It seems like each sport is different in how close they can get to emulating the realism you see on your TV screens and from the stands. Obviously, the more moving parts there are, the tougher it is to emulate the realism. I also believe this is why you see some games achieve greatness and others simply fall flat.
So what four sports are the absolute toughest to emulate in a sports simulation from the designer and programmers point of view? Let's find out...
4. Basketball - The thing with basketball is that there are 10 moving parts on the court at any given time, but those parts should have to be considering a lot of things at the same time. Do you roll over for help-side defense or stick on your man? Do you screen at the top of the key or do you cut to the basket? Off the ball interaction and the overall physical-ness of the game make basketball an incredibly tough thing to simulate.
3. Hockey - Hockeyis an incredibly fast sport. As such, with so much fluidity it's tough to get the AI to completely mimic what should be going on at all times on the ice. While the lack of a huge amount of players does help from a programming standpoint -- the sheer amount of motion and different aspects each player on the ice must mimic (much less the goalies and save logic) at all times is astounding. To correctly mimic hockey takes a lot of effort.
2. Olympic Sports - Olympic games have been, for nearly two decades now, running from the same playbook. Whoever mashes the buttons fastest wins. Whoever mashes slowest loses. But that's not how it's done in the Olympics -- and video game developers can't seem to figure out how to mimic the events that are unique to the games properly. Unlike the rest of the sports here, this is a control and design issue more-so than an AI issue. To be the best high jumper on the planet means you have to time your jump perfect, you have to have your full body in sync, and you have to have trained to have you physically in the best shape of your life. That's a little tougher than pushing the right analog stick to mimic.
1. Football - 22 players, each with different assignments and goals who must react to what the other 21 players are doing in order to correctly mimic the sport. Oh, and all 22 assignments must be carried out in what is likely the span of less than one second. The average football play is six to seven seconds long, and the majority of the decisions you need to make are made within the first second or two. And then, on top of all of that, there is a very physical side of the game which you must correctly model. So you not only have to program a complex AI that is more complex than any other sport, but you have to program a physics engine which requires more depth and detail than any other sport. Good luck with that.
So what about you OS'ers? Which sports do you think it'd be toughest to program for to emulate realism?