Notes on Thief: Deadly Shadows

Slow Learner

You know, when Thief first came out I wasn't into it. I thought it was too much waiting, and too punishing when you got caught. I didn't give it enough of a chance. But enough people raved about it, so later I played Thief 2 and realized that the game wasn't so much about hiding in the shadows, but about sneaking up behind people and knocking them out. That knowledge, and an increased reliance on quicksave, made it one of my favorite games ever.
The Thief 3 tutorial makes sure you know the bread-and-butter gameplay of Thief before it turns you loose, including a demonstration of how to use flash bombs, which I never used that much in Thief 2 but now realize are very useful.
Something game designers hear a lot: "I didn't like your game because such-and-such was too hard." At which point the designer says, "Oh, but you can just such-and-such to get past that point." We can't ship the game designer with the box, so we make these tutorials, hoping and praying they'll teach the player everything they need to know to have fun, but all too often don't.

That's What I Would Have Done

If you'd handed me the Thief license and said, "Where can we go from here?" I would have said, "Let's put it in a continuous world. The city can be an overworld and then you can go inside basements and houses - they'll be your dungeons." Which is just what they did. "Ideally, it would be seamless, but we're stuck with the Unreal engine and not made out of money, so people will just have to deal with the load times as you go from district to district." Check. (We were going to do the same thing with Spider-Man 2, but then we decided that we are made out of money, so we went for seamless.) "And we could gradually expand the area of the city that you can get to as you progress." Check. "And the outfitting you used to do between missions can be done in some kind of black-market shops." Check. "And we can have side quests." Check. (Although I didn't actually do any.) "And hidden areas with extra loot." Check. The end result: it's more than a game, it's a world.
Thief 3 has small levels that are exquisitely detailed. To my untrained eye it's as close to photorealistic as action-adventure has gotten. The framerate suffered for it on my poor PC. Still, it's as if they decided to do the opposite of GTA; instead of a large, empty city it's small and jam-packed with interest, and I think that's what allowed them to make the levels look so nice. I think it was a good choice.

Crisp Gameplay

Simulation vs. Game - as graphics get more realistic, one might think it's a good idea for the simulation to keep pace. There are all kinds of places in Thief where the simulation isn't convincing: when you drag a dead body away, you leave his sword there, and nobody seems to notice; bad guys searching for you in the shadows walk right up next to you and don't notice you; the bad guys keep a run of banter going as they search...as if you're telepathically listening to them. (Maybe you are, I'm not clear on whether Garret's keeper training included telepathy or not.) When they find a dead body, they search for you a little while, and then resume their normal patrol. (Shouldn't you guys be calling for backup or the coroner or something?) BUT 'fixing' these flaws would probably ruin the game, making it too punishing.
Underneath all the pretty graphics there is a rule system. Blackjack or shoot an unaware opponent and they die; once they're aware, the blackjack does nothing, and the arrows hurt but do not kill. Not a very good simulation (surely a well placed blackjack or arrow would kill them whether they were aware or not, eh?) but a good game, one that encourages stealth above all. Once you've used your flashbomb, you can run, but not fight - hitting or shooting your opponent magically restores their sight. Again, not a good simulation, but a good game. Your opponents hear you if you run, but they don't hear the sound of their friend's body slumping to the floor after you've sapped him. Learning the rules takes time, but once you've learned them, you start feeling the sensation of mastery; you can play this game perfectly. (And when you do slip up, there's always the flash bomb.)
I imagine the split between the quality of the visual simulation and the abstraction of the gameplay bothers some people. I can imagine people saying, "Why didn't that kill him?" and "Why was I caught just now?" It doesn't bother me. (Who was it who said, "I don't see Mario; I see a cursor"? I can't find the reference.) For me, good graphics on an abstract game is like having a high quality chess set.

This Industry Chews Up Directors

So Randy Smith quit after Thief 3; Harvey Smith quit after Deus Ex 2; Chris Soares stepped down after Spider-Man; it's looking like Tomo may step down after Spider-Man 2...he certainly seems battered by the experience...; am I just focusing on anecdotes or is directing really rough? It seems like they're flogged for every imperfection, and they get it from both ends: people in upper management want to make sure their thumbprint mars the director's vision, and people on the team get pissed off because they think the director is working them too hard, or is focusing on the wrong things, or whatever.
(I just looked for that bit in One Flew Over the Cuckoo's Nest about how the bird at the head of the V can't hold that position forever and needs to take turns but couldn't find it.)
We have an idea for making the job easier next project; I'll tell you how it worked out in three years.

Comments

[Factory]

"I don't see Mario; I see a cursor"
Hmm when Elite was being ported to the NES, they came up with a problem in user testing wherein the players (particularly young ones) were confused as to who they were playing in the game, perhaps the quote came out then?

[Jeffool]

>>(I just looked for that bit in One Flew Over the Cuckoo's Nest about how the bird at the head of the V can't hold that position forever and needs to take turns but couldn't find it.)

That would be geese flying during migration, and it's true that they often trade lead and split/join during the trip to curb exhaustion. At least that's the popular theory.

Here's the quick and dirty of it:
http://www.newscientist.com/lastword/article.jsp?id=lw827

Unless I'm just imagining it, I think I remember hearing stories about how this has been applied, with at least some success, to leadership roles in other industries as well. Though I wouldn't be the one to guess how it would work in any given group. I assume the more powerful geese lead more, and the weaker geese fall back more often, for the betterment of the group. Think people could pull that off?

[anon]

"Underneath all the pretty graphics there is a rule system. Blackjack or shoot an unaware opponent and they die; once they're aware, the blackjack does nothing, and the arrows hurt but do not kill. Not a very good simulation [...] but a good game, one that encourages stealth above all."

While a system like this could arguably make for a better game, I don’t think I would consider it. It simply defies ALL logic. A system where enemies are more vulnerable to attacks from the back than from the front is justifiable. A system where enemies “raise their shields” once they are alerted makes more sense. But there is no somewhat believable reason for Thief 3’s system.

It completely turns me off when a game imposes a “rule” on me that does not have a justification. It pulls be out of the universe, and reminds me that I’m playing a game. The least designers can do is come up with a paper-thin reason for their game design decisions.

[Chris Busse]

"It completely turns me off when a game imposes a “rule” on me that does not have a justification."

Fun gameplay is it's own justification. I agree, come up with something that makes "logical sense" if you can, but if you can't good gameplay trumps logic every time.

[J.]

That Thief:Deadly Shadows was so good after Deus Ex:Invisible War was so bad, is more than a little phenomenal. Then again, they had six extra months, and NO HARVEY SMITH.

For anyone who's got to the Cradle mission in T3, blame Jordan Thomas for that one. That guy's some kind of fuckin' game designer, there.

[anon]

Here are some examples of where illogical gameplay makes the game more fun:

- I can see the direction enemies and cameras are facing on my radar in Metal Gear Solid.

- There are Precursor Orbs sprinkled all over the world, and they happen to give me hints about where to go next in Jak & Daxter.

- I can carry a rocket launcher, sniper rifle, and submachine gun on my person at the same time and still be “sneaky” in Deus Ex.

Here are some examples of where illogical gameplay makes me want to chuck the controller:

- You cannot kill your fellow UNATCO agents in until predetermined missions Deus Ex.

- You can only aim in first person with certain weapons in Grand Theft Auto 3.

I’ve come to the conclusion that it is the inconsistency in the gameplay that makes the game less fun for me. The game teaches me to play one way, then for an otherwise inexplicable reason the “rules of the universe” are changed. I will take illogical and consistent over sometimes-logical but inconsistent any day.

If you have to change the rules on the player, try to come up with a reason for why they’ve changed. And if you can’t justify it, at least spell it out to the player. Otherwise you might be in chucksville.

[Jamie Fristrom]

I liked the Thief series better than the Deus Ex series as well, but I think that's simply because they focus on one kind of core gameplay, (and a kind of core gameplay I like...shooters tend to bore me) and Deus Ex suffers from having to do it all. But that's just me (and you): in the reviews Deus Ex is only lagging a point or two behind, which could be explained by random variation. So all it really means is Thief is more our cup of tea. Maybe that's because we're game designers. If only the whole world were populated with game designers...

The Cradle level was great for storytelling and atmosphere, but it was broken in a couple of ways: why would I go through the effort of collecting more toys when I have quicksave? (You could argue that excessive use of quicksave is meta-game, that I was cheating or being dishonorable, that the extra toys are just there for people playing with an iron-man/hardcore style (a concession to a meta-game) but it bothered me none the less.) Also, it was too ambitious for its budget - going into the past would have been So Cool if the geometry changed, or if once you were in the past you saw events play out that gave you clues about what to do in the future, but just seeing the same geometry with a different FOV and a green filter was disappointing. I suspect they wanted to do the full, cool thing and then scaled back because they ran out of time or hardware limitations messed them up. Still, like I said a few days ago, I'd rather see something ambitious done poorly than something stale done well. The Cradle level really was one of the high points of the game.

[Jurie Horneman]

Uhh... do tell me about that director protection technique please... I wrote about that issue here: http://www.intelligent-artifice.com/2004/06/bad_stress.html not that long ago.

[Jamie Fristrom]

In a nutshell, we're going to have two "creative directors", splitting the work that Tomo did all by himself on the previous project. Which is really the same thing that you were talking about: too much work? Delegate. Some warning bells might be going off, though:
- how are we going to divide responsibilities?
I'm not terribly worried about this; we'll draw up a czar chart and shift things around as necessary.
- who gets final say?
I'm not terribly worried about this either; it's pretty easy for two people to achieve consensus.

Oh, and thank you.