Timing is everything.
Today Raph has a post showing a 3-part “History of MMOs” video. (well worth watching btw, especially for those who started playing post-2004)
Also today Tobold has a post about how bots could easily play certain MMOs better than players.
In the video, the narrator credits WoW being more linear and accessible as a major source of its success.
The more linear/accessible your game, the easier it is to create a better-than-the-player bot for it.
The… oh, mild connection between ‘dumb as bots’ gameplay and ‘mass market’ is hopefully not lost here.
This of course is not entirely negative. WoW is/was, after all, a great ‘intro to MMOs’ game for many. Whether that same crowd takes the next step into ‘real’ MMOs is up for debate. Certainly a title like SW:TOR is not helping people take that next step, but on the other hand SW tanking BECAUSE it’s an entry-level title in a market of vets (I use that term very loosely) will do some good. If we take one step further, buy into the hype, and assuming GW2 is indeed an MMO that fixes all previous MMO woes while not being a ‘dumb as bots’ title, and it’s successful, then we (MMO players) all win going forward.
Or you continue to laugh/cry at the genre while FiS.
Hopefully both.
Hardcore Casual 
Guild Wars 2….why can it not already be here! =(
Thanks for the link to that video (which I would’ve missed for sure).
If anything, to the extent it’s even technically an MMO, GW has proven that other human players aren’t necessary at all. or even all that desirable. I don’t see that GW2 is going to deviate from that formula, though I could be wrong. So long story short, if requiring players to outperform and thus be more desirable than AI is your yardstick for progress, then I wouldn’t look for GW2 to carry the genre forward all that much.
Whatever the case, though, I am definitely looking forward to it.
I’m gonna be honest, I was looking to make a point, and needed an example, and the best I could come up with was GW2. AAA ‘MMO’ genre sucks.
I assume you’re referring to the fact that in GW1 everything except outposts is instanced, and/or to the Hero & Henchman mechanics that allow you to use a full party of NPC allies. (FWIW, the result is a poor MMO but a pretty complex and enjoyable party-building game.)
That’s not the GW2 formula, though. GW2 has a shared world, there’s no H&H mechanic, and the designers have really gone out of their way to make sure working with other players is always a positive thing. Combat benefits from cross-profession combos, events scale up with more players present, nobody can leech your XP or compete for your resource nodes, etc.
Yes, fanboy much. :)
Kinda interesting to think about whether this game (or any other) truly REQUIRES a human rather than a bot, though. For that you need gameplay based at least partly on intuition, which is basically instant recognition of extremely complex patterns, rather than calculation or reflexes. Chess grandmasters can’t beat computers by “thinking 20 moves ahead,” but they can do it by leading the game down paths where calculation is less valuable than insight.
Eve is the easiest MMO to create bots for. Even as a non-programmer it’s simple to record a macro to mine 23.5 hours per day.
Yeah, I found the sentiment that EVE somehow escapes easily bottable gameplay to be quite bizarre.
A bot that can mine better than a human, sure. But a bot that can do combat in Eve?
Dream on.
Why?
Since Eve automates much of the players movement (orbit at …), all the bot would need to do is identify the correct target and activate the weapons/webbers/etc.
If anything, I would argue that a bot for Eve PvP would perform better than one for most other games.
Tobolds message was that a bot can outperform a player in ANY game (given enough information) and he is right.
@Reaper
Advanced level PVP requires a hell of a lot more than the things you describe, particularly small gangs. Sure, you can design fleets that would be just as good or better in the hands of bots but most fleet compositions simply require too many situational decisions and too much intuition to automate (like nano and kiting gangs).
Even on a small scale, there are a lot of decisions that require intuition. Do I kill drones first or go for the ship? Do I attempt to hold range on my target, or close in? What ammo type do I use? Is target A running fit X, which I can easily kill, or fit Y, which will slaughter me? They’re all decisions that require a lot of intuition, from knowing the enemy pilot or corp, to knowing the common fits, to comparing the speed the ship is going with it’s base, to subtle differences in the way your enemy is flying and acting, and more.
Besides that, most PvP in EVE is about shaping the battlefield and the outcome is often determined before the shooting starts. That’s something completely beyond what current level AI is capable of.
Bots would be useless in EVE combat, they’d simply be baited and blobbed and slaughtered. Or people would find their weakness (tacklers let themselves get isolated, they can’t handle nano fleets that duck in and out of range, etc.) and pick them apart.
The point is, modern generation bots require having complete information. They know all the possible factors that can affect them and all the possible actions they can perform in response.
Chess is an incredibly complex and difficult game that most people just aren’t any good at. And yet, it can be ‘botted’ relatively easily and be far more effective than all but the very best grandmasters (if that, now that computation is getting so cheap). Walking outside, on the other hand, is something nearly every mouth breathing asshole can do with some competence – but it’s something which has proven extremely difficult for AI.
EVE sandbox style makes that impossible. The possible actions and reactions are essentially limitless, and we simply don’t have AI that can handle that yet. It’s the same reason any strategy gamer knows the AI cheats – it’s the only way they can compete once things get too complex to script. It’s the reason the stock market isn’t (completely?) dominated by AI ‘bots’.
Just to add on here, there is a major difference between botting one side activity (mining) vs intentionally limiting bots during your main game (PvE) due to the possibility of outperforming humans.
Just to add on here, there is a major difference between botting one side activity (mining) vs intentionally limiting bots during your main game (PvE) due to the possibility of outperforming humans.
Are you suggesting EVE bots couldn’t run PvE missions? Or provide target-of-target support of human players, potentially to a greater degree than another human player doing the same?
All the PvP examples are irrelevant considering a PvP bot wouldn’t function to any high degree in WoW either, for the same reasons. And let us not forget the difference between a 3rd-party bot you download off the internet, and companion AI programmed by the designers themselves (and made immune to AoE, etc).
The original question was whether bots could OUTPERFORM humans in a core activity (in a themepark, PvE, especially instanced ‘group’ PvE). A bot in EVE could not OUTPERFORM a human in the core activities of PvP, Econ, or PvE (PvE because of mission ganks and wreck stealing).
Could a bot be somewhat decent in those areas in EVE? Sure. But OUTPERFORM a human to the point of wanting a bot in your fleet over another person? Nope.