SamuraiProgrmmr
Posts: 353
Joined: 10/17/2004 From: Paducah, Kentucky Status: offline
|
Hello Everyone, I have just found this forum and am enthralled by the encouraging news that WiF will be ported to a computer game WITH an AI. I was part of the 'beta test' with Chris's CWiF a few years ago and kind of lost heart for reasons I will describe here. I started playing WiF with Version 4. It was a wonderful game. I liked the way that each type of combat (land, air, and naval) had its own flavor and how each flavor seemed to capture the decisions that were important. I especially liked the sea area concept for naval combat that modeled the ability of a task force to have a more effective patrol when closer to port. PLEASE keep the sea boxes intact. I liked the global aspect of the game and the ability to choose what types of units needed to be produced. This is certainly the best WWII grand strategic game I have ever played. I was absolutely ecstatic about Days Of Decision. The way we played, we would start with DoD in 1936 and work our way to the war. This allowed for wonderful alternate histories. (Building the Maginot line all the way to the Channel made things a little more difficult for a Germany that did not like to throw an attack that was not at max or near max odds.) I bought version 5 and like it as well. I bought version 6, but sadly have never played it. In fact, the counters may still be unpunched. Changes in jobs, marriage status, and priorities made it impossible to find time to play this wonderful game any more. The version 6 rules were enough different that I found it hard to motivate to play solitaire, so I went on to other things. I think it is crucial that there be an AI that is capable of at least teaching a beginner how to play well enough to learn from a good player later. Losing is inevitable (someone has to lose), but it is more acceptable when the player can learn from the loss and come back better. As a programmer, I understand the complexity of writing an AI. Although I have never done it, I have done some reading on the subject. The concept of an AI for this game is daunting, to be sure, but it can be done. Just like any programming project it must be categorized and broken up into pieces that are not so daunting in themselves. Have faith, fellow grognards, we may have the holy grail in sight. I know you don't believe me. I would make one suggestion, however, that might make you feel a little more hopeful for an AI that can, in fact, challenge the best of us and keep us on our toes. This suggestion could also have some bearing on the awful dilemma of "how much control do we give up to the AI when playing by e-mail?" Now, open your minds very very wide. Wider. A little bit more...... Thats it... What if the program were written in such a way that the Artificial Intelligence were, in fact, a module (in the form of a DLL) that accepted the state of affairs as a data object and returned the answer to specific questions when asked by the game engine? This is not too far of a reach since the game was originally designed to ask the user interface for all of those decisions. Surely this would not be that hard to arrange. Think about the major benefit: Long after Matrix Games has moved on to another project and cannot monetarily justify improving the AI, the gaming community will have the ability to do so. Also, players that are so inclined can write their own sub modules that handle their interceptions (air and naval) to speed up PBEM without giving up control. Just an idea.... think about it before you discard it. Sincerely Dean
< Message edited by SamuraiProgrammer -- 10/18/2004 5:09:45 PM >
|