Monday, May 25, 2009

Game Designer Dialogue

I had the privileged opportunity to speak with a young man who was thinking about becoming a game designer. What follows is an interesting exchange regarding Mount&Blade which I felt had value to other folks who read this blog. The young game designer graciously gave his permission to allow me to post our conversation.
----------------------------------------------------------------------
About my thoughts, I guess I will start with Mount and Blade since I've been thinking about it recently and the fact that we are on the site. I'm pretty sure lots of people would agree with me when I say that one of Mount and Blade's biggest flaws is its AI. Although the AI itself isn't anything terrible, its just that in a game like this, AI-only opponents, can make the even most emersive combat boring after a while. It eliminates the need for dynamic strategies and makes battles tedious and unrealistic. Although Warbands will include a multiplayer mode I still feel a bit unsatisfied as that doesn't really give any long term gameplay rewards. I don't really know how to get around this but maybe if the AI implements some kind of detection algorithm that will replace their apparently omniscient sight so that they react more like a human and maybe implement a collection of general strategies for them for each kind of situation (outnumbered - try to take on little parts of enemy at a time or hit run, player out numbered - stick together and attack at once or attack enemy from two sides etc...). The AI leader unit's tactics skill level determines how many different strategies they can use and how effectively they can pull them off.

On another matter, I believe that the Mount and Blade combat experience, although innovative and fun, is far from perfect. If the horses behaved better, the combat's experience will be a lot more fun. I was thinking of the horses in assassin's creed which were a blast to just ride (although assassin's creed's combat wasn't fun at all). I think that horse behavior from Assassin's creed combined with mount and blade's ragdoll physics would make the horse combat the most realistic and fun ever introduced in the industry. Also i think that the riding skill shouldn't increase the speed of the horse because that should solely depend on the horse itself but rather, the things the player can do on horseback (more effective attack animations, allow to use more weapons on horseback, more responsive horse and later on, warhorse tricks such as levade). As for unmounted combat, i think that proficiencies should also unlock more effective attack animations (faster swings, lounge attacks, jump attacks etc).

Anyways those are just my ideas to make mount and blade less repetitive and more emersive. I have no idea how possible those features are to implement because I have no experience or knowledge to evaluate the difficulty of those features. They sound pretty feasible (especially horse behavior since its already been done in a mainstream game)

-------------------------------------------------------------
My reply:

Thanks for opening up this dialogue. I find your insights interesting and insightful and look forward to where this goes.

When we look at a product such as Mount&Blade we have a natural tendency to look how to improve it.

We can list out many issues with the AI, models, strategic elements, story, and dialogue, which the list can go on for pages. This is both the blessing and the curse of being a game designer. When your work becomes your play, playing always then becomes work. We cannot enjoy an experience without at some point engaging our analytical game-designing mind. I digress however, my apologies.

The difficulty with fixing any game is to determine what is important to fix and why.

While I agree that the AI is not at the best it could be… is it the most important thing to address? The answer to that is up to the product owner and what they are trying to do with this product moving forward. This is an undeniable fact in our business, is that the business owners make the decisions on what to do, not the designers. We are given tasks to design out something that someone else has had the vision for. I know that this is depressing, but it is the way of things. For our conversation let us assume that we have the ability to make the overall design and vision decisions.

With this base assumption we have to embrace and understand what is happening in the market place and what is worth doing. This is the business aspect of gaming, which we need to comprehend and with which have some measure of fluency. When we make design decisions on this level we have to understand that if we spend 200K on development, it has to bring back at least 500K in revenue. If we fail at this target, then we have not done our job.

With all this in mind lets address the original intent of your thoughts.

I know that Armagan is going to work on the AI in Warband. This is good and one reason I have stayed completely out of this part of the code. Could the AI be better? Yes, but if I were to stack rank the problems of Mount&Blade, improving the AI would not be near the top. It would be in the top 10 things to do, but not critical.

Technically to your question: it is completely doable and it can be done in side the module system and outside of the game engine for the most part. It is just work to determine what we want to do and make it happen.

Your comment of “Although Warbands will include a multiplayer mode I still feel a bit unsatisfied as that doesn't really give any long term gameplay rewards.”, is your most important insight.

All technology, entertainment experiences, game design, etc always revolves not around the nuts and bolts of technology, but around people and their perceptions. It is about satisfying need. You have to think in terms of providing satisfaction. The question back to you is, does improving the AI, animation and battle improve the offering’s satisfaction level to the point you really want? My feeling is it does not.

In this case, and what I consider the most lacking piece of this offering in it’s native form is lack of context. The wrapping of why we are doing the things we are doing are paper thin and lacks the ability to suspend disbelief. In playing the game at this level, there is a definite lack of what we call immersion.

What happens then is very quickly we have engaged all of the possible enemies, and know what to expect. The game degenerates into a repetition of serial battles in order to “win”. The battles vary in terrain, as well as taking castles and towns, and we have to learn some elementary tactics that is challenging; and it is challenging to learn how to really fight in the game. But in the end, you are right; it is an exercise in repetition.

You are left focusing on the act of the conflict, the battle, since this has become the most salient aspect of the game and it follows then that we start looking to improve this most obvious aspect. The trick and power of the game designer is to see past what is obvious and ask questions about, you likely guessed this, what provides satisfaction.

Again I draw your attention back to your original question. Does improved animation, improved AI solve for this repetition and ultimately provide higher level of satisfaction? The answer I feel is no. It helps, but it does not give us the long-term replay value that we as players want from a product. All it does is make the repetitious battles better repetitious battles. The key is not the visual experience of how you fight, but what, why and the manner in which you fight.

The vast vast majority of games on the market today have this very same problem. Most offerings are much worse than Mount&Blade in this regard which by the way, is one of the reasons that Mount&Blade is so popular with the elite and hard to satisfy customers.

The key, as I mentioned before is a combination of identifying satisfaction factors and envisioning what is not present rather than working on what is already there and improving it.

For example: Imagine a mount&Blade game where you cannot see the entire game map, as only the places you have “mapped” are visible to you. A whole new dimension to the game has been introduced: exploration. Now add a variable where the game map is slightly randomized every game, making each game unique. We have just added strategic variance as well as a factor of uncertainty. Between these two factors and with literally just a few hours of work we have managed to improve the replay value of the game greatly.

I will stop here on this note regarding your initial thoughts. I am fairly certain this was not the expected direction you were looking for when you engaged me so it is ok to redirect the conversation in your reply.

----------------------------------

More later..

best,

Jim