Briareos H on 12/6/2013 at 16:21

---

I was thinking of writing part of this in one of the debate threads, but I think they all have gone overboard with insults by people from both sides who don't know how to debate. Better keep it where it is sane. Play nice.
So, from what we have seen, it's time for a heads-up on what we know has been kept and what has been left out or radically changed of all the innovations Thief brought, also identifying what we cannot clearly pin for the time being.
I'll start as objectively as I can with the list, before going on to discuss a personal opinion that tries to take into account factors which were argumented by both sides of the pro/against Thief 2014 debacle.

Please correct me when I am wrong.

Thief 1

* Even more non-game related objects can be picked up - This was confirmed when the player in the second demo picked a bottle to throw it as a distraction.
* Engine-generated shadows as a visibility moderator, calculated in an analog way - Light and shadows are indeed a visibility moderator. The lead level designer says: "it might look like a binary system but it's not, it's a very complicated sensory system".
* Game elements as an audibility moderator (footsteps on different surfaces, noisemakers, broken objects) - Noisemakers and decoy objects are in. Sound affects AI. In the second video, the lead level designer confirms that surface is very important for audibility.
* Realistic sound propagation - Realistic sound propagation is possible in UE3, but it remains to be seen in the demo.
* Tools as a way to dynamically change the environment (create shadows or distract AI) - Noisemakers and water arrows are in, they are sufficient to tick that box.
* Light gem allowing to evaluate said visibility (analog) and audibility - Very surprisingly, the lightgem seems to have only three states, which is hardly analog. There is a bright flash when the player goes into light and a very dim intermediary. This means that, while the game actually implements detection states properly, the analog information is not exposed to the player. In addition, I wasn't able to discern an audibility gem.
* Limited, but realistic physics engine - UE3 implements a modern physics engine.
* Lockpicking - Lockpicking is confirmed.
* Unobtrusive, minimalistic interface with 3D elements - As shown in both demos, the interface can be all over the place. Experience points indicators, togglable alert levels indicators (togglable, because they aren't in the second demo), takedown indicators, combat and focus interface. The shroud (fog applied to vignette the screen) indicating how well you are hidden in addition to the lightgem is anything but inobtrusive. Third-person segments are limited to specific climbing sequences: a third-person view is a new and unnecessary layer of interface, a scripted sequence separated from the main game is another layer of interface (for lack of a better checklist item when I wrote this).
* Loot count (inherent replay value) - There is a loot increase when stealing stuff and pickpocketing, so there surely is a total loot count
* Object and ammo store between missions - This wasn't shown.
* Complex and dynamic AI reaction scripts to the environment - Confirmed, AI react dynamically to the player and the environment.
* Scripted and unscripted alert levels and between-AI conversation - Confirmed, AI react dynamically to the player and the environment.
* AI factions - It remains to be seen whether there can be more than player / neutral / enemy factions.
* AI knockouts - Yep.
* Carrying and moving knocked-out or dead bodies - Dragging and shouldering bodies shown in action.
* No requirement (with rare exceptions) to kill or even engage AI - Ghosting confirmed.
* Wide range of safety, danger and failure states with several degrees of allowance, playing on the edge by design - If we go by what the guy says, then yes.
* Complex, story-advancing objectives that can be tackled in non-linear order - Remains to be seen.
* Minimalistic approach to providing pre-planning, guidance and information, focusing rather on a dynamic understanding of the level - Can't say for sure, but the "find out more in the library" that appears when you enter the heist location points towards more directed guidance and always being able to know where you are.
* Additional objectives depending on the difficulty level (inherent replay value) - Remains to be seen.
* Completely player-driven skill advancement - Maybe the most obvious departure from Thief, there is now an experience system for skill advancement.
* Mantling and vertical movement through rope arrows and climbing - There is mantling and rope arrows (in first-person!). However, the rope arrow was used on a very obvious attachment with white markings. Whether it can be used on any wooden surface remains to be seen.
* Large, architecturally-realistic levels - The level shown was very believeable, how large it is remains to be seen.
* Complete freedom of movement and alternate routes in a closed level environment (inherent replay value) - Alternate routes are confirmed, and closed levels are as well. Scripted escape sequences in third person with maybe QTEs seem to be in from another preview source, limiting the player control over his avatar.
* Disinterested hero with efficient characterization - From what was introduced, the hero is far from disinterested and his lines in the teaser trailer are hardly efficient, he is confirmed to be more talkative.
* Extremely strong audio direction, excellent voices, excellent music and atmospheric auditive backgrounds - The perception of "extremely strong audio direction" is slightly subjective, and cannot be judged from a video displayed at a loud booth.
* Dynamic weather - The visual direction for weather is strong, and there are lots of fog. Will there be rain and snow?
* Movable light sources - Dynamic lighting engine.
* Dark universe in a believable city - That's for sure.
* An unusual take on steampunk where technology is crude and cold but always retains an element of magic, copper doesn't feature prominently - No steampunk copper, but then there isn't a lot of technology shown besides light fixtures and a few buttons. Still OK, but no element of magic to be seen in the visual design.
* Feature-complete level editor (inherent replay value) - There are no plans for a level editor.

Thief 2

* Cameras and turrets functioning as AI - I think I heard that turrets were in. Tentative green.
* Even larger levels - N/A
* More complex scripting - N/A
* More complex objectives, such as tailing someone - How creative the objectives are remains to be seen.
* Dynamic, moving lights that affect visibility - Dynamic lighting engine.
* Fogging - This is Fog 4: The Fog Project

One point I forgot in my initial list (I'll add it back later) is the supernatural AI (zombies, ghosts and burricks) which all might out, burricks being confirmed not in. It was hinted, however, that the supernatural will play a role as the story progresses.
One other point which isn't in the list, because it wasn't an innovation, is the abundance of readables and AI conversations. Both are confirmed, although how prominently the readables will be featured remains to be seen.

In addition to all this, there are new features. In my original post, I was vehement to say that innovations were necessary to honor the LGS legacy so, what are those innovations?

Thief 2014

* Fast-paced moves for quick transition between shadows: dash and swoop. - I am not sure of what the swoop exactly is. The dash is a fast movement between shadows, in first person. It doesn't impact negatively previous elements, but bears similarity to Dishonored's blink. A true innovation?
* Combat assistance and contextual help as part of the game universe (through The Eye): focus mode. - Focus mode interacts heavily and negatively with previous elements. It is however totally optional. In actuality, it really is less of an innovation than a set of game mechanics extremely popular with the current generation of games.
* Thick, fog-heavy atmosphere. - Purely atmospheric and visual, so has no impact on gameplay.

So what does this all mean for the different kinds of Thief players that we are?

Let's start with the good stuff. Many game mechanics, the available tools of the trade and level configuration seem to be in line with what to expect of Thief. Usually, movement from A to B will be done by navigating the shadows in first person, on the edge of being detected, pickpocketing, blackjacking or ghosting guards all along, grabbing valuables, finding alternate routes towards an objective. If this is enough for you to define Thief, then look no further.
On a more personal level, I quite like the idea of an additional dash movement: it is not too complicated an addition controls-wise, and can add interesting dynamic elements to a mostly static situation. Increased agility is not a requirement, nor is it fully desirable, but a simple dash move strikes a good balance of increased player agency over the game while remaining vulnerable.

Now, for what I think needs to be changed, in other words arguing why I think some of the elements that have been left out made Thief. This is a subject that has barely been touched here recently, with only the draft of an argument by a couple people who opposed the new game, immediately shot down by... let's say people who don't get the whole picture. Those two points are minimalism and player agency. There are several instances of LGS developers acknowledging those two core concepts in terms of game development, so I'm not pulling anything out of my ass. The two concepts work together and, in fact, merge perfectly in the case of Thief in a way that was hardly ever done before or after, forming the backbone of the game's immersive nature.

Minimalism in gameplay, different from visual minimalism, is -at its very core- the idea of reducing the number of layers of interface between the player and the game. It means giving the most direct access to the game systems and the least number of abstractions. Everything in Thief is geared towards minimalism: the controls are simple and constant and static, they do not change upon context, because a contextual action requires a new layer of abstraction, often explained through on-screen messages. You move, you frob, you strike and use your tools and, provided you have those tools available at a given moment, the way you use them never changes. The user interface is small, very well spatially-defined (down left and right corners) and hidden most of the time. There are no confirmation popups, no menus during a mission. Provided you have seen one already, and bar a few exceptions, you can instantly recognize an object which can be frobbed. The lightgem reflects precisely how the game calculates your visibility, it doesn't tell you: "all things considered, and trust me, you probably are visible right now" because it's not the game's decision to make, it's yours. In other words, because there are so few layers between the player and the game, the player gathers information directly from the game of their own volition and reacts accordingly using the most straightforward interface (once they know they keys) ; the instances where the player is told something directly by the game ("you earned XP!") while playing are extremely rare if not non-existent. This doesn't happen by chance, but because that way your brain won't even have to think about the game itself.
The less time your brain spends on thinking "can this door be frobbed?" or "how can I maximize my XP gain when taking out this guard?", the more time it spends role-playing Garrett and feeling inside the game. To quote famous critic H.S. Plinkett, "you might not have noticed it -- but your brain did".

Then, there is control, or player agency, as the second facet of immersion. Non-contextual actions, the absence of in-engine cutscenes or forced third person view have one reason: in addition to reducing the number of abstraction layers, they never shift the control away from the player. Control is a wonderful thing and our psyche is wired to feel frustration and injustice whenever it is taken away. After the first level of Thief 1, the player establishes that their control is constant and unwavering from the moment a mission starts to the mission complete outro. This helps immersion as a safety net which guarantees that nothing will prevent you from enjoying your immersion in the way that feels habitual and natural. Another point is that the game never changes to accommodate the player, but the other way around. The player becomes better at Thief because his control is more precise, because he knows what tools to use and when (so he has spared them), because he knows how to analyze better what is happening in the game world, and he learns that by practise rather than by repeated tips.
Considering that there is very little margin for user input error, so an action is always properly interpreted, that makes the control over Garrett feel transparent (minimalism) and absolute (player agency). Both conditions for a fully-working avatar are met, immersion is complete, you are not playing a video game anymore. And this is exactly what makes playing Thief so special, what could turn the new Thief game from a good video game to a good Thief game.

Here is my advice to do that from what we know. It was confirmed that many on-screen indicators can be turned off. Make it all of them. Do not remind me on-screen that I need to press T for a takedown. Do not flash how much XP I won in front of my eyes (transparent interface → gamified interface). If XP must absolutely be there, then never (ever) spend it in-game but instead do it at any time I have established that the game is not. Do not use XP to augment the main character's skills (absolute control → relative control), use it to augment his tools only. Do not create arbitrary layers where control is taken: be uncompromising on the absence of quick-time events (absolute control → contextual and impoverished control), never show a third-person camera (transparent interface → gamified interface). Never give me gameplay advice while in-game. Do not interpret my visibility and feed me a synthetic result (transparent interface → gamified interface), give me the raw information and let me decide for myself. Do not force a decrease in information displayed on the screen at the times I most need it, especially not with a new redundant interface: the shroud (transparent interface → gamified interface). Finally, let me use my rope arrows on all wooden surfaces, as it was established before ; Do not, for the love of all that is holy, succumb to the trend of marking very specific rope-able places with some white paint, because that's the game telling me what I can do (transparent interface → gamified interface).

I can live with an unlikeable Garrett because, in my case, he is not as much as the third facet of immersion as he can be with others. I can live with less magic in the world of Thief, with more grit and realism, because it's an artistic vision that I can accept in the case of a reboot. But those above, those core mechanics are what's all the magic of Thief to me.

To conclude, a word about some of the undecided stuff. I'm fairly sure that EM know how important sound design is, so I'll just wait for more information.
I cannot, however, stress enough how important it is that the player is not fed the information but acts voluntarily to gather it. Design your objectives around that, in the case of the demo, don't tell me to look for information in the library if this wasn't established already in the initial briefing and objectives.
And as a final open letter to EM, no level editor means no legacy for your game, people will still play Thief 1, Thief 2, Thief:DS and TDM missions in 10 years -- this was the opportunity to future-proof and consolidate the fanbase for the future. I know it can be hard to do but if your game is Thief, it deserves an editor.

Judith on 12/6/2013 at 16:44

---

Nope, watch one of these videos carefully.

(http://www.youtube.com/watch?v=XMqdNzCNl5Q)

(http://www.youtube.com/watch?v=U0KS2Hm6_oM)

The transition is pretty fast, because he moves quickly, but the gem indication is not binary. You can see different states of that icon. Watch it frame by frame if you need to.

Also:
[Edit] Nickie was faster ;)

Judith on 12/6/2013 at 16:54

---

Designer said it's a binary system? :O

(http://youtu.be/XMqdNzCNl5Q?t=3m9s) from this moment to 3:15 it has 3 states at least.

[Edit] Yup, although I thought I'll see more states, I hope it's a matter of lightning, not the detection system...

nickie on 12/6/2013 at 17:04

---

I see you've amended the drag body, but I still think I saw a body carried. I also think someone here mentioned that. I'll see if I can find it.

Briareos H on 12/6/2013 at 17:09

---

fuuuuu***k
(shows how much i pay attention)