Your forgetting that different units have different movement rates. There is no reason to run a unit with higher movement like a wolf into the attack zone of a bunch low movement units. If all the units have the same attack speed then it's obviously more complicated however as Heavenfall said you can still use formations, and retreating only delays the inevitable a few turns because maps are pretty small. If nothing else it is important the AI protects it's glass cannons. I routinely cut the attack of AI armies in half on the first turn of combat by fixing it so all my units get first strike and focus on the high damage and low armor enemy units. Not to mention I retreat my damaged units.

Also the AI needs to learn not to stagger it's forces. It's easy isolate and dispatch individual groups. This is really obvious when the AI hangs back because it has ranged superiority. The AI doesn't move it's troops at all until you get into melee so you can target the units on one side of the map and it takes a few turns for the rest of the AI to cross the map and engage, giving you a huge advantage. When faced with large local superiority the AI should really avoid combat for a turn or two in order to reposition it's units so that they all engage at the same time. It also needs to retreat ranged units behind melee ones.

And ranged units only move back one space at a time when they do retreat. Why?

I don't plan to work on tactical AI until the decision comes down as to whether terrain is going to be added or not.  I would like there to be terrain bonuses but there is concern that this could make tactical battles tedious for those (the vast majority) who don't care that much (yes, I know and the team knows that the people on the forums are very into tactical combat just like they are really into multiplayer).

The magic is getting a lot more interesting, I can say that (in tactical).  Beta 4A (not beta 4) will have a new tactical battle UI that will help the player see a lot more information that will help them know which units they should attack and in what order.

[a voice from outside of the forums] I'd love terrain bonuses!

Seconded....

Terrain bonuses would be nice - just make the AI ignore them at lower difficulties and voila! Instant simplification for people who don't care.

But think about it this way: you could remove a lot of interesting 4X or tactical features to simplify the game following the same line of reasoning. Customizing units is a hassle for some people! Having to choose where to put your outposts or improvements can be a hassle! In the end, I think 4X is a genre for people who like these details.

My point is the following: if you have tactical battles in the game, you might as well make these worth playing. People who don't care for terrain bonuses would already not care for many other tactical battle features, and would probably autoresolve anyway.

Thirdeded...

Looking good.  I was disappointed that the video ended.  More!

I'm finding it quite amusing that people are criticizing Brad's tactical play in a demo where he was flying thru.

Is good.

I think terrain is nice, but if it makes the AI more diffcult to implement it should not be added, because it is not as important as an excellent AI. The same applies to multiplayer in my opinion. An excellent singleplayer game is more important.

Exactly, but i think the tactical movement of your units and choosing the right targets are more important battle features.

I'll pay a dollar for terrain bonuses DLC!!! (queue impromptu survey)

I've followed these forums for a long time, and I don't think you've quite got the implied multiplayer / tactical ratio right.  Whenever MP comes up, the majority of comments seem to lean 'meh' or 'don't care', probably because turned based tactical games aren't usually that fun / easy to play multiplayer. By contrast, I can't remember hardly anyone saying the same about tactical combat.

That said, while in my imagination terrain features (I picture actual things / structures that perform active actions) could be awesome - especially in special locations or cities/outposts - they might not be workable. On the other hand, the new Kings Bounty reboot seemed extremely successful (didn't actually look at any numbers, though) and that was centered almost entirely around tactical combat, so it implies that TC can be a strong selling point.

That's my $.02, change is always welcome

Yes, but note how we went from "the AI suffers every first strike and loses" to "the player can no longer just charge in and do all first strikes". Of course he is not _forced_ to attack, he can just auto-defend or reposition his troops. When the AI moves, it will be within striking distance the next turn. If the human player keeps falling back, he will reach the edge of the map in two or three moves. Then the AI will get first strike. I really think a stalemate is extremely unlikely just because the AI plays better. The goal of a tactical AI is not to survive but to amass maximum damage on the enemy given its limited military strength in that battle.

I only meant my post as an example of how the AI was doing a very poor tactic, when it could be doing far better within the scope of the mechanics in beta3. I should have worded my post better (it was pretty arrogant). I just don't want the AI to use the worst tactic imaginable.

I'm one of the guys that couldn't care less about terrain bonuses in tactical combat. I'd rather have line of sight or more action abilities (mutually exclusive) selectable for unit designs. I think those would improve the tactical battles far more than terrain bonuses, without slowing them down like terrain would.

Quoting myself from another topic:

harpo