I would make a case that Electronic Art’s EA Hockey released back in 1993 was one of the first sports games that replicated fairly convincingly the look and feel of the exercise they were simulating, and I do have some fond memories playing it on my Sega Megadrive (Genesis in the US). Later on, EA Hockey evolved into the famous NHL –series followed by other sports series as well, which have seen annual iterations ever since.
Indeed, release after release, player models, arenas, and crowds have become increasingly more accurate and definite in terms of polycount, texture-resolution and simulation of light. In fact, the visual quality in the latest installments of any EA Sports’ title is closing in photorealism on certain conditions, which is particularly true in broadcast –like camera angles. Yes, it’s not uncommon to hear people saying that while playing an EA Sports game, someone who walked into the room thought for a moment that the game was an actual TV broadcast.
However, when I was recently watching a trailer of NHL 12, one thing became painstakingly obvious, that resonates with something Lev Manovich describes in his opus Language of New Media as uneven realism. The gist of it is that the reality we occupy is infinite in terms of mathematical complexity, which has lead graphic researches to develop a host of unrelated solutions to deal with different areas of synthetic imagery.
The problem that rises is two-fold. Firstly, some portions of reality can be considered to be more complex to model than the others, and secondly, our human perception is tuned in to recognize certain traits of reality exceptionally effortlessly as artificial.
So, while the lighting and geometry were fairly believable in that NHL 12 trailer I watched, the movement of the players wasn’t so much. The fact is, the simulation of human motion is still in its infancy, even though EA has now implemented Euphoria –like physics engine to simulate body impacts to an extent. What makes the situation especially complicated for sports games is the fact that they are more often than not based on behavior of human body. And there’s nothing more difficult to simulate than the human muscle and nervous system, since we, as human beings, are conditioned to spot any flaw or inaccuracy in how a human body works.
Yes, the disconnection between how, for instance, NHL 12 looks on the surface and how it behaves is enormous, and I think the gap has nothing but increased over the years. And continues to increase if something radical doesn’t happen in the field of (human) physics engines, which is unlikely.
Interestingly, the problem in hand is virtually absent in games like Gran Turismo 5 or any other vehicle-based games, since the physics of, for instance, a racing car are fairly easy to carry out with a simple mathematical simulation – or at least that’s what our biased human perception lets us to believe. Indeed, if we were cars like in Pixar’s Cars instead of humans, we would (or at least I would) moan how unnaturally cars behave even in the most sophisticated racing games.