Beauty in flawed design or flawed analysis of complex design?

I just finished my yearly playthrough of Thief: The Metal Age. A game that, despite revolutionizing stealth, has no clones made and no worthy successors. It's a shame, and it made me overanalyze Thief again.

The thing about Thief is that, as great as it may be, it is an imperfect game. Some things are clearly easy to fix, like improving movement and mantling. But there are some fundamental design issues that are very complicated. First, it encourages careful planning; you need to get your bearings with a poorly drawn map, listen to footsteps through doors, pay attention to the briefing, in-game conversations and readables, use the scout orb to see what's around a corner, etc. This is all well and good. Second, it focus on a stealth experience, which means when you get caught the game design gets thrown out the window. I mean, if the guards know where you are it's not stealth anymore, is it? Now it's either a running away game or a combat game. Or you quickload to previous state. 

Let us look into two of these scenarios. Let's say you quickload when you get caught. Now you have privileged information about the mission, rendering the careful planning aforementioned useless and the experience less fulfilling. Why stand in a corner for 15 minutes counting footsteps to know how many guards are patrolling nearby, when you can just run, look for yourself, and load?

Let's say you stay and fight. Fighting isn't fun or rewarding or easy, and that is by design, as empowering the player would go against the philosophy of the game. Stealth is meant to be the player's super power, the dominant strategy. 

Since Thief might be less popular, we can use Sonic to do a similar analysis. A platformer that encourages you to go fast! And then put spikes right on your face because you went too fast. I mean, what is it supposed to be? A free flowing running game? a platformer?

These issues are apparently not a deal breaker to either of these games. They are revolutionizing games with a cult following, but the cracks are definitely visible! And they have no simple solution either. Sure, you can think of ways to mitigate the issues. For example, Sonic has the ring system, where if you get hit you lose your coins and all you have to do is pick a single ring to be able to get hit again without dying. As long as you have a ring you won't die. Thief mitigates those flaws by trying to empower the player in combat just enough to discourage loading. Giving tools like the flash bomb work as well.

What if instead of mitigating the design you try to actually fix it? I thought about this a lot. Turns out it's either very difficult or impossible.

Let's fix Thief by tweaking the design just a bit. Let's empower the player a bit more and make getting caught also fun and engaging instead of frustrating. Well now it's just Dishonored! That is exactly what they did. And as much as I like Dishonored, it's definitely not Thief, it is a different experience. Therefore we didn't fix Thief, we just created something completely new with it as starting point. 

Let's fix Sonic by removing the high speed sections. Same problem as before, we just turned Sonic into Mario. Let's fix Sonic by removing the slow parts and spikes along the fast paced sections. Well, ok, now you got something a bit more unique than Mario but not unheard of. There are lots of free flowing platformers out there. However none of them resemble Sonic in the slightest! So any attempts at fixing Sonic seems to yield anything but Sonic.

The question then is: Are these game designs unfixable? In this case we should embrace flaws and imperfections, as we have solid proof that amazing master pieces can only emerge when the design is not elegantly symmetric. Like a chord progression that ends in the wrong key but, hey, it sounds good!

That sounds so profound it is almost hard to deny. But I personally don't think it is true. I believe the analysis is just flawed, and these games are examples of more sophisticated designs and are hard to replicate because of it. One needs to look beyond the 'main game loop' to see the true design take shape. Is Sonic really encouraging and discouraging running randomly? or is it encouraging memorization of level layout through several playthroughs? Is Thief really incohesive in its encouraging of planning and subsequent encounters messing it all up? or is it forcing you to let go of your plans and asking you to learn how to improvise as well?

These are rhetorical questions, of course. It is hard to know what my own feelings are while playing. While I did feel frustrated and did abuse quickloading in Thief, I also have fond memories of moments that happened once I let go of my 'perfect run' and just improvised. The design might still be flawed, in the way that if that was the intention then it could have been made more explicit. For example, reward improvisation skills as much as stealth, and replace save/load with something more sensible. Or, in Sonic, give stats at the end of each level showing your time, compare it with previous runs, unlock levels to play any time once you reached it, etc.

These were not implemented probably because the main designs mentioned in this article are... well, likely accidental is my guess. These are complex and unusual games that came about from trial and error and deadlines. Thief started its life as a King Arthur game! I argue their design are fundamentally solid but too complex. This explains any roughness around the edges and why we don't see these games replicated as often as simpler but equally revolutionary games. Thief and Sonic stand solitary in the world of game design while Doom and Mario have a monthly quota of games iterating on their designs.