Friday, May 16, 2014

Munchkin Opposition

What is the point of the Opposed Check mechanic in Warhammer 3rd?  I only recently figured out the answer to that question, and until I did so, I hated the mechanic. Now, I realize it's kinda awesome and absolutely vital to the health of the game.

The point of that rule is certainly _not_ to provide results where both parties' stats have equal weight. Not by a long shot. Blue dice and purple dice just don't have anywhere near mirrored results, and really there is no accurate conversion ratio (because Chaos Stars complicate all the math). Even if they did equate better, the initiating participant in the challenge would still have the advantage of Green, Red, or Yellow dice, none of which are available to the opposing participant in the roll. The defender's stance makes no difference, and the defender's skills are devalued in comparison to the aggressors.

So there must be some other reason for using the opposed mechanic.

Having run a lot of math (see below), I've come to a likely answer to the above question. The answer is interesting, and it's a shame that the designers over at FFG never discussed it in something akin to one of those "Behind the Curtain" sidebars you sometimes see in D&D books. Why does the Opposed Check mechanic exist, and work the way it does? For the singular purpose of preserving character niche for those who didn't take a 2 in Fellowship and Intelligence.  It's main point is to ensure that the wacky non-combat careers in Warhammer (such as Agitator, Courtier or Scribe) actually have a relevant specialty that isn't easily encroached upon by the brute-force combat types just throwing 2 XP at a non-career skill.  Tangential to this, the Opposed Checks ensure that those who optimized their stats actually pay some trade-off for those 5 or 6 blue dice they're getting in their best areas -- there are some things they just can't do at all.

To elaborate: Warhammer 3rd is a game where characters generally begin with a high degree of competence. Starting characters can readily get above a 90% success rate from their best skills and actions. Even if something is well outside your focus, you've still almost always got a better than 50% chance of success for a standard roll.

There's two areas where this high success rate isn't true:
  • one is Opposed checks by characters with a low stat,
  • the other is arbitrarily high difficulty ratings used on plot-relevant checks specifically called out in published adventures. 

The later exist as speed-bumps to prevent PCs from whistling through a scenario due to dumb luck in a way that's not dramatically satisfying. If the PC does manage to short-circuit drama by grasping at straws / thinking outside the box / spamming some random skill at every NPC, the drama that was lost by short-cutting the plot is (hopefully) replaced by the thrill of having beat a high-difficulty (often 4 Purple) check.

The former, those Opposed checks when one side or the other has a 2 in the relevant stat, is there only to make sure that if you put a 2 in your Fellowship or Intelligence you will never succeed at a default roll of that stat. It does this to protect the character niche for those who did invest in Fellowship or Intelligence.  Tangential to this it also makes the person with the 2's occassionally and infrequently get brutally pwned by NPCs with high stats. Said pwnage is just the icing on the cake, it's not the main point of the opposed checks. The game already puts high success rates on the high-stat (N)PCs and if providing PC vulnerability was your only goal you'd solve it with something less like opposed checks and more like the elegant 1-purple of vs Target Defense rolls. There's easier ways to empower GM's manhandling of PCs, but there's not easier ways to protect character niche.

That's probably a good point to segue into all that math I alluded to. After I broke down the success-rate percentages of the official opposed check rule and the most common house-rule to replace it (many GM's use purple dice = opposed stat minus 2), I got a comment from one of my players that perhaps the problem with those numbers was due to the minus two, and the way it zero'd out the difficulties at the low end. If I did 1/2 or 1/3 the stat as purple dice, and any remainder as blacks, maybe that would fit a more reasonable curve. So a 5 in the opposed stat would always produce either 2 purple and a black, or 1 purple and 2 black depending on which variant I used, and a 1 in the opposed stat would always be a single black die no matter how high or low the aggressors stat was. That was actually pretty close to a house-rule I actually used in my first couple one shots, but had eventually dropped because I wanted to try a campaign with a minimum of house-rules to see how the game ran by the books.

Anyhow, I ran a bunch more numbers, and here (after a weird gap that blogger insists on inserting before all tables) are the results:









































Opposed Stats original success rate Purple = Stat -2 Purple = Stat / 2, Remainder = Black Purple = Stat /3, Remainder = Black 1 Purple
2 opposed by 1 44% 75% 58% 58%44%
2 opposed by 2 25% 75% 44% 44%44%
2 opposed by 3 14% 44% 33% 44%44%
2 opposed by 4 8% 25% 25% 33%44%
2 opposed by 5 8% 14% 19% 25%44%
2 opposed by 6 8% 8% 14% 25%44%
2 opposed by 7 8% 4% 10% 19%44%
3 opposed by 1 88% 88% 75% 75%60%
3 opposed by 2 59% 88% 59% 63%60%
3 opposed by 3 38% 59% 49% 59%60%
3 opposed by 4 24% 38% 38% 49%60%
3 opposed by 5 24% 24% 31% 40%60%
3 opposed by 6 14% 14% 24% 39%60%
3 opposed by 7 14% 9% 19% 31%60%
4 opposed by 1 94% 94% 85% 85%72%
4 opposed by 2 72% 94% 72% 76%72%
4 opposed by 3 72% 72% 63% 72%72%
4 opposed by 4 51% 51% 51% 63%72%
4 opposed by 5 37% 37% 43% 53%72%
4 opposed by 6 37% 23% 37% 51%72%
4 opposed by 7 37% 14% 29% 43%72%
5 opposed by 1 97% 97% 92% 92%81%
5 opposed by 2 97% 97% 81% 85%81%
5 opposed by 3 81% 81% 73% 81%81%
5 opposed by 4 81% 63% 63% 73%81%
5 opposed by 5 63% 46% 55% 65%81%
5 opposed by 6 46% 32% 46% 63%81%
5 opposed by 7 46% 21% 39% 55%81%
6 opposed by 1 98% 98% 95% 95%88%
6 opposed by 2 98% 98% 88% 91%88%
6 opposed by 3 88% 88% 82% 88%88%
6 opposed by 4 88% 72% 72% 82%88%
6 opposed by 5 88% 56% 65% 75%88%
6 opposed by 6 72% 42% 56% 73%88%
6 opposed by 7 56% 30% 49% 66%88%


See how every variant I tried* resulted in far more favorable numbers for those with a 2 in the relevant stat? It's a feature, not a bug. It was as if the designers had specifically chosen the official method just to screw with people who had 2's.  It's not monsters they're trying to hamstring, as a quick glance through the NPC stats and PC's Action cards will show you that even the most dim-witted beast has Willpower enough to resist any opposed check the PCs want to throw at them (it's almost always easier to stab a monster than to intimidate or charm it). The only rolls that these opposed mechanics specifically penalize are rolls only PCs ever make: haggling checks, all-or-nothing improvised social rolls, and intution checks to determine NPC motive. PCs with high stats are (as expected) very good at them, but PCs with low stats are terribly handicapped (far more so than a low-strength character swinging a sword).

* = (EDIT: running some more numbers, I just stumbled across a version that probably would work. More about that in a future blogging, as adding it here now would further bloat an already unreasonably-long post.)

Every character in the game needs a decent (at least a 3, and really it should be a 4) Toughness and Willpower, or else they'll spend most of the campaign carrying around a dangerous number of critical wounds and insanities, and slump over at the first sign of Fatigue or Stress. 2's in either of those stats will kill you. You also need either a high Strength, or a high Agility, to power your attacks. Wizards can get away with a high Intelligence + Willpower, and (at least some varieties of) priests can get away with a high Fellowship + Willpower. Between Strength and Agility, whichever of those stats you don't focus on, the other can be dropped down to a 2 with almost no associated downside or penalty. Likewise, each party needs a single character good at Fellowship, and a single character good at Intelligence, and everyone else could effectively dump those stats down to a 2 as well.

The opposed checks mechanic exists only to make that sort of min-maxing dangerous, and to reward the one person per party that decided to invest in the stats everybody else is ignoring. Opposed checks generate unfair results that double-dip on the stats and penalize those with 2's and 3's. Because the game fails to ever discuss the logic behind this, it strikes most readers as a bug. The moment you realize that Hans the shy dockworker can't ever pull a fast one on his coworkers, but Prof. Moriarty can always pass a Guile check vs Sherlock Holmes, you feel like you've just found a shocking loophole that the designers somehow missed. You didn't. That the system works like that was actually the entire point of their design.

There's one thing I'll say about FFG's crew: somebody over there (probably Jay Little) really knows their math. Words aren't always their strong suite, but numbers clearly are. The stance dice, and their complex interactions with the two sides of the action cards, are absolutely brilliant. There is a lot of very complicated math going on behind the scenes all throughout this game system, and so when I see something where the obfuscated math does something very pointed and unusual, I can only assume that it was put there for a reason.

As it turns out, if opposed checks it didn't work like that, there would be no downside to having a 2 in Fellowship. The game has such low difficulties across the board (outside of a low-stat Opposed check) that anyone can try their hand at any non-opposed check. If you're using the default difficulties for non-opposed checks (which is usually around 1 purple), anyone can reasonably do anything. "I absolutely suck at this" means I've only got a 62% chance of success. Unless it's an opposed check, that is.

While the game makes some nods towards Social encounters and progress trackers moved by "Influence the Target", the obvious truth is that most social situations likely to come up in play will be solved by a single roll of a single social skill. GM's might like the idea of building intricate social encounters, but it's a lot of work to do so, and is not an easy thing to improvise and still make it fun. The game is rather light in concrete examples of it in action, and off the top of my head I can't think of a single published adventure that has a really good one at it's core. If there is a good example, it's probably in the adventure from Lure of Power, but that came out very late in the game's development cycle.

This is Warhammer. The game has several-decades of reputation as the game where your party includes the dregs of the Empire. The party is more likely to be a ratcatcher, a blacksmith and a coachman than it is to be a knight, a wizard and a healer. That's a vital part of the setting and product identity, and FFG wasn't about to eliminate it entirely (even though they were interesting in making the game more cinematic and heroic than the previous editions). Part of the fun is being the unlikeliest of heroes. The spoiled nobleman fop, the illiterate peasant, or the malpracticing barber-surgeon that is thrust into the deep end and has to make do. Someone was going to draw those careers, and feel compelled to invest in non-combat stats just to be true to the character.

It's not fun for those players if everyone else (especially those lucky few who drew legitimate warrior careers) can sink those same points into killing things faster, and then still get by when forced to roll their one non-combat check of the session (because the difficulties are always low and success means the puzzle or social encounter is over). To make those colorful non-combatant careers worth playing at all, it was necessary to make 2s in non-combat skills suck.

The lengths the developers went to in the attempt to make these wackier careers worth playing is actually kinda cool and commendable. A vague framework of a social encounter system that takes a lot of GM prep time? That's to give the nobleman or agitator a chance to shine, and if you don't have one at the table the GM can skip it in favor of a single die roll. Obnoxious rarity and haggling system that relies on too many opposed checks? That's so merchants and conmen have an area where they feel needed, and if you lack that sort of character you can handwave it. Byzantine healing rules that involve a million die rolls? That's so barber-surgeons and physicians can be vital to the party, but if you don't have one, a liberal dose of GM fiat will whisk it all away.  All the clunkiest, most annoying parts of the rules exist just to make players of certain careers feel good. The only problem with all this is that FFG is so tight-lipped about design philosophy, they never actually say "skip this if you don't have a character at the table who cares about it."

Maybe I'm wrong and they don't intend you to skip them -- it could be that they expect high PC mortality rates to make all those careers relevant eventually -- but we'll never know because FFG never explains their motives or decisions.



Arcane Afterthought: Wizards and Priests sort of break this dynamic I was just raving about. They both use a non-combat stat as a combat stat, and so get a large number of spotlight moments both in and out of battle from that single investment. The game tries to compensate for this by making 1st rank spells and blessings weaker than default actions, so they don't come into power until late in the campaign. The game also makes them vulnerable to a bad channel/curry roll, so they have to invest in a second stat (Willpower) to fuel their effects. Intelligence has the most skills, so Wizards in particular are also subject to nasty miscasts (there are also thematic reasons for this, but that's beside the point) and minor restrictions on armor. Priests avoid those downsides, but require extra XP investment to unlock their "magical" abilities in the first place, so they trail behind wizards in the early game (at a time when wizards themselves trail briefly behind others in combat at least). None of which really manages to bring wizards or priests back in line with the other characters in a long campaign -- any Wizard (and any Priest of Rank 2 or higher) has tons more options and spotlight moments than a similar-rank scholar or merchant.

Honestly, the biggest thing keeping this in "balance" there is the low frequency of these character types. To be a wizard or priest, you have to get that one particular career card as one of your three random draws from the pile of 50 basic careers (and not have anything of greater personal interest on the other two cards). If none of the players get a lucky draw on day 1, it's a non-issue. If you have every expansion and follow the character creation rules as written, there's a less than 25% chance of even having an apprentice wizard in a 4-player campaign.

Obviously this mitigating factor goes out the window if the GM lets players pick any starting career instead of drawing randomly, but even then it's a least partially mitigated by the lesser chance of anyone playing the random non-heroic careers who get eclipsed early on by Priests and Wizards. Even in these circumstances, a min-maxed caster will eventually grow to be better than everyone else at nearly everything, but you'll get 20-30 sessions of balanced play before that happens.

So again, perhaps the designers just figured the game's high lethality and grimdark grittiness would prevent characters ever living long enough for that to be an issue.

No comments: