Games
Interview with Josh Sawyer of Obsidian Entertainment

Interview: Part 1





One of the first things mentioned about Project Eternity was the concept of Souls being important and a source of power, that you were interested what world building ideas are generated from that design mechanic.
Where did the inspiration behind this concept come from ? I've heard many people mention NWN2: Mask of the Betrayer, although I personally thought of Lionheart: Legacy of the Crusader.



It actually came out of some thoughts I had about the physical and metaphysical underpinnings of our own world. When worldbuilding, I think a lot of designers want to explain everything up front. There's obvious value in defining how the world works because it helps everyone wrap their heads around what the setting is about. Over the years, I've felt that breaking down the supernatural into easy-to-comprehend chunks drains the magic from it.

Compare this to our own observation and understanding of the physical world. Public reaction to the discovery of the Higgs boson particle was very telling. Despite the scientific community's general requests to stop calling it the "God particle", the public and media couldn't help themselves. A discovery that potentially explains, if not the "why", at least the "how" of existence is appealing.

Project Eternity's world has a similar level of flawed understanding. They can perceive souls, they can detect and record some data about them, they can verify their findings to a certain extent, but they still don't really "get" how it all works. Arguably of greater importance, they don't understand why souls work (and don't work) the way that they do. Individuals also don't agree on the role the gods play in the cycle. Last week, our art director, Rob Nesler, came in and asked me, "Are the gods actually gods or just beings of immense power?" While there aren't many people in the world of Project Eternity who deny the existence of gods, that question is one that people in the setting have debated for millennia.



Project Eternity now contains 11 playable classes. What is your high-level design perspective from a class point of view? Are you thinking of using a completely homogeneous system (a la D&D 4E) or a system that is more fractal (like 2E or 3E)? Do you think it will be challenging to make all of the classes feel completely different?



With 11 classes, we have a lot of challenges. The ones I'm focused on are ensuring each class feels distinct, has genuine value to the player, can be built in myriad distinct ways, and does not rely on obtuse UI elements to function. 4E's classes are more homogenized than we will shoot for. I think our classes and level progression options will fall somewhere between 3.5E and 4E.

(They are releasing an update next week including the system mechanics of classes, which is why I assume this information is vague)




I saw Feargus mention that the idea in Project Eternity is have the party be able to make it to around level 12 in D&D Power level, which is absolutely perfect in my opinion. What are your thoughts on the leveling rate, power curve and Experience Table in Project Eternity? Will the level curve be slower like Baldur's Gate 1? Do you plan on making lower level play more exciting like D&D 4E? Will experience gain be static or scaled? My favourite gain rate was 2nd edition with the Fighter Experience table personally.



I think leveling will be a bit quicker than in the original BG and Pool of Radiance, but not as brisk as Icewind Dale, so somewhere in the 10th-12th level range (A/D&D equivalent) seems likely. I would also want the low-level play to be more enjoyable than it was in earlier A/D&D incarnations. What 4E did was essentially take Dark Sun's idea of functionally starting all characters at the equivalent of 3rd level. The most relevant aspect of this was that hit points tended to be double-digit right away, which minimized the possibility of random monstrous crits outright murdering a character in one shot.



Regarding the art style of the game, what games in particular have inspired the look and feel of the 2D Backdrops, Portraits, Avatars and Paperdolls? (for example IWD1 for backdrops, ToEE for avatars etc)



We look at Icewind Dale levels for a lot of our inspiration. They were beautiful settings full of atmosphere, interesting architecture, and a ton of cool, hand-painted details. Also the relatively subdued color palettes of that art falls in line with what we want to explore, similar in saturation what you might see in the art of the Hudson River School. We also look at Icewind Dale portraits because the brushwork of all of the artists tended to have an enjoyable mix of loose and tight strokes.



How customizable would you like to make the player paperdoll and to what extent would that be reflected on the avatar ?



Customization is important to us, but we need to see exactly how large the avatar is going to be on screen before we know the full extent of customization details we want to support. Also, we need to be cognizant of the logistics of how many assets we need to create across every race and sex combination. Thankfully, this analysis is something we've had to do in a number of previous titles, so we can answer those questions pretty quickly.



At first I was a bit scared of the combat being like Dragon Age or an MMO when I heard that the rounds/d20 system was being done away with and some form of cooldowns being used, but after a while it dawned on me that you guys have the potential to create a freaking amazing combat system unlike anything we've seen before in this style of game. Tim Cain has mentioned Real-Time Strategy games as a drawn influence for the system design of this game. What real-time strategy games have influenced some of the design decisions? One wanted feature I've seen mentioned on the forums is a Warcraft 3 style mini map.



You'd have to ask Tim for his specific influences, but personally I think unit responsiveness is very important (and often lacking in RPGs), positioning is important, and the ability to combine units in different arrangements for different challenges is important. We want people to feel like they have a lot of tactical options for approaching a challenge. A challenge that requires one specific build to overcome isn't really tactical gameplay; it's more of a combat puzzle. I think that's less satisfying in a party-based game where you control the composition and build of your group.



A good example of a tried and true RTS combat system is the one used by Warcraft 3/Defense of the Ancients/DotA 2. Units and Heroes have a Base Attack Time in seconds (1.7), as well as an Attack Animation time (eg 0.5) a Damage point, and a backswing time (eg 0.7) and attack speed is calculated by scaling these values [ Animation or BAT / (1 + attack speed bonus)]. Is this a similar direction to the one you're going with moving combat system design away from the d20 system ? The ability to cancel attack animations/casting animations has disappeared in more recent games as well but these were handy in the IE games.



Calculating actions (and cancels) in frames (or seconds, from a user perspective) is something we've had to do on most of our recent titles, especially Dungeon Siege III and Fallout: New Vegas. It's less important for the player to perceive when it's not a direct input system (like Dungeon Siege III or New Vegas), but working in real-time means, yep, you have to think about things on a real-time scale. In a round-based system, increments to attack rates need to be made on a coarse-granularity scale. That's why older editions of AD&D had attack rates like 3/2 before 2/1 and GURPS gets into some complicated time unit subdivisions. When the round is the basis for combat timing, you don't have many other choices.



To make the game feel like an Infinity Engine game but also encompass a balance of combat and non-combat options, you will need a lot of systems that use checks (to hit rolls, skill checks). Are you planning on implementing a homogeneous system for all checks? Now that you are not limited by a 20-sided die, what are some of the design possibilities regarding checks?



Personally, I've never felt that the d20 was a limitation to resolving challenges. I did like the 3E switch to a unified mechanical system because it made it easier to understand the relative difficulties of tasks. Instead of using d100 for some checks, d20 low for other checks, and d20 high for other other checks, the resolution mechanics focused on rolling high on a d20 for almost everything. People seem to respond well to a 100-point/percentile scale because it's a common scale of measurement, but I'm not sure what we'll do to quantify our checks.

I do know that we want to allow the player to account for a marginal shortfall in skill requirement with resource consumption. E.g. if you need a 75 to pick a lock but you have a 65 Lockpicking skill, you can still open the lock for a cost of 10 lockpicks. This wouldn't apply if the skill is too far from the requirement (e.g. a 45 Lockpicking skill would be too low), but it does allow for a little more flexibility.



Implementing Some of the UI features that the IE games had would go a long way to adding that IE feeling to the game. I would just like to say that if you do a select all button, have it in the bottom right corner of the screen like Baldur's Gate! Some other important features of the IE games UI were the Combat log and the Object Highlighter. These are features that could potentially be linked to expert mode. Would you like to include a number crunching combat log and an object Highlighter ? If so I would like to cast a vote for easter eggs in tiny holes on the map like the Ring of Wizardry that don't show up on the highlighter.



I think combat logs are great for understanding the mechanics in detail. Not everyone uses them, but I definitely think they have value. I also think select all is very handy. Many people like to do a fast marquee select, but a select all button is nice as well. The object highlighter feels almost like a requirement to me. I've never seen the fun in pixel hunting, especially when you're often dealing with something that the characters could likely see very easily. This doesn't preclude the existence of "secret" containers or doors, though, since those can be turned off by default.



Computer Hardware has come a long way since the heyday of the Infinity Engine games and these days many hardcore PC gamers are running high-end rigs and 120Hz Monitors. Many games these days are optimized to run at 30 or 60FPS due to console limitations. Since Project Eternity will be a PC-only title, one feature that many of my readers will be interested in is the Run-Time Performance of the game. Are these factors a consideration for the development of the game? Or will the game be using a frame cap like the older sprite-based games such as Baldur's Gate and Diablo 2 ?



We'll have to see what our performance considerations are, but we want this game to run on a large number of machines. Even if we do have some fancy features, we will allow people to turn those down or off if they are using lower-end computers.
Next PagePrevious Page
Back to Top