Josh Sawyer at GDC Europe 2011

[funcroc]

http://schedule.gdceurope.com/speaker/Joshua-Sawyer

 

The Evolution of RPG Mechanics: From Die Rolls to Hit Volumes

SPEAKER/S: Joshua Sawyer (Obsidian Entertainment)

DAY / TIME / LOCATION: TBD

TRACK / DURATION / FORMAT: Game Design / 50-Minute / Session

 

DESCRIPTION: This session illuminates valuable lessons developers have learned in designing and implementing mechanics for contemporary cross-platform RPGs. Though classic RPGs have their roots in simulating tabletop games, recent titles such as Dragon Age, Mass Effect, Fallout 3 and New Vegas, Alpha Protocol, and Dungeon Siege 3 commonly borrow more player-driven mechanics from other genres. From the pitfalls of virtual die rolls to the challenges of balancing player skill with character statistics, the session breaks down what works, what doesn't, and why RPG fans are often their own worst enemies.

 

TAKEAWAY: Attendees will learn how to avoid the most common difficulties and pitfalls developers face when attempting to create and balance RPG mechanics. Specific examples from high-profile titles will be used to ground the lessons in practical terms for easy reference.

 

INTENDED AUDIENCE: The session is aimed at game and interface designers, animators, gameplay programmers, and anyone else involved in conceiving or adapting mechanics for role-playing games. Developers with a tabletop gaming background or who are involved in cross-platform RPG development should find this session especially helpful.

 

Joshua Sawyer

PROJECT DIRECTOR

OBSIDIAN ENTERTAINMENT

 

Josh Sawyer is a project director at Obsidian Entertainment. He has been working in the industry since 1999, when he started his career as a designer on the Icewlnd Dale series at Black Isle Studios. Since 2005, he has been working at Obsidian Entertainment and has contributed to Neverwinter Nights 2, Alpha Protocol, and was the Project Director and Lead Designer on Fallout: New Vegas.

 

Starting with the "red box" Dungeons & Dragons basic set, Josh has a long history with playing, modifying, and developing tabletop RPGs. Much of his development background has focused on adapting and updating classic rule systems for contemporary gamers and gaming platforms.

 

A Wisconsin native and graduate of Lawrence University, Josh currently resides in Huntington Beach, California with his girlfriend and cats.

Edited June 17, 2011 by funcroc

[funcroc]

OT: Avellone to attend GDC Online this year?

 

Drew Holmes (the lead writer at Volition): Also, I hope GDC Austin is ready for this October. We're getting the band back together! @SteveJaros (Volition's senior writer working on inSANE) @bluedev (the creative director on Reckoning) @ChrisAvellone

[Wombat]

Sounds interesting. For, at the price tag of Alpha Protocol and Dungeon Siege III, I think the designers need to check out established successful (mainly game-play) models and weave them into what they are doing especially when they'd like to seek for their originality in story-weaving. An attempt and ambition which tries to defy every single existing format such as Planescape: Torment is admirable but risking too much, at least, with that pricing.

That said, the presentation must be most about technical aspects...in reality, besides technical problems, AP and DSIII seem to have suffered the reactions triggered by disappointments through complex presumptions and expectations in the cross-genres, which is rather an issue of product design and marketing strategy area. In AP's case, I clearly saw it coming but, in DSIII's case, my lack of knowledge in online multi-player gaming blinded me.

[Wrath of Dagon]

I'd really like to know why RPG players are their own worst enemies. I thought greedy publishers and sloppy developers are the RPG players' worst enemies.

[Wombat]

Yea, it's quite easy to blame each other when things going wrong.

 

As for rule-sets, I'm not a great fan of D&D in general since it focused too much on combat rules rather than making interesting sessions. Even Avellone seems to have been quite surprised by how FO rules work when he came across with the game, which probably means he hadn't played outside of D&D rule-set.

 

Personally, I felt at home with FO and, partly, Morrowind. FO-like skill systems tend to be tied to how characters interact with given "worlds," where setting can play a quite important role in game-plays. If I can think of most convincing reason why skill systems are more interesting (at least, to me) than class systems in D&D is that it consciously takes characters as a medium to interact with an imaginary (or theme-focused real) world rather than combat units, a mere borrowed concept from predating combat strategy games. In fact, if someone is conscious of the core-idea above at all, how should the rule-set update be such a big issue?

 

IMHO, outside nostalgia, I don't believe D&D rule-set is a good place to start the discussion...for this part "why RPG fans are often their own worst enemies", maybe, but how constructive is it, going back to the first line of this post? Probably, what makes me annoy is that I find the image of RPG by CRPG only players tend to be quite limited so that I'm inclined to think it is more or less related with the dominance of a certain style of game-play introduced to some classic CRPGs.

[funcroc]

http://twitter.com/#!/Motte/status/103479805181308928

 

[Tale]

I'd love to see the rest of those slides. Unless maybe it was entirely a speech with just the one slide.

 

In which case, I wonder if there's a video or audio floating out there.

[Bendu]

Maybe Josh could post a transcript and the sildes on the new Obsidian blog when he returns.

[Lexx]

Josh said that there isn't much text on the slides, so a video or such would be much better.

[funcroc]

http://gaming-gratitude.blogspot.com/2011/...rope-day-2.html

Right after that highlight I switched rooms into a smaller room but not to a session less awesome. Joshua Sawyer from Obsidian gave a speech about implementing rulesets in RPGs and how this has involved over the years. Very interesting and bull****free. He wasn't afraid of displaying his own work (Obsidian games) for examples of bad decisions. Really liked the presentation. Could have gone on for ever.

 

http://twitter.com/#!/devilkitten

- Joshua Sawyer talks about his roots. And I recognize myself <3!

 

 

- Joshua Sawyer talks about the casino system in Fallout New Vegas and how random gambling was made fun.

 

 

- JS talks about the sentimental and analytical resistance to making gameplay feel better.

 

- Make sure that the changes in stats are big enough to make the player perceive the change.

 

- Players in digital games can just reload. Which makes the standard random challenge pointless.

 

 

- Strategic failures - if you don't have a thief or a bard you fail. Strategic planning has to be transparent.

 

 

- More forgiving games make it less varied for the experienced player, but better when it comes to strategy.

 

 

- You have to be aware of what you ask of players. Don't expect the player to manage too much.

Edited August 17, 2011 by funcroc

[funcroc]

SAF's Alpha Protocol thread

 

Look what I saw at GDC Europe:

 

[ShadowScythe]

SAF's Alpha Protocol thread

 

Look what I saw at GDC Europe:

 

*snip*

 

I think according to the SA forums he's talking about how Alpha Protocol implemented the rng into the shooting mechanics poorly.

 

EDIT: No wait, it's the fact that the reticule for the shotgun is ridiculously huge.

 

This presentation looks really interesting- I do hope someone posts it on youtube like that Chris Avellone one.

Edited August 17, 2011 by ShadowScythe

[Lexx]

I never used shotguns, because the reticule is too big for far away targets. This is probably as well what the image above says.

[Nightshape]

And here I was thinking, no human being likes to shoot hot chicks with shotguns, especially when your eye is pulled to it by a massive yellow cirlce.

 

The message could also be, no human being likes AP - which would give the impression AP fan's are inhuman

[Tale]

- Make sure that the changes in stats are big enough to make the player perceive the change.

 

This is something I need to hear more about. My current line of thinking has been that games tend to inflate stats too much, causing problems such as encounters becoming too trivial, and maybe responsible for the problems with leveled encounters, such that if you tone down stat advancement you may not need the leveled encounters at all.

 

I want to know where I'm going wrong in that line of thinking.

[Enoch]

- Make sure that the changes in stats are big enough to make the player perceive the change.

 

This is something I need to hear more about. My current line of thinking has been that games tend to inflate stats too much, causing problems such as encounters becoming too trivial, and maybe responsible for the problems with leveled encounters, such that if you tone down stat advancement you may not need the leveled encounters at all.

 

I want to know where I'm going wrong in that line of thinking.

I think you're confusing the granularity of RPG character progression with its scope. Josh is talking (I think) about how big each step is in the progression of stat/skill/ability increases. I think you're talking about how large the difference is between beginning characters and advanced characters. They're related concepts, but not exactly the same. You can get from "basic bumpkin-with-a-sword" to "awe-inspiring colossus of power" in few large steps, or in lots of barely-perceivable steps. (Alternately, you can progress from "competent hero" to "somewhat more competent hero" over the course of a game in a similar manner.)

[Tale]

- Make sure that the changes in stats are big enough to make the player perceive the change.

 

This is something I need to hear more about. My current line of thinking has been that games tend to inflate stats too much, causing problems such as encounters becoming too trivial, and maybe responsible for the problems with leveled encounters, such that if you tone down stat advancement you may not need the leveled encounters at all.

 

I want to know where I'm going wrong in that line of thinking.

I think you're confusing the granularity of RPG character progression with its scope. Josh is talking (I think) about how big each step is in the progression of stat/skill/ability increases. I think you're talking about how large the difference is between beginning characters and advanced characters. They're related concepts, but not exactly the same. You can get from "basic bumpkin-with-a-sword" to "awe-inspiring colossus of power" in few large steps, or in lots of barely-perceivable steps. (Alternately, you can progress from "competent hero" to "somewhat more competent hero" over the course of a game in a similar manner.)

Ehh, it's not really worth sidetracking the thread. I could easily be reading the one line wrong. "big enough to make the player perceive" doesn't mean big. And I'd agree on the point that the player does need to be able to recognize a change has occured in his character. I just want smaller increments than we typically see, for reasons I'm pretty attached to.

[Nathaniel Chapman]

Here's an example and a counter-example of what Josh is talking about.

 

In Deus Ex, when you buy a skill rank, the change is instantly noticeable because the granularity is so large. You have 5 ranks and so each rank can afford to be a massive, obvious improvement to players.

 

In Fallout, a single point in a skill makes no perceptible difference in the actual player experience.

 

The idea is that it's better in most cases to give fewer large rewards than many small rewards. This is especially true when you have a fixed-endpoint single-player game. Obviously it's harder for something like an MMO. Also you can mix the two - Perks are an example of a less granular, more perceptible change in Fallout.

[Giantevilhead]

Frequent small perceptible changes should be the most effective. Also, persistence can be created by stretching the ratio of reinforcements. Of course that's already built into the system with greater experience requirements for higher levels but then level caps tend to screw up the system.

Edited August 17, 2011 by Giantevilhead

[funcroc]

GDC Europe: Obsidian's Five Hard Lessons Of RPG Design (Gamasutra)

[Giantevilhead]

I don't agree with the first point. I would say it's more of a matter of giving the player the ability to mitigate the risks involved in randomization. Some players like risk, while others do not. Developers should offer options for both kinds of players. For example in an RPG, the player should be given access to both skills and item that with a fixed stat/bonus, like a gun that does 50 damage, and skills and items with a variable stat/bonus, like a gun that does 30 to 60 damage.

[J.E. Sawyer]

Even tabletop players know it's better to have an attack/weapon that does more dice of damage than one with about the same range but fewer dice. For example: a spell that does 6d4 damage vs. one that does 4d6+1. Even though the latter has a slightly higher average (and max), the former will result in a smaller bell curve. An even better example is a 2nd Ed. longsword +1 vs. a broadsword. Ignoring the obvious additional benefits of a longsword +1 being magical and being more likely to hit, 1d8+1 has a broader bell curve than 2d4. Generally speaking, more reliability = better than.

 

In my talk, I didn't advocate completely eliminating randomization, but I did advocate eliminating extreme randomization and randomization that allows for "uncontested" re-rolls via reload. E.g.: lockpicking, speech checks, crafting checks, etc. A good combat example would be the "Old School" Disintegrate spell. Make the save or you're annihilated. It's an all-or-nothing spell. People used it to great degenerate gameplay lengths in old Infinity Engine games to kill powerful enemies on the first round of combat. If the creature made the save, they'd just reload. It's all-or-nothing mechanical chaos. Newer versions of Disintegrate simply do a lot of damage on a failed save, with less damage on a successful save. Still potent, but not live-or-die, and less likely to encourage save scumming in a CRPG environment.

[Niten_Ryu]

SAF's Alpha Protocol thread

 

Look what I saw at GDC Europe:

 

Oh yeah. I've been saying same for years... except I actually liked it but players in general hate it

 

That feature certainly don't have place in modern game design.

[Giantevilhead]

Even tabletop players know it's better to have an attack/weapon that does more dice of damage than one with about the same range but fewer dice. For example: a spell that does 6d4 damage vs. one that does 4d6+1. Even though the latter has a slightly higher average (and max), the former will result in a smaller bell curve. An even better example is a 2nd Ed. longsword +1 vs. a broadsword. Ignoring the obvious additional benefits of a longsword +1 being magical and being more likely to hit, 1d8+1 has a broader bell curve than 2d4. Generally speaking, more reliability = better than.

 

But "better than" is an opinion of the player. That's why you have classes like the Wild Mage, which is all about risk.

 

In my talk, I didn't advocate completely eliminating randomization, but I did advocate eliminating extreme randomization and randomization that allows for "uncontested" re-rolls via reload. E.g.: lockpicking, speech checks, crafting checks, etc. A good combat example would be the "Old School" Disintegrate spell. Make the save or you're annihilated. It's an all-or-nothing spell. People used it to great degenerate gameplay lengths in old Infinity Engine games to kill powerful enemies on the first round of combat. If the creature made the save, they'd just reload. It's all-or-nothing mechanical chaos. Newer versions of Disintegrate simply do a lot of damage on a failed save, with less damage on a successful save. Still potent, but not live-or-die, and less likely to encourage save scumming in a CRPG environment.

 

Wouldn't the logical solution be to introduce a "degrees of success" type system? Making a random check save or die will make the player reload until they succeed but making the player meet a specific skill requirement also has problems. If you see a door you don't have the skill to unlock, it's pretty annoying having to remember to come back when you raise your skills enough. You'll either forget or just cheat to unlock it. You have similar problems with other types of skill checks. If you meet a speech challenge you can't beat, then you can just reload an earlier save, level up and then come back to beat the check.

 

So I guess the way to do it is to lower the skill check required to succeed but add rewards or punishment based on the player's skill rank vs. the DC check. For example, a lock may have a DC of 25 but you can beat it if you have 5 lockpicking skill at the cost of a bit of extra resources, but don't make the cost too high, and maybe a "frustrated" or "hand cramp" debuff. If you have 25 lockpicking skill, you unlock it for "free" and maybe gain some bonus experience. If you have 50 lockpicking skill, you can recover some parts from the lock which you can use for crafting or sell for cash.

Edited August 18, 2011 by Giantevilhead

[Sannom]

What I find interesting about Alpha Protocol is that although it randomized 'normal' fighting way too much, it completely took away the random element of the random element by excellence : critical hits. Heck, I even wonder if those big spreads weren't meant as an incentive to use the critical hit mechanics of each weapon, and indeed someone playing those games should learn how to use them the most effectively.