1) The first mistake was playing (not to said testing) the first french edition during an evening. We don't do a turn. Played less than half an hour and spend 4 hours in the rules looking for solutions to the numerous problems arising while crawling in the sequence of play. Of six or seven only two wanted to continue to play. The friend that had bought the game and me. So we went to the shop that was kind enough to let us copy the english rule while at the same time we found a ton of errata to the french rule.
2) How. Some luck. A good part of hindsight. And a lot of badly written rules and mechanism. I explain : first turn saw a new queen in England rated 9/8/6 while a gold mine was discovered in the North of Anglia. If you remember the state of England in 1492, those even made a lot of difference for GB. The 3/4 next turns while Fr and Sp where fighting each other to the last man following the preprogrammed victory table; saw England "build" a lot of merchants ship in trade zone to gain some full controlled TZ, all+1 allowed manufactures and do some ship reseachs. The army was just a strong garrison and fortress in Calais, stronger looking for the Fr than the Hasburg's one. At the same time build ships to have a navy looking mighty like the one of the 19 century. No other expense, no war, no costly diplomaty except to keep England at peace, only administration and building. Then during one or two turn, sending small 2 or 4 ships squadrons at the same time to discover the northern atlantic in wave of contiguous sea zones. Trade some of those discoveries with SP as Sp hasn't yet interest in North America. Then when Canada coast are discovered, US coast traded and next naval technology is discovered, i send a huge fleet with a huge army on board in an impulse to Canada coast then split it in two ships fleets and armies aboard and let the armies disembarks in the coast lands and discovers those. Then reimbark the surviving armies aboard the ships and regroup. Came back across the atlantic with a couple of ships and men but more than half the territories where the landing took place discovered. Redo the same next turn while starting to build trading post and colony to let the modifiers dissapears before the true attempts. That's it.
During this time, EU had a set of VP tables than are follwing history line while GB had totally moved out of history table. So, nobody had interest to DOW to GB at that time. Same still happens in the computer version but to a less extend due to bad boy factor.
Now about CEiA AI, do you really think that someone would be able to program an AI that will be more a challenge (without cheating) to someone older than a 10 years old kid? In all group i played, we used the week between to game encounter planning the next quarter's stategy... looking to it
- strategically : diplomatic, VP, balance of forces, alliance, naval move, forced peace, minors state, what if i surrender now (thinking forward to peace condition) ...
- tactically : position of others corps, city and depot garissons cutting or hampering supply, supply line with a 4 corps per depot limit, french double move, terrain effect, enemy goals (conquer, destroy, capital occupation...), own goals, tactical advantage (morale, cavalry superiority), country perticularity (cossack, Austria light infantry, feudals, insurrection corps...), time to go to peace, PP combat's effect, extended or regrouped formation, leaders effect ...
In all games i played the first turn of an evening is always twice or thrice more well thinked and resolved that the last one when you are tired and doing less planed thinking (only spend a quarter of an hour rather than a full week). And you really want a AI to be well enough programmed aka all the previous translated complex formulas resulting in simple binaries solutions aka yes/no instructions to balances all those parameters and translate those in coherent moves? I really think you're dreaming. I never yet found a computer game that offered a challenging AI without cheating others than Chess and other game with simple rules and AI that are programmed since the start of computer history. Huge planned computer hit like warcraft have poor AI with full team of programmers. What are you really expecting from a single programmer that doesn't know the board game when starting?
It's possible to program an AI for a linear wargame like one on the battle of Stalingrad. Defend to X turn or until a konw X/X battle factor ratio; attack starting on X turn looking for X:1 local superiority before that on XL1 ratio, going to hex XX/XX direction and planning X hexes surrounding. Otherwise, don't buy a computer wargame or diplo game to play against the AI.
And now to finish about rule change. Why change rules that were tested on the boardgame and mostly working and that all the first buyers (and last of the first ones doesn't do a good critic) of the computer game have already played and enjoyed. I'm pretty sure that all of us that are chearing/bashing good/bad news from Marchall and the playtesters are vet's from the boardgame. Ok nearly all cause someone will post claiming he is the exception.
At the start of the forum, someone...think it was Michel was a complete newbie and posted a lot of interesting ideas that were different from the board game. I remember one idea being hidden geography and corp location on the map reflecting the lack of communication of the time. Was a very good and interesting idea but got completely trashed by the ones (and was mostly one of those) than wanted an orthodox translation of the EiA boardgame.
To finish, the goal is not to bash Marshall. I'm sure he has a lot of goodwill and endurance. The problem is that he started something without a clear and fianl plan of what this project will be with a cost/benefict, targeted market, programming time needed and planned line to reduce the difficult step in programming ... I think that ADG and Matrix are more to blame than Marshall on this.
Now what's the use of this if we don't learn from the mistake done :
What is the estimated time (for programmer) to finish the AI?
What is considered a decent AI?
What is the ET to remove the bugs?
What is the ET to add more scenarios?
What is the ET to add TCP/IP?
What is the ET to add more chrome (optional rules)?
What is the ET to add an editor (letting players created mod and other scenarios) and attracting more consumers... (let's continue if you have other ideas)
What price are you ready to spend for those parts?
ET may be replaced by actual % progress and time already spend.
I really would like Marshall and David reply to this. And some polls or forums help like WiF may then be launched.
< Message edited by Titi -- 9/15/2007 6:34:25 AM >