Focus’ Twitter account posted the image below yesterday. I’m equal part excitement and sadness. Yes, I “love” me some Blood Bowl (even though I think the teams need an overhaul). But, could there be any worse developer than Cyanide? They can’t design GUIs, their games are loaded with bugs, they make half-baked games, and have I mentioned their poor interfaces?
I have this odd feeling we may see a mobile version of BB or possibly one that is also for consoles. Neither is good news for PC users. I wish I didn’t feel this way but they have burned me with every BB release and pretty much all of their others games too.
I enjoyed Blood bowl, but like many others, I’m not fan of the unbalanced team, despite the fact that the game was ‘designed’ that way. I would enjoy it more if it got the dawn of war treatment. Keep the theme and spirit, but recreate a game that can take advantage of the benefits that a computer had to offer.
The existing game is already about as faithful a remake of the Board Game as is possible on a computer.
Of course, it is quite possible to improve the user interface, but firstly, there is no way that can be sold as BB2, and secondly, I have no belief that Cyanide is capable of handling that given their track record.
Which means - no doubt - that BB2 is going to feature Cyanide’s own branch of “improvements”…
I’d just really love to see a more robust single player experience this time. Let me set up a full sized league and play out an entire season against a hopefully improved AI.
If Cyanide is doing this, I’m probably out automatically. What BB needs (since as strategy points out it’s already fairly faithful as an adaptation) is a good AI for each team and each teams sub-strategies (ideally) as well as a robust single player experience, such as with leagues full of ranomly made (or a mix of random and manually created) teams battling it out with yours but this time actually simulate the damn games, and make the experience as a whole very entertaining. Then I’d be in. But if Cyanide is doing this, I can’t imagine it turning out well.
Also, that UI needs some sort of overlay or something to help inform the player about things like dodge rolls needed to move to a specific square, stuff like that. The visuals were never a weak point in the game, but everything can be improved upon. I hope it’s something to look forward to, don’t get me wrong, and I’ll even keep an eye on it, but for the moment I’m leery.
This sounds aweomse, I really think the game needs a graphic, animation and general usefulness update.
Its never really been a good single player, but I can’t care less, however I do understand if you buy it for that purpose, you should get the goods I guess, but you ARE MISSING OUT on the awesomeness that is MP.
I can tell everyone now. If you’re hoping for human competitive AI you’re going to be disappointed. Bloodbowl was designed as a board game, not a computer game. The ruleset is incredibly difficult to program an AI for.
A computer is very good at some tasks that are difficult for humans, and vise versa. Getting a computer AI to play Bloodbowl properly would be a very expensive proposition for any game studio, not just Cyanide. It just makes no financial sense.
While the conclusion is obviously true, I don’t think there is any intrinsic reason why BB would be more difficult for an AI than chess or Go. It’s just not popular enough to merit the same level of research and investment.
While my AI coding experience is limited to 1 semester dabbling back in uni, I still think Bloodbowl is inherently much more difficult to code AI for than chess.
The rules of Bloodbowl are more complicated.
There are far more exceptions to each rule.
The sides are asymetric.
There are many more teams and units.
There is randomness.
There are many, many, many more possible moves on each turn.
And then, as you say, theres this:
Chess has a large organised library of theory.
Chess has a long history of AI programming to learn from.
Lots more people wanted to see a working chess AI.
Again I’m no expert on game development, but my experience is that even in AAA games the AI coding is done by either 1 programmer or a very small group of programmers working to limited time and budget. I think its insurmountable.
I’d love to be proved wrong though, because I’m a single player gamer.
Uncertainty (randomness) makes all the difference. You just can’t compare any moderately complex game with uncertainty to something such as Go or Chess.
That being said, it should definitely be possible to develop something significantly better than what Cyanide has made up. At its core, there are some pretty basic “rules” that are smart to follow when playing BB (low-risk moves first, always place a man into TD range on last turn, etc), and the AI fails to adhere to any of them.
Yes, but I’m not convinced this puts humans in a favorable spot. Humans are notoriously terrible at dealing with probabilities and complexity. They cope with it by following a few intuitive guidelines, as strategy said. Yet complex systems like BB often have equally complex but totally unintuitive solutions. I wouldn’t be surprised if the dominant BB strategy was some incredibly boring and absurd-looking pattern that no human mind would ever consider.
I think you’d be surprised. The one thing human brains are incredibly good at doing is rapidly pruning bad choices. Our brains can parallel process by applying simple heuristics at a rate that puts even the fastest computers to shame. This doesn’t always lead to optimal solutions, but it does quickly get you to the point of making good and sometimes great decisions, and in a game like Blood Bowl that ends up snowballing. Blood Bowl has a tremendous number of options, but the probabilities behind each of them are actually quite simple and can be gauged relative to one another easily. Humans shine at this sort of task - often you don’t even realize just how much cognitive capacity you’re wielding, but consider that when you look at any given board position in Blood Bowl there are probably tens of thousands of possible moves, but within seconds you’re usually down to debating between less than a dozen; how did you know to disregard those other possible moves so quickly?
I feel like Chess is the example that counters this. There are far less moves, but the end pool of “acceptible” moves is probably the same. In chess single mistakes, even after you pare out the obvious bad choices, can still be crippling. In a game with chance, a mistake can be negated by a “bad” roll by the opponent making the optimal play. And that’s before you include things relative to random factors like confirmation bias and the gamblers fallacy.
I do think the biggest reason is that nobody cares enough to do it. Unless someone tried, we don’t really have any idea how hard it might actually be.
