Brian Reynolds's Article

[Ceci n'est pas Snapcase]

http://www.gamespy.com/devcorner/april01/reynolds/

Comment.

[SnowFire]

BR is one to talk about this, sure. Now I love Colonization!, Civ2, and SMACaBRD, but these games are all PRIME OFFENDERS in the gameplay balance category. It's not so much the moronic AI that's the problem (odd that Colonization, the oldest of these games, has the best AI in relation to the system), it's the simple fact that none of these games scale the cheating. All of these games are challenging in the early game when everybody is equal (sometimes even too challenging), because of the computer cheating. But with superior human production ability (and the ability to do things like drill boreholes), all of these games lets you win one war towards the beginning, and use your superior population from that to overwhelm everybody else, just as he describes about Civ I. Now the easy way to solve this is to give an option for graded cheating, where the AI starts out cheating a little and then cheats more and more as the games goes on (unless the human player starts falling apart). That, or else give more scenario options so that the scenario can have the other guys cheat.

Heroes II does not have the greatest AI ever, but with its scenario editor you can easily make global events to give the computer money, troops, etc. and plus you can create stacked scenarios where the human player is far outnumbered. And because of that, Heroes II is still challenging, while no BR game is anymore.

[MidKnight Lament]

An excellent article.

Anyone who's played a few games in their time (particularly strategy) will know all these guidelines already, but it's encouraging to see Brian express it clearly. It's obvious (As you'd suspect) that he's put a lot of time into thinking over these things. The fact that it's an instructive article makes it interesting too. "Yeah, that's right. I could do that."

------------------
- MKL ... "And a sun that doesn't set but settles" - Augie March
Shameless Plug: http://www.poetic-license.org ............. All welcome.

[MikeH]

We were discussing this on ACOL. We thought there were some holes in his arguments about RTS balance in particular. Shining1 made some very good comments.
http://www.planetacol.com/ubb/Forum6/HTML/014782.html

[Grumbold]

Many things in his article make sense, although Shining1 does point out some fallacies. In all it seems a case of easier to preach than to practice. All game designers *think* that what they are developing will be playable, fun and have long term appeal. Certainly getting a crude alpha playable version up and running early in the design process could help in many cases where they are mistaken. Ultimately though the major problem seems to be proper project management. Almost all games are completed under immense pressure to meet final deadlines, programmers working 80+ hour weeks for the final months while simultaneously stripping out unworkable or unimplemented features at the last minute. No wonder what appears on the shelves often fails to deliver what the reviewed beta code promised, isn't balanced well and has had truly moronic bugs sneak into the gold version.

Without the internet available to issue post-release updates this kind of sloppy development could not be sustained because the game return rate would be prohibitively expensive. I can remember being disappointed with software in the 80's but I never returned any because it crashed 2 seconds after I began play. Someone would have tested the main game elements to completion using the final code before releasing it to the public. I admit the proliferation of differing standards for sound, graphics and other hardware has added greatly to compatability difficulties. It is the main reason why I expect the X-box has a viable chance of swallowing the PC games market if it is marketed properly.