Jason's Rant on Skill Costs

From Skillful Roleplaying Game
Jump to: navigation, search

The tiered skill system provides an excellent way of representing characters' specialized and general competencies, but it doesn't do a very good job of encouraging people to choose skills based on their character's "flavor" or narrative. On the contrary, the current skill system overwhelmingly incentivizes characters to invest almost exclusively in first-tier skills, even if the character's "feel" is meant to be that of a narrow specialist. Except in the corner case where the character literally only wants to learn a single skill, he or she would be better off investing almost all of his or her XP in generalist skills.

Suppose I want to learn both Electrical Engineering and Mechanical Engineering. I have 300 XP to spend, so I buy 5 ranks in each: (10 + 20 + 30 + 40 + 50) * 2 = 300.

A strictly better way of doing this would be to simply buy 5 ranks in general Engineering: (20 + 40 + 60 + 80 + 100) = 300 XP. This gives me the same +5 bonus to both Mechanical and Electrical Engineering, in addition to a +5 bonus to any other kind of Engineering I might want to do, like Programming or Chemical Engineering.

Similar logic applies to literally any pair of two skills that are both covered by any higher-level skill, such as Pharmacy and Diseases, or Throw/Catch and Zero Gravity. Moving up the tier ladder only makes things worse. For example, suppose I want to learn both Agility and Speed. 5 ranks in each would cost 600 XP: (20 + 40 + 60 + 80 + 100) * 2 = 600. Instead, I could simply buy 5 ranks in Body, for (30 + 60 + 90 + 120 + 150) = 450 XP. I would thereby get the +5 bonus to Agility and Speed at a steep discount, and *also* get a +5 bonus in Power and Fortitude.

Similar logic also applies to any triplet of third-tier skills that are all covered by the same first-tier skill. For example, suppose I wish to learn Soft Melee, Dodging, and Point Shooting, all of which fall under Combat. There is no sense in actually investing in any of the third-level skills, because I could just take 5 ranks in Combat for the same price: (30 + 60 + 90 + 120 + 150) = (10 + 20 + 30 + 40 + 50) * 3. I would then get my +5 bonus to my preferred skills, as well as to any other combat skills I ever cared to use. Of course, if I wish to learn more than three of the sixteen or so third-tier skills that fall under a typical first-tier skill, the reasons for only investing in first-tier skills will be that much more apparent.

This is a major problem because it penalizes players who think carefully about their character's ethos. If Patrick is sort of an amateur urban ninja who dodges out of the way of gunfire, uses aikido-style moves to remove relatively innocent obstacles from the line of fire, and then pops the real bad guy between the eyes, he should be able to represent that by specializing in Soft Melee, Dodging, and Point Shooting. He shouldn't behave exactly the same (in terms of skill checks) as Bob, who specializes in Hard Melee, Covering, and Indirect Fire because he's a middle-class trucker who's used to brawling in bars but prefers to stay out of direct gunfights in favor of shooting railguns from the next corner over. Yet both Patrick and Bob have strong incentives under the current system to just take 5 ranks in Combat, leaving them mechanically indistinguishable from each other.

It could be objected that characteristics can give players a relative advantage in their ability to buy a third-tier skill if they so choose, but even a double major aptitude only provides a multiplier of 1.2, and it is not clear that double majors in third-tier skills are so very much cheaper than double majors in first-tier skills as to help balance the scales in favor of specialists. Most likely they are simply proportionately cheaper, and thus have little or no effect on the specialist vs. generalist balance.

It could be objected that the above examples rely on all-or-nothing shifts of skill from one tier to another tier, and that in practice players tend to put at least some ranks in each tier. Even with a relatively balanced skill portfolio, though, the point remains that whoever tilts their skill set toward the general skills will tend to do much better.

Suppose we give Bob and Patrick 1000 XP each to set up their combat skills. Bob is a meta-gamer, so he puts 6 ranks in Combat (30 + 60 + 90 + 120 + 150 + 180 = 630 XP), 2 ranks each in Melee, Defense, and Artillery (20 + 40) * 3 = 180 XP, and 3 ranks each in Hard Melee, Covering, and Indirect Fire (10 + 20 + 30) * 3 = 180 XP. All together, 630 + 180 + 180 = 990 XP. Bob will get a bonus of +11 to his chosen specialist skills, as well as a bonus of +8 to 'nearby' skills and a minimum bonus of +6 to any vaguely combat-related skill.

Patrick is placing his skills 'artistically', to try to capture his character as accurately as possible, so he puts 2 ranks in Combat (30 + 60) = 90 XP, 3 ranks each in Melee, Defense, and Guns (20 + 40 + 60) * 3 = 360 XP, and 6 ranks each in Soft Melee, Dodging, and Point Shooting (10 + 20 + 30 + 40 + 50 + 60) * 3 = 630 XP. Patrick will get the same bonus of +11 to his chosen specialist skills, but will only get a bonus of +5 to nearby skills, and a piddly bonus of +2 to general combat skills.

Patrick's bonuses are more realistic, in the sense that they more accurately reflect the way that training intensely for a narrow specialty provides only mild bonuses to related skills. However, once players realize how much weaker Patrick is than Bob, they will start to meta-game. The rules will punish players who strive to be realistic, and reward players who exploit the system. This isn't the end of the world, but it strongly suggests that we change the system, because half the point of the rules is to reward players for behaving realistically so that they don't have to exercise willpower and self-discipline just to keep the simulation running smoothly.

A minor tweak like increasing the cost basis for a first-tier skill from 30 to 40 or even 50 is unlikely to suffice, because the geometric increase in the cost of any given skill will force players to put something close to a third of their XP into the top-tier skills no matter how expensive the base cost for first-tier skills is, and then as long as the players wish to learn more than 2 or 3 skills within that first-tier skill, the first-tier skill will start to look incredibly attractive.

Suppose, for example, that we make the base costs 10, 25, 60 instead of 10, 20, 30. We kick the total XP up a notch to compensate for the higher average costs, so players get 1300 XP for their combat. How would Bob and Patrick spend their 1300 XP? Patrick stubbornly continues to put the bulk of his points into third-tier skills. He sinks 1 rank into Combat (60 XP), 2 ranks each into Guns, Melee, and Defense (25 + 50) * 2 = 150 XP, and 8 ranks each into Point Shooting, Soft Melee, and Dodging (10 + 20 + 30 + 40 + 50 + 60 + 70 + 80) * 3 = 1080 XP. 1080 + 150 + 60 = 1290 XP. He gets a +11 bonus to his specialties, but a weak +3 to nearby areas and a pathetic +1 to general combat.

Bob, meanwhile, shrewdly splits his XP evenly, putting 3 ranks into Combat (60 + 120 + 180) = 360 XP, 3 ranks each into Artillery, Melee and Defense (25 + 50 + 75) * 3 = 450 XP, and 5 ranks each into Indirect Fire, Hard Melee, and Covering (10 + 20 + 30 + 40 + 50) * 3 = 450 XP. 450 + 450 + 360 = 1260 XP. Bob gets the same +11 bonus to specialties as Patrick does, but Bob enjoys a substantial +6 bonus to nearby skills, as well as a somewhat useful +3 to general combat. Despite spending fewer XP, Bob gets a strictly better result.

By "strictly better," I mean that there is no conceivable circumstance in which Patrick could have a statistical advantage over Bob in combat. The problem is not just that, on average, Bob does better. The problem is that Bob does *just as good* as Patrick does *even* in the areas in which Patrick has supposedly specialized, whereas Bob does better everywhere else. Compared to Bob, Patrick is thus not really a specialist -- he is simply someone who is ignorant about all areas outside his favorite skills. It is unlikely that there is any base cost we could set for first and second-tier skills that would fix this problem short of simply making first and second tier skills ridiculously unaffordable. This would over-correct the problem and force everyone to play like Patrick, whereas the real goal is to allow players to choose a generalist or specialist outlook based on narrative concerns without suffering unwarranted penalties. It's fine if 'radical' characters who invest *only* in first-tier or third-tier skills to the exclusion of all else are penalized, but a skill set that's _tilted_ toward first-tier skills should be roughly as powerful as a skill set that's _tilted_ toward third-tier skills.

How can we accomplish this?

One option is to dump the geometric skill progression, at least in the early stages, in favor of adding a 'penalty' to skill costs in excess of a relevant player statistic. For example, if a player wants to take a 12th rank in Electronic Engineering, but only has an Intelligence score of 8, the 12th rank might cost +40 compared to its base cost. I'm not sure I like the idea of making player statistics a necessary part of skill development.

Another option is to halve the skill 'bonus' that a higher-tier skill confers on skill checks for skills underneath it. For example, if Bob has 6 ranks in Combat, 2 ranks in Melee, and 3 ranks in Hard Melee, then his total skill ranks in Hard Melee would effectively be (6/4) + (2/2) + (3/1) = 1.5 + 1 + 3 = 5.5 ranks, which we might round up to 6 ranks. By contrast, if Patrick had 2 ranks in Combat, 3 ranks in Melee, and 6 ranks in Soft Melee, then his total skill ranks in Soft Melee would be (2/4) + (3/2) + (6/1) = 0.5 + 1.5 + 6 = 8 ranks. Thus, in unarmed combat, where both players have a chance to work within their specialties (Hard Melee & Soft Melee) Patrick has a narrow edge of +2 over Bob when both of them are working within their specialties.

On the other hand, if we give Patrick and Bob an unfamiliar weapon like crossbows, e.g., then the generalist Bob will do better. With 6 ranks in Combat, he gets (6/4) = 1.5 ranks, which we might round up to 2 ranks in Archery. Patrick, on the other hand, only has (2/4) ranks to work with, which might round up to 1 rank in Archery. Bob will thus have a narrow edge of +1 over Patrick whenever both of them are working outside their specialities.

This provides the appropriate cost-benefit tradeoff for people who are considering becoming specialists: if they specialize, they will tend to beat equivalently experienced generalists when they have a chance to use their specialty, but they will tend to lose to equivalently experienced generalists when their specialty is inapplicable.

Personal tools