Reverse Gamification - Manzikert 1071 AD

Alan, my favourite game of all time is F355 challenge, which was released on the Dreamcast games console (a fantastic console before it's time, much better than the original playstation but which unfortunately wasn't marketed properly). The Game was basically not a game but as close to a driving simulator as you could get. The lead developer Yu Suzuki, is a massively successful Games engineer, who had built such a reputation and following, he was able to commission the development of the game as a faithful labour of love and physics simulation with virtually no "dumbing down" to suite a gaming audience (he had made his money and made the game as a tribute to the F355). The interesting thing is, of all the driving games out there (I do have a soft spot for them), I found it to be by far the best. It was best precisely because it wasn't engineered to provide rewards and progression at a constant rate. It's just bloody hard, simulating real car behaviour and then when you finally succeed in being able to stay up with the pack, 50 X better than any other racing game for that fact. I think Gamification (game style crafting of user experiences to encourage engagement), works very effectively as a lead-in to an activity but I suspect that, like Game-Theory, can easily lead to what is not always the most positive outcome (though as a business you perhaps don't care so much if you have built great market share getting there). I find it interesting that true simulation seems to counteract the negative "least optimal satisfaction levels" effect. There is a link I think between simulation and the heights of satisfaction achievable (and could it be living real-life ends up being the most satisfying of all - surely not!). As I understand them Gamification theories recognise this but still also understand the value of the easy lead in to encourage engagement.

Of course agent based modelling relates more to the psychology of decision making than physics simulation, however the trend has been in the games business for some time. The AI of in game opponents is very important. F355 implemented agent based modelling for the reactions of other computer modelled drivers on the race course - and as an interesting feature you could train your own agent to then race against - he would take on your driving style/characteristics. It worked well. Importantly the agents decisions and inputs were held to the same car physics model as you the player, so there was no shortcuts just to hide deficiencies in the AI. Of course other simulations, like Sim City, have been very popular for some years- albeit with less emphasis on the kind of strategic thinking an Army General might undertake.