Dissapointed

[Aikamun]

Also, tactics designed to exploit the tactical combat ai should be squashed. One example I read was to attack a large army with a force of only hydras. When the hydras routed(which was planned) they would leave a trail of poison into which the enemy army would run. This evidently results in a VERY positive kill:loss ratio. Maybe, this is part of the reason Pythium is considered overpowered. I would be very disappointed if these types of engine exploits were used in a multi-player game.

Aikamun

That is an expensive and slow way. The more prevelant and exploitive way of using the AI is against indeps (for early expansion). Their Archers/xbows/missile units are all set to fire closest. Thus you can leave a single unit with a shield in front of a mass line of archers. Most of the time if you have a decent archer brigade, the lone shield bearing lamb will surive due to the spread of arrowfire, while you incure no losses and most indeps break before they get to your line.

Another is that all normal fighting troops are on attack closest (as far as I can tell, I've never had anyone attack a flank that didn't engage them first). So you can exploit it in that fashion with spells, retreating options, etc.

Side Note: Nearly all Indeps (with the exception of some Amber Tritons, Amazonians) have only normal priests for their morale support. So you can easily use fear to cause routs without worrying about any losses.

[Coffeedragon]

Originally posted by Kjeld:
What I would like is the same (with some adaptations of course, like removing spellcasting, obviously, and adding a fire "single round command" on the topmost list) orders screen for units as the one for commanders.

Not sure if units should really be as flexible as commanders.

[Coffeedragon]

Originally posted by Aikamun:
One example I read was to attack a large army with a force of only hydras. When the hydras routed(which was planned) they would leave a trail of poison into which the enemy army would run. This evidently results in a VERY positive kill:loss ratio. Maybe, this is part of the reason Pythium is considered overpowered. I would be very disappointed if these types of engine exploits were used in a multi-player game.

Engine exploit? Are you sure this wouldn´t work in real life?

I think this is cute Hydra tactics, rather than an exploit.

[Coffeedragon]

Originally posted by Zen:
The more prevelant and exploitive way of using the AI is against indeps (for early expansion). Their Archers/xbows/missile units are all set to fire closest. Thus you can leave a single unit with a shield in front of a mass line of archers. Most of the time if you have a decent archer brigade, the lone shield bearing lamb will surive due to the spread of arrowfire, while you incure no losses and most indeps break before they get to your line.

Another is that all normal fighting troops are on attack closest (as far as I can tell, I've never had anyone attack a flank that didn't engage them first). So you can exploit it in that fashion with spells, retreating options, etc.

I haven´t really noticed that yet. But if it is so, perhaps the Indies could be added some tactical scripting?

[OG_Gleep]

The spellbook idea, or any sort of implementation that would allow control of what spells are allowed to be cast would be great.

Also just a chime in on what Jonas said earlier. Every forum I visit is full of people that submit bugs which are really just mistakes or misunderstanding on the their part. In almost all cases I am aware of, in order to fix a bug the developer must be able to replicate it.

Just my 2 cents.

[PvK]

Overall, the AI is very impressive, considering the huge amount of complexity in the game!

Suggested improvement for blood mages: Don't be so eager to kill all the blood slaves without a good reason. I have a Mictlani Prophet who is scripted to cast Smite five times. If we have blood slaves, though, he kills off all the slaves during the battle casting things which may be helpful, but I'd rather keep the blood slaves, especially when Smite is available for zero cost.

Similarly, conserving gems should be a factor.

I have seen multi-casting self-defense spells, but only rarely. One cause of stupid spellcasting is when the caster doesn't have many choices (low research). I'd suggest the AI should value resting higher than casting ineffective spells.

I think the body count for fire flies and stone shards is extremely low, and mostly friendly casualties rather than enemies. Missile & spell fire should be a bit less dispersed, and of course, the friendly fire risk needs to be weighed about 100 times more important than it currently is - killing your own men should be very very rare, rather than currently very common.

Probably the spellcasting AI should consider desperation - that is, compare the strength of both sides, and whether the spellcaster himself is currently at risk, and above all, whether the spell would risk own casualties, and then decide whether it makes sense to risk own units, or use gems or blood slaves, or even fatigue. (When the enemy is running away, it's probably a really bad time to massacre blood slaves to fling a badly-aimed magma bolt, etc.)

PvK