Frustrating AGRICOLA

[McMonkey]

I was just using it as a quick example. It looks like Techumseh will make a permanent version that will include point 7

[Turno71]

[joke mode]

Naaahh... i was only pointing out that the most helpfull hand to you usually is at the end of your own arm.

Got lazy after so many years, Werd?

[/joke mode].

Great idea. Way to go, Tech!

[civ2units]

I´ve noticed during my latest playtest with my Civilwar scenario that the AI only produces units when the improvements are disabled for the AI. But it isn´t really good to give the AI many different units, only the (for the AI) best unit will be produced, all others will be ignored.

I´m giving the AI now only four different units. Tecumseh wrote that the unit role also is important, so I gave them different roles. The AI now produces all units. Another thing I noticed when I gave the AI too many units, some units aren´t available for the AI controlled civ. I think that is because of the stats or maybe a bug. So the AI only gets a few ones.

Aggie noticed during the playtest, that the unit production won´t be updated correctly when researching the techs. I´ve now given them different roles and they updates correctly.

For my Civilwar scenario I going with these settings:

- Improvements are only available for the human player.
- Most of the wonders are not available. Only a few.
- Barracks are now very expensive. Nearly all units in the
Civilwar got their experience on the battle field and not in a camp.
- Nearly all techs are disabled. I think when the AI knows that there
isn´t anything for research, he will produce more units instead.
- I´m using in my case two folders. It gives me more freedom
to use many different units and stats.

The same goes for the Independence War scenario which I´m updating.

[Turno71]

Nice post.
For my in development scenario, after carefully reading this post, i've decided for this settings:

1 - improvements reduced to minimum.
2 - only a few wonders - all builded at the start of the game.
3 - barracks - renamed "Training officer" - available late in game at medium cost. only a few builded from the start, an mainly for one civ (see point 6, that's why. I don't want my main guest to be easily overwelmed early in the game... ).
4 - AI stats changed to maximum aggressivity.
5 - almost all of the research techs will give a unit. Long time for research
6 - a hell-of-a-lot-barbarians as buffer between civs

[techumseh]

I've reorganized and updated the article and posted it on the SL site, here: http://sleague.apolyton.net/index.ph..._Expert_Player

[Turno71]

Thanks. Downloading it right now.

[AGRICOLA]

The thinking in this thread has gradually evolved towards scens where the developer removes as many non-essential things as possible. This can work very well and a bare bones scen that has been pared to the absolute minimum can be a lot of fun to play.

A couple of months ago, when I had had more than enough of micro(mis)managing large armies and economies, I tried two old, classic scens . . . . . Case's THE CRUEL SEA and Nemo's unfinished THE BLITZ. Both involve minimal city 'housekeeping' and no land units of significance. They are all about ships and/or planes and fighting.

There are two ways to play The Cruel Sea . . . . as a historically more or less accurate U-boat campaign with an inevitable defeat and a lousy score . . . . . . . or as a completely ahistorical Battle of the Atlantic that racks up a huge score. I have no doubt that Tech would choose the high road whereas I like to rewrite history. Fortunately, both of us would have fun.

The Blitz is playable only up to Day 30 or so. At this point the German bombers start to stack up outside London without attacking the city. From all reports, this is what caused Captain Nemo to abandon further development of the scen.

Up to this point, the Fighter Command has a thoroughly difficult time in trying to cope with constant Luftwaffe attacks. There is only limited scope for imaginative play but players are more than busy enough fighting the Germans and trying to bring freighters into port.

I unsuccessfully tried a few things to fix the problem, but failed miserably. However, a good designer with time to spare and TOT at his disposal might actually make the scen work.


The one thing that I would caution against is turning out a scen that leaves too little room for player initiative. IMHO, one of the classic scens, Nemo's SECOND FRONT, falls in this category. The following thread outlines my sour grapes about the scen and LaFayettes sage comments after he played it.

http://apolyton.net/forums/showthread.php?s=&threadid=105479&highlight=scenarios+from+the+community

Feel free to ignore my ignorant kvetching but please do pay attention to LaFayette's words of wisdom.

[techumseh]

The one thing that I would caution against is turning out a scen that leaves too little room for player initiative. IMHO, one of the classic scens, Nemo's SECOND FRONT, falls in this category. The following thread outlines my sour grapes about the scen and LaFayettes sage comments after he played it.

http://apolyton.net/forums/showthread.php?s=&threadid=105479&highlight=scenarios+from+the+community

Feel free to ignore my ignorant kvetching but please do pay attention to LaFayette's words of wisdom.

See also this thread re: Darth Veda's 'Herbstnebel'. http://apolyton.net/forums/showthrea...ht=herbstnebel

It was also a great scenario which attempted to stretch FW/MGE through the use of multiple, sequential events files, but ended up being a prisoner to history. This is a common pattern, starting with 'Sparticus', followed by 'Second Front' and 'Herbtsnebel'. All wonderful scenarios, but very linear. New and innovative strategies cannot be easily accomodated when multiple FW/MGE events are used, as each new file must use a fixed set of assumptions as it's starting point.

[curtsibling]

I've always day-dreamed about updating Nemo's 'Blitz' and trying to get around the bomber madness...

It would look nice in ToT as well...

[kobayashi]

Seems to me the goal here is the timeless one of making the same game hard for hardcore gamers, but easy for casual gamers. And as a proper tribute to the man, Agricola has been picked as the gold standard to beat. I have tried everything mentioned here and then some (e.g. jacking up operating costs to make the economy a basket case, making new land units only available by random spawning) and still the aforementioned player keeps thrashing the AI.

Admittedly, I have not been entirely successful in thwarting the Cola, especially during beta testing. But I have figured out a few things…

The way to keep madmen like him in check (and still make the game playable to normal humans) is to disincentivise them from trying their shenanigans. Whenever something 'ungentlemanly' is detected, say when a certain city is captured or a certain unit type is destroyed, then a punishment is meted out. Punishments could be: allow the AI to skip ahead in the tech tree, force an allied AI to declare war on the human, make some important human unit obsolete, give the AI an endless stream of free units...you get the picture, ideally something bad that will happen eventually, with aggressive play making it happen earlier.

A more devious method is to use the ANY flags which apply only to the first civilization which triggers the event. The beauty here being the human player won't know if and when another of the AI civs has already triggered the event. If you minus 5000 gold from an AI, it won't care, but if you did that to the human player, that would hurt.

Another way to build in some self scaling AI augmentation events. If A is killed spawn B. If B is killed spawn A. It scales the game such that a novice player won’t be affected.

[gapetit]

Hello to the whole world.
Techumseh I have a doubt that I believe that it does not appear in your article. A time ago modifies the scenario Barbarossa to play it with the Soviets. I prepared an assault very beast for the Nazi and though the scenario is too big in order that the AI prepares powerful assaults distantly, provided to the cities Nazi of an enormous potential to construct units... but instead of constructing units his devotes himself to the capitalization though the technological advance that allows his discovery and use only can be used by the Soviets ... which can be the reason of this disorder?
Saludos,
Petit

[AGRICOLA]

This is a common problem in large scens such as Barbarossa or Red Front, where an AI civ starts the scen with a very large numer of units. When the number of units possessed by an AI civ reaches some limit, it stops producing units and starts producing capitalization or city improvements. I don't know exactly how high the limit is, possibly 500-600 units. My best guess is that this behavior is built into CIV2.exe so that one AI civ does not use up too many of the 2048 available units slots.

However, as a large AI civ starts losing units, it will resume building units when its unit total decreases to somewhere in the 100-200 range.

The limitations of the CIV2 MGE events language make it very difficult, if not impossible, to overcome this. In TOT, with Count= and Randomize, events can be used to give an AI civ any units that it needs.

[gapetit]

Thanks Agricola.
Saludos,
Petit