The Value of Good Bots

In an interview with Jim Brown, lead level designer for Unreal Tournament, Jim made the comment "Statistically speaking, more people play offline than they do online"(source: http://www.1up.com/do/feature?pager.offset=1&cId=3140119).

To many people, this is mind boggling. With other games it might be more understandable, but this is Unreal Tournament, a game that appears to have been designed entirely for multiplayer with a couple single player modes (which basically mimic multiplayer but with bots) thrown in at the end.

Likewise, in the DotA official forums, there is a thread for the AI version of the map. There is little discussion there (mostly just "When is the next AI map coming out?" posted over and over), yet the thread tends to range from 50-200 active viewers at any given time. As there's no real reason for more than one person in each group of friends to visit the thread for more than a few minutes per week, mathematically, that means that there are a TON of people who want good, up-to-date bots to play against.

Like Unreal Tournament, DotA was made as a hardcore multiplayer game with no real story, pre-scripted events, or single-player modes. So what makes so many people want to play against bots instead of players?

Definition of "Bot":
For the purposes of this article, I'm going to use the term "bot" to refer to any time the computer fills a role that can be filled by a player. Thus, it would include the skirmish AI on an RTS but would not include the scenario AI.

So why would people choose to play against bots?:
There are, of course, the obvious answers of practicing for multiplayer, slow internet, and experimenting, but these don't seem sufficient to account for over half of gaming, or really even close to it. Knowing quite a few people who prefer to play against bots and spending time on a variety of forums to attempt to research why others do, I've feel like I've come up with a pretty good list of why:

1. Some people are not competitive: Too many people make the mistake of equating competitive to hardcore. Many people will dive deep into a game and love a good challenge, but they have no desire to prove that they are better than anyone else. For these people, a major part of the appeal to competitive multiplayer is gone.

2. Playing against bots allows you to control your experience: How much of a challenge do you want? If you lose, do you want a rematch (or as many rematches as it takes to win)? Do you feel like fighting more difficult opponents of matching numbers, or would you like to play against hordes of weaker opponents? Do you need to pause frequently because you're multitasking? Are you tired of fighting the latest "OP" faction or character and want to mix it up?

Online, you have no control over any of this. Some, you'll almost never get online (the option for frequent pausing or a few good players against hordes of weaker ones), and the rest are just a shot in the dark. You might have a good challenging game, or you may go up against an opponent/team with a vastly different skill level than your own (even good matchmaking doesn't always stop this). On the other hand, you may want to leap into a challenge that's way over your head to see if you can rise to it, and the matchmaking will stop you (and your opponents may be bored even if you do manage).

3. Offline play often provides more variety: Before you argue with this one, consider how many people are out there who do nothing more than repeatedly use the most popular tactic. Whether it's the current "overpowered" weapon in an FPS, the unbeatable combo in a fighting game, or the standard rush tactic in an RTS, I find that online games are often far more predictable than those against bots. Sometimes that makes easy games (if there's a reasonable counter to their overused tactic) or really hard games (if there isn't), but it gets boring either way.

Admittedly, this depends heavily on the specific game, patch, and community, but a complete lack of variety is the main reason I quit playing Dawn of War online. Too few people were interested in anything but the current "win" tactic.

4. Some people just aren't that good at games: Whether because they don't play them as much or because they simply don't have the timing or finger dexterity, some people just aren't as good. With a few exceptions, even the weaker online players are quite competent at the game. If you're one of the many people who will only be spending 2-5 hours per week on a game, you'll probably never be good enough to compete online.

5. Online lobbies, jerks, profiles, droppers, hackers, laggers, and other such can be more trouble than they're worth: This almost made me quit League of Legends, and it did make me shift to only playing with an entirely premade group of friends, which dramatically reduced how much I play, and thus, how much money I'll spend. I have no interest in being on a team with jerks. It sucks to be 30 minutes into a good game and have somebody drop. There's just too much burned time around the good games.

Why aren't these people more active on forums?
This is a fair question. A game designer trying to get the pulse of a gaming community is unlikely to see much demand for good bots when browsing forums. It's still there, certainly (just google "Killzone 2" and "Bots" and you'll see a ton of positive comments all over the net from the announcement that the game would have them). However, they are certainly less vocal than the competitive community. There are a few reasons for that:

1. People fighting bots aren't nearly as interested in balance: Quite possibly the most heated topic on gaming forums is balance. Bots typically aren't very good at exploiting, so a tactic that is a bit overpowered is unlikely to make much of a difference for this type of gamer.

2. People who prefer playing against bots are ridiculed online: Go check out some forums, and when you can find somebody who admits that they'd rather team up with their friends against bots than play random people online, you'll usually see insults or derision in many of the subsequent posts. Considering the fact that many people choose to play offline to avoid these people, it makes sense that they wouldn't choose to seek out their company in forums.

3. People who play offline are often less internet savvy or have slow connections: This one should be pretty self-explanatory.

4. Bot-fighters tend to be the type whose gaming life is with real friends, rather than random people online: If your playing and discussion is only with people you know, you are less likely to post on forums.

Wouldn't these people get more satisfaction playing dedicated single player or co-op games than a single player version of what was inherently a multiplayer game?
There is a philosophical difference between designing a single player game and a multiplayer. In short, multiplayer is designed to be a game and single player is designed to be a story.

Look at some more traditional games (chess, football, etc.). In each of these, the game is nothing more than a set of rules that allow a lot of room for developing your own approach and style. The creators of these games weren't trying to create an exact experience, but rather, they give you a context to create your own.

Single player video games, on the other hand, take a very controlled approach to the player experience. Designers of these games typically seek to control what the player is feeling at a given time, what tactics will be good or bad for particular situations, difficulty at any given moment, and how much the player is capable of achieving. In fact, if a player comes up with a good enough strategy to interfere with this experience, it will be labeled a "bug" and "fixed" in a later version (if the platform allows patching).

I'm not saying that one of these is better than the other but that they appeal to different audiences. While some people find that a heavily scripted and controlled game creates a more enjoyable experience, others are simply frustrated by their lack of ability to make meaningful choices. This second type will typically get more entertainment from either fighting against bots or playing through less scripted game modes (i.e. I found Horde mode in Gears of War 2 to be far more enjoyable than the campaign).

Try it!:
If you still think that bot fighters are missing out on the good part of the game, try giving yourself a wild challenge and see if you can rise to it. Can you and one friend beat 6 normal computers in Warcraft 3? How about 8? 10? (It can be done).

Taking on crazy tasks like this will reveal depth to a game that you never thought was there. You'll find yourself inventing bizarre tactics and experimenting in ways you never would have thought to try in regular multiplayer matches.

I've seen avid competitive gamers have a blast fighting AI opponents when the challenge is great enough.

Conclusion: Of all the ways to expand the audience appeal of multiplayer games, creating high quality bots is probably the best return on investment available to you. No new graphics, physics, landscapes, and most importantly, no features that diminish the experience of the core gamer. Unreal tournament and DotA have proven that there are tons of gamers out there who want good bots. It's time for other games to take advantage of this.