d% roll under system and combat.

[Psychic Robot]

Needing some ideas here. Basically, trying to make the WFRP system workable because I like it a lot but, as many of you may have noted, the system is awful. My first order of business is to work with the combat system.

Here's how combat in WFRP works:

1. You have two stats, Weapon Skill and Ballistics Skill (WS and BS, respectively). Whenever you are attacking, you make an d% roll. If you roll under your WS/BS, your attack hits.

2. Your opponent can then attempt to negate your attack. He may try to dodge it (by rolling under his Dodge Blow skill) or by parrying it (by rolling under his WS). Someone may dodge once per round, and he may parry once per round. Shields grant a free parry each round, as does two-weapon fighting (normally, you have to enter a parrying stance to avoid blows, which is pretty much bullshit).

3. If you hit, roll damage, which is 1d10 + your various modifiers. Your opponent reduces that damage by an amount set by his armor and his Toughness stat.

3. Once someone is reduced to zero or fewer wounds, he begins to take critical hits with all attacks that do at least one damage. This is where the system gets really bad, as it involves rolling on a chart in a horrible, time-consuming process that slows the game to a crawl. An opponent cannot be knocked out; he can only be killed due to critical injury.

Combat is really, really bad. The way that it works is bullshit. The RNG is completely fucked. Not only do players have a very low percent chance of hitting someone (usually about 50% to start), but that person usually has a 40% or greater chance of completely negating said hit. This means that combat is only effective when the party engages in dogpile tactics to gain large WS boosts and exhaust the enemy's dodges and parries for the round. Once that enemy is down, they all charge the next. (The most effective use of tactics in my WFRP game? The bounty hunter throws a net on someone and everyone starts stabbing him.)

However, once the enemy is reduced to 0 wounds, he can continue to fight on for several rounds because of the critical hit system. If you don't hit him hard enough to kill him, he can stay fighting with little to no penalty.

On top of this, the system discourages any kind of movement. At early levels (well, careers), players are clamoring for every bonus they can get so they can actually hit the enemy, so they spend a full round action to get a +10% bonus on their attack. At later levels/careers, players use their full attack action and trade dodges and parries.

The system, overall, is just balls. I'm trying to think of a way to fix it.

The easiest solution, in my opinion, is invert the d% system so that it works more like the d20 system: roll a d%, add your skill, and see if it hits the appropriate DC. However, this leaves a bad taste in my mouth, as it seems a betrayal of the system. I desire to leave the system with the roll under d% rather than reconverting everything.

This leaves me with a dilemma on how to do to-hit rolls, which will form the foundations of the system. My first order of business is eliminating the "you can only dodge one attack per round," as that is retarded. Combining the ability to dodge into the attack roll (as with D&D) seems to be a good way to deal with this issue. The first seems easy enough: you roll your WS against your opponent's Dodge Blow skill for every attack, and see whoever rolls better (whoever has the most "degrees of success," which is basically your d% roll - your skill and then divided by ten). I fear that this will lead to too much rolling, however, so I am considering the following:

1. Each character has an avoidance rating. This is equal to his Dodge Blow skill - 50% (the average on a d% roll). This number can be negative or positive. This represents his ability to, well, avoid attacks.

2. Each character making an attack subtracts his target's avoidance from his WS/BS when making an attack roll. This new number is the WS/BS that the attacker must roll under to hit the target.

3. Parrying and whatnot will be covered as an alternative form of combat, perhaps with that whole "openings" system I mused upon.

My concern with this system is that it's a little more complicated than just opposed rolling, but I think it might work when players get used to it.

Thoughts? I'll obviously have to wait on working on the rest of the system until I get this combat business squared away.

[DSMatticus]

I always hated roll under for the exact problems you're facing: it's just so mathematically inflexible. You can get the exact probabilities using a roll over, addition, and a global DC for every check, and it makes it a hundred times easier to do a straight opposed roll.

It's like they asked, "what can we do to get the exact same sorts of probability as a d20 + mod vs DC roll, but make it so it can't handle arbitrarily sized inputs?" Not to mention, the idea of having 'opposed contests' like to-hit and dodge be separate rolls against static DC's as opposed to eachother means as combatants get better, the frequency of hits tends rapidly towards "once in a blue moon" because everybody rocks their dodge checks.

You already discarded it, but for it's awesomeness I have to recommend roll + modifier vs DC 100 (101, actually, but let's keep it simple, 1% doesn't matter) or the opponent's result. Your probabilities work out basically the same for single rolls, and opposed rolls should come out to 50/50 hit and miss between equally skilled opponents, instead of less than 1/4th hits. Having the floating number "100" would probably make players angry, though. "Why is everything the same difficulty? Makes no sense." Just for cognitive reasons, you'd probably want to add situational bonuses that typically effect the skill to the DC instead, but... that's getting complex.

Blame the designers for starting with a bad framework.

But if you don't want to do that, and it seems like you don't, you're getting close to the next best thing. A few options:

1) You said it, add 50 to the WS, then subtract the target's dodge. This zeroes you in on a 50 WS vs 50 dodge at 50% chance to hit. Every point in WS or dodge shifts this exactly 1 percentage point. If WS is typically higher than dodge, as in your example, hits will happen most of the time.

2) If you wanted to be closer to the original system, where hits happen once in a blue moon, you could just subtract 1/2th the dodge skill from the attacker's WS. This zeroes you in on a 50 WS vs 50 dodge at 25% chance to hit. It also effectively increases the cost of dodge, it uses multiplication/division so it scales horribly (but this is a roll under skill system, so the RNG is already set up to scale horribly), and all around sucks. I don't really know why I even brought this one up, except perhaps as a suggestion not do do this.

3)

which is basically your d% roll - your skill and then divided by ten)

It's too much math, really, not rolls. Subtraction and division? Cut that shit right out, obviously. The rolling isn't a big deal - you already had opposed rolls for combat in the first place (albeit only one dodge per round). If you have to find degrees of success, make that a roll over system. Again, main clue here is to definitely not do this.

I have to stand by either converting to a roll over with a static global DC (roll + skill +/- mods vs 100 or opponent's check), OR using what you suggested if you must have roll under, option #1.

[CCarter]

Another option might be to convert the d100 to a die pool.

For a d6 pool equal to the 10s place , TN 5, 2 hits required for success, the d% probability isn't too different to what you'd get on the d100...at least, its within 10% between about 30% and 90%, though it falls apart at the far ends (particularly lower end).

2d- 11%
3d- 26%
4d- 41%
5d- 54%
6d- 65%
7d- 74%
8d- 80%
9d- 86%
10d-90%

First worked this out the long way but there's a calculator here I used to check:
http://www.pvv.org/~bcd/SR/dicerollcalc.html

[Username17]

The easiest solution, in my opinion, is invert the d% system so that it works more like the d20 system: roll a d%, add your skill, and see if it hits the appropriate DC. However, this leaves a bad taste in my mouth, as it seems a betrayal of the system. I desire to leave the system with the roll under d% rather than reconverting everything.

OK, now consider the exact same argument in light of THAC0 in the move from 2e to 3e. Even if you go with the fixed Target Numbers model, it is fucking daft to use a "roll lower than skill" as your fixed Target Number when you can get the same percentages and a lot more functionality by adding skill to roll and having the Target Number always be 100. Like, degree of success then becomes "the last two digits" on all successful rolls, allowing you do opposes rolls with two less subtraction steps.

% Roll Under is a dinosaur. It's one of those systems that was "obvious" back in the 1980s, but it is in all ways inferior to rolling and adding against a Target Number. All ways.

-Username17

[mean_liar]

RollOver doesn't cleanly model "X% of all hits are crits" with a single roll. A RollUnder can do it cleanly by having it be 10% of your target number (ie, an 87% chance to hit also is a crit result on 01-08).

RollOver generally has to do crits as a "results X% over the target number are crits", so that all results of d100+Skill-Defense (or whatever) > (100 + X) are crits.

[Ferret]

If you're looking for 10% of your rolls to be crits, it's easy enough to just say that any roll where the single digit on the percentile dice is a 0 is a critical (so you crit on 10, 20, 30, etc) if the roll would otherwise be successful.

And you'd probably want to do something special for 00/0 rolls, because hey, everybody loves rolling double-aught.

But again, that mandates a RollOver system, which PR is trying to avoid.

PR, given your requirements, I think you're on the right track. Convert dodge / parry to a static defense score (how you do it is arbitrary, really) and let the attacker roll under his WS/BS and add the targets defense as a penalty.

[RobbyPants]

RollOver doesn't cleanly model "X% of all hits are crits" with a single roll. A RollUnder can do it cleanly by having it be 10% of your target number (ie, an 87% chance to hit also is a crit result on 01-08).

RollOver generally has to do crits as a "results X% over the target number are crits", so that all results of d100+Skill-Defense (or whatever) > (100 + X) are crits.

Couldn't you just roll a d10 along side of the attack roll, and if the attack hits, check the d10 to see if it's a 10? That'd make 10% of your hits crits without having to do additional math to check if the attack roll is above X, wouldn't it?

[mean_liar]

I assumed there were ways to do it, I just didn't put much thought into it. I only mentioned the X% as crits deal because it shows up in an old-school game I like using roll-under... as 5% and 15%.

I don't think that's a good system - it was a pain in the ass - but it would've been worse in roll-over.

[Username17]

I assumed there were ways to do it, I just didn't put much thought into it. I only mentioned the X% as crits deal because it shows up in an old-school game I like using roll-under... as 5% and 15%.

I don't think that's a good system - it was a pain in the ass - but it would've been worse in roll-over.

Nah, in Roll-over a Crit is just whenever the 1s place is a magic number. The magic number is set by how you want to round the 10%. If you would round it down, the magic number is 9. If you would round it up, the magic number is 0. If you would round it off, the magic number is 5. And so on. It's actually much easier even than dropping the ones place to determine crit range.

Exra bonuses for the fact that you can actually color in the magic critical number on the ones-place die, thereby calling attention to a potential crit every time.

-Username17

[CCarter]

Another way to do the x% chance of hits are crits is with a second confirmation roll. You can even do it with an additive d20: a 19-20 threat range followed by a second successful attack roll (e.g. 3E D&D longsword critical) equals 10% of hits being crits.

[Blasted]

I would have thought that the obvious method is to base the crit on the target number. i.e. rolling exactly X (5%) or X+1 (10%) etc.

[Username17]

Another way to do the x% chance of hits are crits is with a second confirmation roll. You can even do it with an additive d20: a 19-20 threat range followed by a second successful attack roll (e.g. 3E D&D longsword critical) equals 10% of hits being crits.

The Confirmation Roll is the best and really only way to get variable crit chances going. By varying the threat range you could have a 15% crit rate or an 11% crit rate or whatever. d% has the specific property that it has a ones place, meaning that it is easy to generate an approximately 10, 20, or 30% crit rate without a second roll by having one or more crit numbers pre-designated for the ones-place die.

-Username17

[hogarth]

Like, degree of success then becomes "the last two digits" on all successful rolls, allowing you do opposes rolls with two less subtraction steps.

But note that you can have a "roll under" system that doesn't require calculating "degrees of success" for opposed skill checks; we discussed such a system (the latest version of Runequest? I forget) in another thread.

In that example, you didn't want to roll low, you wanted to roll as close as possible to your skill rank without going over (a la blackjack or "The Price Is Right"). So a successful roll of 60 beats a successful roll of 30 in an opposed skill check, for instance.

[Username17]

Like, degree of success then becomes "the last two digits" on all successful rolls, allowing you do opposes rolls with two less subtraction steps.

But note that you can have a "roll under" system that doesn't require calculating "degrees of success" for opposed skill checks; we discussed such a system (the latest version of Runequest? I forget) in another thread.

In that example, you didn't want to roll low, you wanted to roll as close as possible to your skill rank without going over (a la blackjack or "The Price Is Right"). So a successful roll of 60 beats a successful roll of 30 in an opposed skill check, for instance.

That "works" but only in the way that THAC0 "works". It gives the same results as "roll high, TN 100" except that you have to report two numbers instead of only one.

-Username17

[Orion]

I can't tell, but I *think* Hogarth doesn't actually intend you to do any subtraction. The proposed order of operations for an opposed roll is:

1You both roll d% and compared to your skill
2 Did you succeed? If no, you didn't win. If yes, speak aloud the number showing on your dice
3 If the opponent didn't succeed, you win now. If he did, compare numbers
4 Higher number wins.

[echoVanguard]

it is fucking daft to use a "roll lower than skill" as your fixed Target Number when you can get the same percentages and a lot more functionality by adding skill to roll and having the Target Number always be 100. Like, degree of success then becomes "the last two digits" on all successful rolls, allowing you do opposes rolls with two less subtraction steps.

% Roll Under is a dinosaur. It's one of those systems that was "obvious" back in the 1980s, but it is in all ways inferior to rolling and adding against a Target Number. All ways.

-Username17

I'm afraid I'm having some trouble understanding the principles your argument is based on, Frank.

Let us assume that a character has a value in a skill of, say, 62, and succeeds on their roll by a margin of 54. In a DC 100 system, this requires a roll of 92, but in a roll-under system, it requires a roll of 8. Both systems work identically, except that you want to roll high in a DC 100 system, and low in a roll-under.

Here's the question - how is 62+92=154 (and look at the last 2 numbers) intrinsically simpler than 62-8=54? One could even make the argument that the first equation is actually 62+92=154-100=54, which is obviously *more* complicated.

Furthermore, opposed rolls work the same way in either system - in a roll-under, if a character with skill 22 rolls a 15, while another character with skill 66 rolls a 55. Character 1's margin of success is 7, while Character 2's margin of success is 11. 11 is obviously greater than 7, so no additional steps are needed.

In DC 100, character 1 rolled an 85, and character 2 rolled a 45. Adding 85 to 22 is 107, so margin of success is 7; adding 45 to 66 is 111, so margin of success is 11. Again, 11 is obviously greater than 7, but the number of steps is the same (assuming you perform result-100 intrinsically, which is a pretty brave assumption).

If you wouldn't mind, I'd really appreciate it if you could elucidate what additional benefits DC 100 has over roll-under that I might have overlooked. Does your argument assume that addition is intrinsically easier than subtraction?

echo

[echoVanguard]

But note that you can have a "roll under" system that doesn't require calculating "degrees of success" for opposed skill checks; we discussed such a system (the latest version of Runequest? I forget) in another thread.

In that example, you didn't want to roll low, you wanted to roll as close as possible to your skill rank without going over (a la blackjack or "The Price Is Right"). So a successful roll of 60 beats a successful roll of 30 in an opposed skill check, for instance.

That seems pretty fun. How did it go over with your group, Hogarth?

echo

[hogarth]

That "works" but only in the way that THAC0 "works". It gives the same results as "roll high, TN 100" except that you have to report two numbers instead of only one.

-Username17

Frank, how do you propose to do an opposed skill check with only one roll? Does one side "take 10" essentially?

That seems pretty fun. How did it go over with your group, Hogarth?

echo

I've never used it (although I've played various Chaosium "roll-under" systems in the past). It came up as we were discussing Mongoose's RuneQuest v2 in this thread.

[echoVanguard]

The description Orion gives seems to be the only completely opposed dice-based resolution mechanic with no non-binary outcomes (i.e. no actual math done other than evaluation of 1-2 inequalities) that I've yet heard about.

It has interesting potential in other ways, too, since you could tie the quality of the effort to the number rolled in a fairly linear fashion - in other words, a roll if 67 is quantifiably 67% of the maximum possible quality achievable at that check by anyone, and thus intrinsically only achievable by persons with a 68 or higher in the skill. That might actually be both simpler and greater in depth than either of the proposed alternatives.

I'd be very interested in any potential counterarguments, opposing viewpoints, or overlooked potential problems with this resolution mechanic.

echo

[ScottS]

I might have missed a few system-related bitchfests in the past few months, but another recent roll-under system was Eclipse Phase, and I'm pretty sure this same conversation came up in that thread (i.e. the fact that in addition to roll-under somewhat sucking, the opposed-test mechanic is "roll high as you can under your skill", which is in direct opposition to needing to roll low in order to get higher degree of success etc.)

[echoVanguard]

Assuming http://www.tgdmb.com/viewtopic.php?t=50 ... sc&start=0 is the thread in question, I'm reading it now, and thank you for pointing it out to me. However, I'm a little skeptical that "subverting the lower-is-better goal" is a meaningful problem, especially considering how many other problems it seems to solve.

echo

[Username17]

Assuming http://www.tgdmb.com/viewtopic.php?t=50 ... sc&start=0 is the thread in question, I'm reading it now, and thank you for pointing it out to me. However, I'm a little skeptical that "subverting the lower-is-better goal" is a meaningful problem, especially considering how many other problems it seems to solve.

echo

It solves absolutely nothing compared to just counting up instead of fucking around with counting down. There is no advantage to fiddling with roll under. None. It is in no way better than having a fixed TN and counting up.

Let's say for a moment that you rolled really well and you forgot a modifier and wanted to go back and add it in. How is it easier to have your "good rolls" be the ones that are razor's edge from not having succeeded? What the fuck?

-Username17

[echoVanguard]

It solves absolutely nothing compared to just counting up instead of fucking around with counting down. There is no advantage to fiddling with roll under. None. It is in no way better than having a fixed TN and counting up.

Let's say for a moment that you rolled really well and you forgot a modifier and wanted to go back and add it in. How is it easier to have your "good rolls" be the ones that are razor's edge from not having succeeded? What the fuck?

-Username17

Which method are you referring to, Frank? The original method of roll-under and difference is margin of success, or Hogarth's percentile-chicken method?

Also, you're not actually responding to any of the arguments folks brought up regarding the different number of steps in calculating for each of the different methods. I sincerely hope I'm misreading your statement, because at first glance, it seems to be that your argument is "It's the same, it's a waste of time, and it doesn't improve anything" without actually presenting evidence or arguments to support any of those assertions.

echo

[CCarter]

It has interesting potential in other ways, too, since you could tie the quality of the effort to the number rolled in a fairly linear fashion - in other words, a roll if 67 is quantifiably 67% of the maximum possible quality achievable at that check by anyone, and thus intrinsically only achievable by persons with a 68 or higher in the skill.
echo

Its not really much different to having an additive system capping bonuses at +100%, and having the total be the maximum possible quality achievable on a linear scale out of 200. However, with this method, you can no longer easily compare 'fail' results and 'success' results on the same linear scale, since the 'fail' results are the scores that are higher.

[Psychic Robot]

To note: I am aware that it is easier to use a straight d20 target number system and that there is little reason not to use a roll over system in place of a roll under system. However, like I said, I believe that the roll under system is part of the game and to abolish it would be a travesty, akin to playing D&D with 3d6 instead of a d20.