Here's another roundup of developer posts in official Obsidian and Bethesda forums, as well as J.E. Sawyer's Formspring. A lot of them are modding-related, as J.E. Sawyer frequently answers to modders' questions.
“The 2D borders around the Mojave Wasteland world map do not have any relationship to the size of the game world relative to F3's Capital Wasteland. The Mojave Wasteland is not a perfect square. It is taller than it is wide and the boundaries are mostly defined by natural features: the Spring Mountains, Colorado River, etc.
The approximate size of the Mojave Wasteland is 60-65 tiles tall by 40-45 tiles wide, tallest around Nellis AFB, widest around Jacobstown. The size of the Capital Wasteland is 51 tiles by 51 tiles. If anyone wants to open the GECK and load WastelandNV, you can see its exact size.”— J.E. Sawyer
“First, let me get this out of the way: I do not believe that it should be a design goal to make one skill better than another. While it's impossible to perfectly balance all skills, I don't think that Unarmed should be designed with the goal of being inferior to Guns, or that Energy Weapons should be designed with the goal of being inferior to Explosives. I believe the goal should be to make the weapon skills (specifically, in this case) distinct but comparable in overall utility. This means that I think Energy Weapons should not be inherently better than Guns and I don't think that someone who focuses in Unarmed should drop dead in every open fight because it would be realistic to do so.
The goal with Energy Weapons was ultimately to make them "outside the boundaries" of Guns. By that I mean that Laser Weapons would tend to be more accurate, faster firing, and lower DAM (usually, not not always) with higher DPS than its equivalent gun. Conversely, Plasma Weapons would tend to be slower firing and higher DAM. There are some cases where this isn't true, the most notable example being the Cowboy Repeater and Plasma Rifle, both of which are considered "tier 2" weapons for their classes. When balancing some of the Guns toward the end of the project, I did bump the Cowboy Repeater up relative to other Guns but did not bump up the Plasma Rifle, which is an oversight that I regret. The Plasma Rifle does have a much faster reload rate than the Cowboy Repeater, but it should also do a higher base damage. Ultimately I do think that I did tune some of the Energy Weapons too far out of "spec" compared to Guns, mostly at the low and mid-range. I did not intend to make Energy Weapons "bad" and I apologize if that is the perceived result.”— J.E. Sawyer
“I really don't remember "hyping" alternate currency, though the subject was certainly a very hot topic on the forums and folks speculated (lol) about it a lot.
Unlike Pre-War money, Legion Money and NCR Money are not affected by the Barter skill. I.e. $5 NCR is always worth 2 caps and a Legion Aureus is always worth 100 caps. I don't think I ever suggested some Bardi-esque currency inflation/speculation gameplay, just that the different factions had different forms of currency and that each was "backed" differently: caps are water-backed by Hub merchants due to the decline of the inflated NCR fiat currency and Legion money is not "backed" at all, but of value due to the use of commodity rare metals. The currencies are there for flavor and to reflect something larger about the societies that use them.”— J.E. Sawyer
“I never boasted about it as a game mechanic. I explained that the different factions had different currencies (they do) and what the ideas behind the currency were. I explained this because community members were spinning their wheels debating about it for literally pages upon pages.
Yes, exploring more game mechanics with currency could certainly be interesting. Many things could be more interesting if they were more fully developed. We expanded crafting, ammunition subtypes, weapon modification, melee and unarmed special moves, and added gambling minigames, among other things. Devoting time to complex currency mechanics didn't seem like something that was worthwhile given everything else we were doing.”— J.E. Sawyer
“You can skip any number of the steps of the first half of the main quest and the story still makes sense. As with Fallout 1, the steps you follow to reach your goal are information-oriented. You go to Point A to learn about going to Point B, which leads you to Point C, but at these locations you're not doing anything that changes the world in a way that's necessary to continue to the story. Nothing prevents you from going to... the place where the water chip is... right after leaving Vault 13. It's just difficult. And that's the way F:NV's crit path is set up. You're just physically closer to your goal in Goodsprings than Vault 13 is to the water-chip-place.
The second half of the main quest is more open in terms of the order in which you do things, but you are required to do more things to advance the MQ.”— J.E. Sawyer
“Talking with our producers, aside from tech stuff specific to PCs, anything you're seeing in the PC updates should be making it's way into the next Xbox update. We did a supplemental update for PC this week (it's a different process from consoles and doesnt require the same approval process) -- so you can see some of those fixes here.
The "comprehensive" update will include these fixes and other fixes being worked on.”— Matt Grandstaff
“We discovered why companion skills were all over the place (mostly low) and why Boone was the god among scrubs. One issue was a bug related to calculating the leveled skills of characters once they were set to followers (they were reset to 1st level values). The other problem was that Boone, specifically, was having bonuses applied to his skills through his script. This resulted in Boone always having over 100% Guns regardless of when you recruited him. Weapon skill not only increases damage, but reduces weapon wobble during movement, which is why regardless of where Boone was moving/what he was doing, he was able to pull off running headshots that drop the target in one attack.”— J.E. Sawyer
“Reputation is a tricky thing because it's sort of complicated under the hood. I guess I'm to blame for that, but I think the nuance is worth the hassle.
With any given group for which reputation is defined (in GECK), the player has two reputation values under the hood: good and bad. These are separate values. I.e. New California Republic - Good and New California Republic - Bad. If you rescue an NCR lady's pet cat from a tree, you get a bonus to NCR Good. If you steal an NCR baby's sucker in front of its mother, you get a bonus to NCR Bad. These values typically only go UP. If you do something bad, your good reputation doesn't get worse -- your bad reputation goes up. There are exceptions to this in Fallout: New Vegas. At specific points in the game, you are given amnesty by a group even if you have a terrible rep. This effectively wipes out the "Bad" component of your reputation with them.
What the player actually sees in game is an aggregate of the compared good/bad values on a scale defined for the reputation group. If you have only good or mostly good, you are going to be on the good axis. If you have only bad or mostly bad, you are going to be on the bad axis. If you have a relatively even mix of good and bad, you have a mixed reputation. This is what GetReputationThreshold checks, the aggregate that the player sees (Liked, Wild Child, Vilified, etc.). It is important to note that the player can only be on one of these three axes with a group because all reputation titles/ranks are mutually exclusive. Yes, you have a good rep and bad rep meter behind the scenes, but what the player sees and what NPCs should react to is the threshold.
The last piece of this puzzle is reputation scale. Each group that uses reputation has a "scale". This is used to determine how quickly the player moves through the ranks from performing actions that alter his/her good/bad rep scales with a group. Shooting a person in the face will always give you X points of bad reputation, but it will change your threshold more quickly in Goodsprings (which has a scale of 30) than with NCR (scale of 100-ish). If you set the scale high enough, the player will have to perform an enormous number of actions to alter his/her reputation title/threshold. If you set it very low, getting caught stealing a piece of bread can instantly vilify you.
N.B.: Because of the way the system is set-up, players that do a bunch of crazy stuff like save a town and then kill most of its residents will find themselves permanently stuck with a Wild Child reputation. This is intentional, but if modders want to allow players to worm their way out of it, you can always allow some sort of option to reset their reputation values (maybe as the result of a quest, or paying a propagandist, etc.).”— J.E. Sawyer
“0 is the mixed axis, 1 is the good axis, 2 is the bad axis. At any given point, the player's rep is on one of these axes (only) and at a value between 0 and (I think) 6. The higher the value on the axis, the stronger the reputation. Idolized is the highest good, Vilified the highest bad, and Wild Child the highest mixed.”— J.E. Sawyer
“It's almost always better to use the threshold, since checking the individual reputations (good rep, bad rep) doesn't tell you anything about the other one. So you could have 10 good NCR rep and 100 bad NCR rep, but if you use GetReputation to just check what's on the good scale, you don't know what's on the bad -- or vice-versa.
“Rapid Capacitor Weapon.
EDIT: The admittedly simplistic idea is that the ECP loaded into the "drum" area is powering banks of capacitors in the "drum" around it that fire in rapid succession to give the weapon its high RoF.”— J.E. Sawyer
“It is also possible to selectively exclude certain ammo subtypes from a weapon by pointing it to a different form list for its selected ammo (or simply to an ammo form).
For example, if you want the AER14 prototype to only use stock MFCs or to only use stock + OC + MC and not the special projectile-changing version, it can use a separate form list for ammo than the normal laser rifle.
You can do similar things with .38 Special revolvers that can't fire .357 Magnum (for example).”
“Another important tip: the ##sightingnode does not use the orientation you give it. I.e. it will always be aligned with the main axis of the weapon. This means that your ##sightingnode, rear sight, and front sight must be in a straight line parallel to that axis.”— J.E. Sawyer
“Explosives are meant to be used against groups. The Fat Man is pretty far from useless as long as you keep that in mind.”— J.E. Sawyer
“Yes, that's an Explosive. What do you think it should be categorized as?
If you want to take out a single high-HP target quickly, use a weapon like the Anti-Materiel Rifle, Gauss Rifle (if armored/at range), Minigun, Gatling Laser, (if unarmored/at range), Light Machine Gun (mid-range/mid-armor) or 12.7mm SMG (short range/mid-armor). The upgraded Plasma Caster is also very effective when loaded with max charge MFCs. With the right perks, an Unarmed character equipping the Ballistic Fist can wreck deathclaws quickly.
“I'm not sure if this is possible. It may be with 10 INT/Educated/Tag!/Comprehension + all books but you'd have to be determined.”— J.E. Sawyer
Has Obsidian considered the possibility of working on smaller and more focused project? No offense, because I REALLY like your games, but it's painfully clear at this point that you don't have the manpower and expertise to work on AAA projects.
“I'd love to work on smaller games, but ultimately I don't determine what projects the company undertakes.”— J.E. Sawyer
Happy Birthday Josh! May New Vegas be your most successful project ever! That said, there's something I wanted to ask: does your reputation (as in Obsidian's) as 'great writers but poor programmers' feel like a burden on your shoulders?
“I think our programmers often have to deal with incredible stress and very difficult problems for which there often are no easy solutions. They also often get blamed for bugs that are not their fault, which is even worse. They aren't recognized for the work they do and they are blamed for work for which they were never responsible. That's pretty crummy.”— J.E. Sawyer
“Not really. I guess there are really two things to examine in the review. The first are the implications of laziness and/or incompetence. Those implications are irrelevant; Fallout: New Vegas is what's being reviewed, not Obsidian. Additionally, I and the other people on the team know what level of effort we put into the game. People not involved with the development of the game, whether reviewers or endusers, do not.
The second issue is the state of the game. That is the point of the review and the reviewer's comments seem as fair as anyone else's.”— J.E. Sawyer
The writing in New Vegas is truly exemplary, kudos to you and your team, but that seems to not get mentioned in many reviews. I posted some forum comments to that effect and was basically told writing doesn't matter to a lot of people in games. Agree?
“Writing absolutely does not matter to a lot of people playing games. This is something I've accepted for a long time. For a lot of RPG players, game mechanics really don't matter. They will gladly march through a game that they hate if they enjoy the writing and story.”— J.E. Sawyer
What a great game. The writing is absolutely phenomenal. And the whole faction mechanics, the complicity and importance to story and character feedback, is a new landmark in storytelling in video games. Who did the writing on Rose of Sharon Cassidy?
I'm curious (and I mean this question in a very basic way) about how you develop on three platforms at once (PC, PS3, and 360). When it comes time to test a 'build', I imagine that you can just launch the PC version. What do you do to run it on 360/PS3?
“We have proprietary tools on the PC that allow us to transfer local or remote builds to the 360/PS3 dev kits. This transfer copies over any platform-specific data as well as any current user data (plug-ins/local design overrides) to test content. Once it's transferred, we (developers) launch the game from the console's dashboard or remotely from the tool.
Testers do something similar but may be running under optical drive emulation or off of an actual burned DVD depending on what they are testing.”— J.E. Sawyer
“We had to make the songs pretty quickly toward the end of the project. For the guitar songs, the guitarists (Nathaniel Chapman and James Melilli) had to learn the pieces in an hour or two. There wasn't any time to transpose anything, so I had to sing the songs in whatever key we had guitar tabs for. For side content, I think they turned out pretty well.”— J.E. Sawyer
Who wrote most of the dialog options? To be totally honest, I think they are a slight step down from Fallout 3 in terms of wit and humanity. Nonetheless, the rest of the game and its characters are just as good as its predecessor.
“Dialogue writing was shared by a number of designers. A few designers focused heavily on writing while others split writing responsibilities with area design and implementation.”— J.E. Sawyer
“I implemented and tuned all of the weapons, designed the perks, all of the system and combat design modifications, and wrote Chief Hanlon and Arcade Gannon. I also did the high-level concept design for all of the areas, but didn't do any core area design or implementation myself.”— J.E. Sawyer
The Caravan minigame is a great idea, but it was exploited soon after release via 1. the initial discard mechanic and 2. a stacked deck of 7s, 9s, and 10s. Do you have any ideas for improving/fixing the minigame, and d'you think we'd see these in a patch?
“The biggest flaw in the current AI (a known bug) is that the enemy AI will never play face cards against the player. High-value decks like 7/9/10 or 8/10/K don't work very well against an opponent who can play jacks or kings against you. You wind up spending a lot of time recovering from nuked 16/20-value cards or a busted caravan sitting at 40 that forces you to jack your own high-value cards or destroy the caravan entirely.”— J.E. Sawyer
“There are many things about the human companions that are statistically atypical, but there are only four of them, so it shouldn't be surprising that they don't break down along probabilistic lines. If designers strictly followed what was probable in companion design, there would likely never be gay or bisexual characters at all given the relatively small pool players are choosing from in any given game.
Ultimately, the companions who do express their sexuality (of any sort) do so as an aside and usually only when being directly asked about themselves by the player.”— J.E. Sawyer