GH2: Character System Changes

[Joseph Hewitt]

Previously I posted my ideas for revising the damage system. Here's an idea for revising the skill/renown system. As always, I'd like some feedback before I go doing anything stupid.

The number of skills will be cut drastically. Each skill will correspond to a specialty or career path, such as "Mecha Combat" for all mecha fighting skills. Different uses of a particular skill will use different stats- Reflexes for attacking, Speed for defense, etc. Characters may take all skills without penalty. Each skill will have its own individual Renown value; these separate renown values will be used for generating skill-specific plots.

Prospective skill list, subject to change:

• Mecha Combat
• Personal Combat
• Subterfuge: Thief skills (Stealth, CodeBreaking, PickPockets)
• Scouting (Awareness, Survival)
• Interaction: Covers all the speaking skills (Conversation, Intimidation, Taunt)
• Performance: Singing (Charm), making art (Craft). Maybe change name to Art? Humanities?
• Insight: The detective skill, maybe change name to Detection? Investigation?
• Medicine
• Technology: Covers repair and science
• Arcana: Mysticism, secret knowledge and lost technology
• Regimen: Healthy living. Vitality, Athletics, Concentration, and Toughness

There should be plots/story branches based on different skills. The non-combat skills should allow access to different branches rather than simply making the existing branches easier- these different branches should lead to different outcomes. I'll have more to say about that when I get the revamped story system fully written up.

Talents will be expanded to cover many of the special abilities currently handled by skills. Mecha Engineering would be either a talent or series of talents- no more rolls for part destruction, you can either do something or you can't.

Personal history should be more important in character creation. Right now I'm thinking about four steps for the lifepath- social class/birth, education/job, personal hardship, and reason for becoming a cavalier. In addition to setting skills and personality type these lifepath steps could also set special traits that will follow the character for life- key skills (cheaper to advance), interaction bonuses or penalties for certain NPC types (based on job, personality, or faction), mutations and so on. There should, of course, still be an advanced mode to let players pick their exact traits a la carte.

Advantages over the current system: Stat distribution would be more important. The line between skills and talents would be more distinct. Jack of all trade characters would again be possible, but would not have a big advantage over specialist characters. By having separate renown scores for separate skills a character will not be offered unwinnable personal combat missions just because they've advanced far in mecha combat. A smaller number of skills means that it'll be easier to create content catering to each of those skills.

[Lord Afabie]

I like it, but I do have a couple of suggestions.

With the new skills covering multiple old skills they will be used more frequently, and thus improve faster, unless the experience curve is re-balanced.  And if you are already changing the amount of experience you get, I would like to suggest a change in the way experience is calculated.

In one of my vapor-ware projects I had a system where the amount of experience awarded on skill use was based on how close the skill roll was to the minimum roll required for success.  I abandoned the game before I was able to test this idea in play, but I think it has several advantages.  It makes challenging, but not suicidal, fights give the most experience, thus reducing the inclination towards both grinding and power diving so many RPG’s suffer from.  It also seems more realistic to me since you learn fastest when pushed just slightly outside your comfort zone (so perhaps failure should give a bit more experience that success).

Another concern of mine is that reducing the number of skills might reduce the diversity of characters.  This problem can be reduced by making it impractical to become proficient in all skills, but then you run the risk of recreating the problem in the current skill system, where so much of your experience needs to be dumped into ‘essential skills’ everyone will take that you can’t spare enough for adequate customization.

My solution would be an increased reliance on talents to provide that customization.  This would mean giving each character more talents, and providing a larger list of talents arranged in ‘talent trees’.  The initially available talents should be fairly weak, but they would open up the path to more powerful talents that could make two characters with the same stats and skills play very differently.  Each talent tree should have multiple branches (some of which could merge back together) and be large enough that a single character can’t fill one entirely, but not so large that he can’t gain powerful talents from at least two trees.

These suggestions (particularly the second) could have a major effect on how the game feels, and so might not fit with your vision.  I understand that maintaining that vision means you will have to reject even some brilliant ideas, therefore I won’t be upset if you fail to endorse these. 

[Rowanthepreacher]

I wouldn't take it that far, nor do I think that talents are the best way to go.

Gearhead 2 has a suitable number of traits and skills; rarely will I have every trait or skill that I want, but usually I've got enough that the play experience isn't lacking. Reducing the number of skills would abolish the fine balance that players must strike.

I could get behind the "do or do not" functioning of skills though. A lot of skills already operate that way and the only one that doesn't is the biggest pain in the arse to fail (Engineering).

Something I'd like to add in early: Renown needs to go down slower when you fail. I've just quit playing GH2 again because I accidentally committed a couple of crimes and now, whenever I back out of a fight with the police, I lose loads of renown and have to do a dozen missions to raise my mission pay back to normal.

I'm not sure how well "xp based on near-failure" would work though. I can see that it would be relatively easy to pick a fight with a very nimble but weak opponent (Dora, anyone?) and just fire weapons with -2 accuracy penalties and the like. Then, you'd be likely to miss by a bit and rack up the xp. As long as you both miss a fair amount, you could grind to your heart's content. Hell, you'd train up your dodge at the same time.

[Joseph Hewitt]

Thanks for the comments. One thing about the skills is that ever since they've been decoupled from specific stats, there's no reason why a single skill can't do multiple things- just paired with different stats.

Even if the skill system isn't changed, I think that converting MechaEng to a talent + having multiple Renown scores would be a good idea.

[Lord Afabie]

I'm not sure how well "xp based on near-failure" would work though. I can see that it would be relatively easy to pick a fight with a very nimble but weak opponent (Dora, anyone?) and just fire weapons with -2 accuracy penalties and the like. Then, you'd be likely to miss by a bit and rack up the xp. As long as you both miss a fair amount, you could grind to your heart's content. Hell, you'd train up your dodge at the same time.

But if you're missing a lot, then you're missing BY a lot.  Even if you could tune things so that you frequently barely miss, you will be barely hitting about as much, and any enemy that isn't a significant challenge should die off fairly quickly. 

I will admit that with mecha engineering you could design a mech that lets you get experience from a much weaker opponent, however that means you won't be capable of fighting anyone stronger then that without changing mechs.  If your opponent isn't a threat, that means they probably can't hit you effectively, and that means you won't be gaining much experience from dodging, or blocking, which happens much more frequently than attacks.  And you have to work harder to grind efficiently, making it less appealing. 

Plus this whole argument applies only to meha combat skills (in the new system there would be only one of those) there isn't a way to make you less effective with most other skills by more than a couple of points.  it's also harder to predict the difficulty of most non-combat challenges.

[DudeGuyMan]

I'm not terribly enamored with the whole renown concept in the first place. Let's face it, nobody goes "Wow my awesome mecha I've spent the whole game building just got blown up in a random encounter! Time to spend the next week getting garbage mecha shot out from under me until the difficulty slowly scales back down to my new loser status!" They just go "Oh crap time to reload and run from like five missions in a row so I don't get cornholed!" at which point you may as well streamline the process and just give them a difficulty slider.

If you scale the difficulty based on player actions, you're telling the player to game the system.

[Rowanthepreacher]

Lord Afable, perhaps. I regularly run with two or three mechs (one is a mule, admittedly), so I'm not sure that the wrong mech, wrong situation thing will come up much. While you're probably right, in that the combat skills are the only ones that would be much affected by gaming the system, if there's any equipment that can increase your skill levels for other skills (scanner pack, magnifying glass, survival knife, guitar/violin/ukelele, etc), you can take on missions significantly more difficult than your natural skill would allow for, and then dump the bonus gear for near-miss xp.

On the topic of "weak opponents die quickly" It's not much of a challenge at all to strip the arms off of an opponent and then plink it with a machinegun (admittedly, this was easier in GH1, because you could install your SF0 pistol). Weapons with Overload can be used to bring an opponent down to just the right level. If you hit too often, you can stop overloading and let them recover. If you hit too rarely, just pull out a +2 to-hit Overload weapon and get stuck in until they stop dodging so much. Lots of GH weapons are designed to be good at disabling a foe without destroying them, as part of the weapon balance, which makes those weapons crazily good at grinding xp.

The idea has merit, but I'd like to see how it actually works.

[Crucifix]

Oh hoh, I had just been lurking and peering in at all the good news coming out of here lately, but this looks like it's going to be awesome, since with talents and such I'd imagine these basic skills can become as specific as they need to be.

I'd agree that failure and narrow success should give experience (even a hopeless failure should give a tiny amount, you *can* learn by jumping right in and making a mess of things, or by just trying and trying until you finally get something right).

I'm not sure grinding is really a problem in a single player sandbox game, but worst-case, would it be possible to mark each source of exp with diminishing returns? You get decent exp for the first ten near misses, you swiftly drop to getting 1 or 2 since by then you probably know what you're doing wrong.


Concerning Reknown, my problem is verisimilitude - I have absolutely no problem with things rising up to meet the PC - you should influence more at level 30 than you do at level 5, which means bigger, badder battles and ever more extreme situations.... But it shouldn't be the universe that moves around you, a regular "mook patrol of pirates is messing around in the station exterior" should be a simple mission with a small bounty, not magically filled with Bargols rocking 21+ in Mecha Combat and offering a million credits - even legends don't spend 24/7 doing more and more extreme jobs, the big opportunities just don't crop up every day.

There should be elite opponents and incredibly dangerous missions that crop up independent of the player. A massive concert or deadly mega-death dungeon doesn't just show up when the PC is in town, or when she reaches Music-Reknown Level 75, but actually getting in on the gig requires smooth talking and a decent reputation. These should get more common at higher reknown (your social circle changes to incorporate more of such opportunities), but still remain uncommon enough to be interesting and cool.

This circumvents the whole "lost my mech, time to de-rep or reload" to "Lost my mech, time to keep a low profile and take some smaller jobs until I'm back on top".

Lastly:

I personally prefer "Performance" to Art, how about Artistry as an alternative? That's a fairly broad meaning to it.
And I prefer "Investigation", it sounds like more of an "active" process than insight.

[demnpercy]

Joseph-
I like the idea of changing to a system with a few general skills, with the manner in which you can apply those skills determined by your stats and talents. I was wondering if you'd considered breaking down the combat skills a little differently, though. Perhaps instead of Mecha and Personal scale combat, you could have Melee and Ranged combat, and a skill that would cover both Mecha Piloting and personal scale Dodge. It would increase the number of skills by 1, but it would also allow for a greater degree of character customization, with the downside of having one skill cover two activities that don't seem to have a lot in common - driving and dodging blows.

[Rowanthepreacher]

I kind of assumed that Mechs had you suspended in a sort of harness which tracked all your movements, and that the difference between doing an action on the personal scale and SF2 came from the fact that you're trying to pilot a huge, 30-60 tonne vehicle. If you dodge in SF2 like you do in SF0, you're not taking the increased reaction times, momentum and whatnot into account, so you dodge badly. If you're used to dodging in SF2, however, and you try the same dodging techniques in SF0, you'll be adjusting for momentum that you just don't have, and thus, failing to dodge. Same thing for the other skills.

[KA101]

*KA101 launches Analytical Fire Wall O' Text.*

Skills generally: something like Code Breaking could be a valid *overt* skill, whether military ("Hudson.  Run a bypass.") police (if the door alarm goes off, they'll execute the hostages) or piratical (opening the safe you stole *without* risk of breaking the fragile PreZero crystal discs it's supposed to contain).  If skills must be locked together, consider letting one group pinch-hit, at reduced effectiveness, for another skill's domain.

GH1 example: currently, no matter how much Knowledge, Science, and Investigation I have, I just can't understand PreZero languages I find on a monolith (because I don't have Mysticism).  BUT I *can* crack PreZero data-storage devices (the Genome and Ziggurat computer come to mind), presumably coded & written in PreZero languages, and understand their contents just fine.

Thanks for the comments. One thing about the skills is that ever since they've been decoupled from specific stats, there's no reason why a single skill can't do multiple things- just paired with different stats.

Even if the skill system isn't changed, I think that converting MechaEng to a talent + having multiple Renown scores would be a good idea.

Re skill+stat: PLEASE make the connections explicit.  Example: conversing at the party, lines, skill/stat involved, and any significant effects, in that order:
"It's VERY nice to meet you!" (Conversation + Charm) (Flirt)
"So, have you ever used a Reflex System?" (Mech Piloting + Charm)
"I am the invincible $PLAYER_NAME!" (Taunt + Ego) (Passionate)
"Don't you feel a connection between us?" (Mysticism + Ego) (Spiritual) [and/or?] (Flirt)

Re MechEng: Strongly oppose.  If you do make it a trait--something that costs 5K XP to buy and only 5 total slots through the entire character, IIRC--would you kindly let players have better stock mechs?  The idea that "all non-quest-reward mechs should have some flaw which needs MechEng to fix" makes having access to MechEng kinda important.

Re multi-renown: Support.  Performance != Piloting, which need not equal SF0 combat skill.  Possibly detective or med/sci skill (if the PC helps knock back a few epidemics, that should attract attention).

Generally: I think Crucifix is onto something.  If xe's correct about 100ish Renown giving GH2 opposition pilots Mech Piloting 21...well...that explains a LOT.

[PlaintextMan]

The number of skills will be cut drastically. Each skill will correspond to a specialty or career path, such as "Mecha Combat" for all mecha fighting skills. Different uses of a particular skill will use different stats- Reflexes for attacking, Speed for defense, etc. Characters may take all skills without penalty. Each skill will have its own individual Renown value; these separate renown values will be used for generating skill-specific plots.

I generally definitely oppose such radical axing of different skills. Simplifying is good, but a player needs a specific level of granularity of input that should be consistent throughout the game, and I think this would be taking things a bit too far.

However, the whole toughness/fitness thing can do with this, or simply be axed entirely. Crawl's approach to "support skills" comes to mind: almost no skills are purely for boosting attributes; eg. either Spellcasting or Invocations determine amount of mana but are primarily used for (respectively) spell success rate and deity invocations; Fighting skill, which helps melee/ranged combat rolls, also determines HP; Str stat is used for inventory carrying capacity.

Definitely in favour of the idea that different stats get mixed and used with certain skills (cf. approach to ranged combat in my combat concept test).

Also, what KA101 said about

Re skill+stat: PLEASE make the connections explicit.
Re MechEng: Strongly oppose.
Re multi-renown: Support.  Performance != Piloting, which need not equal SF0 combat skill.