not a simple success chance, it's an opposition. The conflict resolution system is meant to mimic the Resistance Table, and so far the only one that's mathematically identical is the one I proposed in another thread:

If the difference is greater than 100, the highest has a flat 95% chance of winning.

Otherwise, subtract a multiple of 100 from each to bring them both back to a 1 to 100 scale, then resolve as normal.

Quick, clean, simple, and mathematically identical to the Resistance Table.

GbajiTheDeceiver said:
The key problem here is that it's not a simple success chance, it's an opposition. The conflict resolution system is meant to mimic the Resistance Table, and so far the only one that's mathematically identical is the one I proposed in another thread:

If the difference is greater than 100, the highest has a flat 95% chance of winning.

Otherwise, subtract a multiple of 100 from each to bring them both back to a 1 to 100 scale, then resolve as normal.

Quick, clean, simple, and mathematically identical to the Resistance Table.

Yes, I know. THat is one of the ironies of removing the resistance table. You system doesn't quite mimic the reistance table though simply becuase the opposed system uses a two roll method. Otherwise Skill A-Skill B +50% could be used.

Even the rolemaster method of adding skill to an open ended D100 and high roll wins would work fine.

But the low roll wins, toss in criticals method still gives good results.

GbajiTheDeceiver said:
If the difference is greater than 100, the highest has a flat 95% chance of winning.

Otherwise, subtract a multiple of 100 from each to bring them both back to a 1 to 100 scale, then resolve as normal.

Quick, clean, simple, and mathematically identical to the Resistance Table.

What do you do if one character has say a 90% chance and the other has 110%? Giving the higher skilled guy a 95% chance of success is unfair, and you can't deduct 100 from the guy with 90% either.

The resistance table approach is to add the difference between the skills to 50, and that's the chance of the higher skilled character winning. In this example the target number would be 70%.

Also, the resistance table used a single die for resolution which will give a flat statistical spread. Any system that uses two rolls combined will produce a statistical bell curve in the results.

Simon Hibbs

simonh said:
What do you do if one character has say a 90% chance and the other has 110%? Giving the higher skilled guy a 95% chance of success is unfair, and you can't deduct 100 from the guy with 90% either.

The resistance table approach is to add the difference between the skills to 50, and that's the chance of the higher skilled character winning. In this example the target number would be 70%.

Also, the resistance table used a single die for resolution which will give a flat statistical spread. Any system that uses two rolls combined will produce a statistical bell curve in the results.

Simon Hibbs

Yeah! Gound sound reasoning and good sound math!

What are the odds?

simonh said:
The resistance table approach is to add the difference between the skills to 50, and that's the chance of the higher skilled character winning.

So just playing Devil's Advocate for a moment... let's say you have a dude at 125% and a dude at 25%. The difference is 100%; added to 50 you've got a 150% percent chance he'll win.

What then?

What then?

Well, exactly what you would think -- the higher-skilled guy wins. IIRC, the resistance tables was based around the concept that at some point, if one opposed quantity completely outclasses the other, success is guaranteed (like a STR 20 person picking up a SIZ 1 book).

Actually, I think the table allowed for a 1% chance at either end of the spectrum for those Hollywood-style 'against impossible odds' lucky instances, though... 8)

But the resistance table was designed to pit numbers of the same overall range against each other (like STR vs STR, POW vs POW, or STR vs SIZ, etc.); I'm not so sure it would apply as well on something with as wide of a range as percentile skills.

The table used to cap off at 5%/95% to give each side some chance of the unexpected.

You know, we could use this idea with two sides rolling. It would slant things in favor of the greater skiilled character though. We would just have to say that the contest continied until one side or the other missed the roll.

SO with the 125% vs 25%, it would be 95% vs 5% (ths is similar to someone elses idea-I think it was Lord Twig's). Both sides would roll until one side failed (probably the 5%).

simonh said:
GbajiTheDeceiver said:
If the difference is greater than 100, the highest has a flat 95% chance of winning.

Otherwise, subtract a multiple of 100 from each to bring them both back to a 1 to 100 scale, then resolve as normal.

Quick, clean, simple, and mathematically identical to the Resistance Table.

What do you do if one character has say a 90% chance and the other has 110%? Giving the higher skilled guy a 95% chance of success is unfair, and you can't deduct 100 from the guy with 90% either.

The resistance table approach is to add the difference between the skills to 50, and that's the chance of the higher skilled character winning. In this example the target number would be 70%.

Also, the resistance table used a single die for resolution which will give a flat statistical spread. Any system that uses two rolls combined will produce a statistical bell curve in the results.

Simon Hibbs
Good point, I hadn't actually thought of that. Very good point on the bell curve too.

Well, the difference isn't greater than 100, so the 95% rule doesn't apply. You can't subtract 100 either because doing so won't bring both to a 1 to 100 scale. Likewise with leaving them alone.

Hmmmm.

Adding difference to 50 is a bit more fiddly than I'm aiming for here, even if it is the obvious RIGHT way. Subtracting, say 82 from 137, involves just about enough think-time to slow things down.

Hmmmm.

I'm tempted to say "drop the ones digit and use d10" with 5 + active - passive as the target number...

SteveMND said:
Well, exactly what you would think -- the higher-skilled guy wins.

Right. Exactly.

Which is why I said some number of posts ago, in one thread or another, if you have such a disparity in skills, why even roll in the first place?

Or why not roll just to insure the equivalent of a fumble isn't rolled?

Really the simplest approach is to subtract the amount over 100 of the highest skill from both skills and resolve normally. So 110 vs 90 becomes 100 vs 80, 125 vs 50 becomes 100 vs 25. 180 vs 50 becomes 100 vs 5 if you want to keep a (very slight) chance for the defender, or you can rule it an auto win.

Because you never reduce the higher skill below a hundred, and always reduce the lower skill, this never results in lowering the higher skilled characters chances.

Methods that give a bonus for each multiple of 100 (bumps and re-rolls) often break when both skills are high, say 102 vs. 98, and the 102 gets a re-roll or automatic upgrade and the 98 doesn't. In the above method it is resolved as 100 vs 96.

I have said in other posts that a method that keeps the difference between two skills the same doesn't work as well with very high skills (in the high hundreds) - if that doesn't bother you or you don't plan on using skills in the hundreds, the above method works very well.

iamtim said:
SteveMND said:
Well, exactly what you would think -- the higher-skilled guy wins.

Right. Exactly.

Or why not roll just to insure the equivalent of a fumble isn't rolled?

I don't mind this idea. It is just that it won't handle the 150 vs 120 stuff.

atgxtg said:
I don't mind this idea. It is just that it won't handle the 150 vs 120 stuff.

Why not?

If both opponents have skills over 100%, I think -- and I've said it before -- they're both badasses, and probably not going to fail.

Same thing -- just make them both roll, whoever rolls the equivalent of a fumble loses. If they both succeed, whoever succeeded by the most wins.

If you wanted to get really wild and crazy, if either rolled the equivalent of a crit, they automatically win.

Which means that in a contest between a 320% guy and a 150% guy, the 320% guy is going to win every time unless he does the equivalent of a fumble or the 150% guy does the equivalent of a crit.

That, uh, seems to be how it should be.

(I use "equivalent of" because there's no fumble/crit in opposed rolls. )

The rules that I am considering are the following:

1. A success beats a failure, and a critical success beats a success or a failure. Ties go to the defender.

2. The attacker can subtract a quantity from her role, and force the defender to subtract the same quantity from his role. The attacker can not reduce her chance below 50% in this way.

3. Somebody will be assigned the role of attacker and somebody the role of defender, by GM fiat if need be. The attacker is normally the one who makes the roll necessary (the thief trying to sneak past the sentry, for example).

atgxtg said:
The table used to cap off at 5%/95% to give each side some chance of the unexpected.

You know, we could use this idea with two sides rolling. It would slant things in favor of the greater skiilled character though. We would just have to say that the contest continied until one side or the other missed the roll.

SO with the 125% vs 25%, it would be 95% vs 5% (ths is similar to someone elses idea-I think it was Lord Twig's). Both sides would roll until one side failed (probably the 5%).

Not my idea. Mine was the one where you could roll again with a -100 to your skill if you had over a 100% skill and add your results. As far as I know this would give you the best (most consistent) improvement curve over any other method suggested so far. And it would give you a final result without having to roll for a test again.

Utgardloki said:
The rules that I am considering are the following:

1. A success beats a failure, and a critical success beats a success or a failure. Ties go to the defender.

2. The attacker can subtract a quantity from her role, and force the defender to subtract the same quantity from his role. The attacker can not reduce her chance below 50% in this way.

3. Somebody will be assigned the role of attacker and somebody the role of defender, by GM fiat if need be. The attacker is normally the one who makes the roll necessary (the thief trying to sneak past the sentry, for example).

I don't like the ties go to the defender bit (unless you mean actuall D100 results, in which case I'm nor worried about 1%).

The problem with equal success leevels gointo to the defender is that it hurts the attacker too much. A 90% vs a 90% turns into a 9% vs a 90%.

Lord Twig said:
atgxtg said:
The table used to cap off at 5%/95% to give each side some chance of the unexpected.

You know, we could use this idea with two sides rolling. It would slant things in favor of the greater skiilled character though. We would just have to say that the contest continied until one side or the other missed the roll.

SO with the 125% vs 25%, it would be 95% vs 5% (ths is similar to someone elses idea-I think it was Lord Twig's). Both sides would roll until one side failed (probably the 5%).

Not my idea. Mine was the one where you could roll again with a -100 to your skill if you had over a 100% skill and add your results. As far as I know this would give you the best (most consistent) improvement curve over any other method suggested so far. And it would give you a final result without having to roll for a test again.

Sorry. By this point I'm nore sure if I could recognize my solution, and I've came up with four or five!

Replies
18
Views
436
Replies
31
Views
1K
Replies
10
Views
761
Replies
41
Views
2K
Replies
7
Views
461