Is online play for multiplayer games all that important? Many of you would likely say “yes.” However, I’m not entirely sure online play is what we want. I think what we really want is a thought in our head that says our victories are worthwhile. After all, A.I. can be worked around. Exploits can be found and computers can be tricked. Beating a bot just doesn’t feel good, but beating a person means that you straight up outplayed someone. You are better than that person. Your victory meant something.
But, honestly, what evidence do we have for thinking that our online players are actually human at all? I mean, sure, there are chat rooms and voice chat and if everyone you ever play with has a mic, you could reasonably expect a person to be on the other end of the line. Heck, you can reasonably expect that there is a person on the other end of the line even if they stay silent. But do you have any actual proof that your opponent is real?
Consider the curious case of the Wii’s Smash Bros. Brawl . When playing with strangers online, you really had no control over how the game’s rules were set. You simply were thrust into a big free for all match and no one could interact with each other except through game mechanics. There was no voice chat, no text chat, not even player monickers over your opponent’s heads. Other than the latency, there was literally nothing in the match to let you know that you were playing online. In fact, in the early days, many people suspected that the game simply replaced anyone who would disconnect on you with bots and no one would ever know.
However, that didn’t prevent people from playing. For all intents and purposes, there was absolutely no difference between an online free-for-all match and a free-for-all match with bots, but people who absolutely swore off the random chaos and lame A.I. of single player modes would play in tons of online free-for-alls. Many people would say that it was the fact that their opponents acted like real people that drew them to the online mode, but even that does not guarantee that you will be facing a human.
Consider the case of the BotPrize competition. This is an entire competition modeled after the Turing Test. The Turing Test is a simple test for A.I. intelligence, and while it has been criticized many times, it’s still a very useful thought experiment. The idea is that if you cannot tell that an A.I. is not a human, then you have no grounds to say the A.I. is not intelligent. That being said, this competition isn’t made for making intelligent A.I.s, just A.I.s that can fool humans into thinking that they are human.
One of the best places to make A.I.s that do this is video games. Many FPS bots have been made to observe human players and mimic them, both their skills and mistakes. Several times, FPS A.I.s have managed to fool players into thinking that they are human. Still, what if you were given a choice. You could either play against an A.I. that perfectly replicates a human or play against a human. In either case, you would be secluded from your opponent, so you couldn’t interact in any way, other than through the game. That means no voice chat, no text chat, nothing but gameplay. What would you rather do?
If it all came down to gameplay, the choice shouldn’t matter, yet I’d wager most people would choose option B. Thus, I posit this: it is not the gameplay, not the experience of playing a human that is important, but the badge of honor of beating a human that is important. At the end of the day, we aren’t happy winning, we just want to make sure that someone else is losing.