To start with the game is a vast improvement over E:WoM, I played almost my entire labor day weekend. I designed a couple of map scripts for CiV IV and really dig cool realistic maps. With any 4x TBS square/hex based game you get into map issues with exploration/expansion, number of opponents, starting position, symmetry/asymmetry asthetic and balance, size of map an number of potential settlements, total number of turns in a game, movement, etc.

Min/Max since E:FE is now single player is kinda thrown out the window. That being said AI is now more important and having a balanced game within the boundary of a asymmtretic asthetic is cruicial to a fun experience. An what that means to me is making the ai, without artifical bonuses, perform near what the player can do within provided mechanics.

The horizontal expansion of resources is directly related to number of cities and city improvements and the vertical progression to the linear conclusion of the game is inversely proportional to that horizontal expansion. I.e. the more cities you have the faster you win because you have more resources. So unless this design paradigm changes what we need is an ai capable of settling as fast and efficient as the player. And doing so in an asymmteric way, i.e. so that at turn 100 every player is not sitting on exactly the same number of cities. This means not making it perfectly symetric(balanced), but instead focusing on throwing out outlier situations (like really fubar starting positions) and either balancing turn saving mechanics (spells, roads, scouts, escorts, monster attack chance/immunity/correct position on end turn) or teaching the AI how to use those mechanics.

The Map

The Good

• Wild areas, A+ best idea you have had yet I love the increased danger, the bosses, and the way it creates an unsettleable area for awhile, would love to see a forest and swamped themed area.
• The balance between fixed city points and unsettlable areas - anything that helps the map look more like a sparse fantasy world with wild forests, expansive deserts, deep dark swamps, etc is a really good thing.
• Random Generation, Its working pretty well although there are still times it creates unattached islands, are these intentional for the raise land spell?

The Bad

• Still not completly stable you can crash it by revealing the map with ctrl-u zooming all the way out and hitting ctrl-n repeatedly. Also weird line artifacts will appear creating vector from a point on the map that is viewable from certain zoom levels. I can also Main menu start new game without exiting the client and lose the water textures after about 10 restarts.
• Losing markers on reload, you lose mine/mob/other markers in the explored but unsighted areas when you reload, not consistently but frequently, rivers can get cut off in the same area etc. Also the resource markers get lost seemingly all the time although the mouse over tool tip shows G/M/E for a square. It would be a HUGE help to always see these markers for settlement planning
• Water water everywhere, you can start a map with a certain number of opponents and the water generation seems to be anywhere from 30% coverage to 80%, which can vastly change the available amount of land in a given game, we definetly need a pre-game option for amount of water
• Fubar starting  locations, I decided to play a few game with fog off so I could watch the ai, and bad starting locations can kill an ai. In one game an ai was in a corner with two small corridors between mountain and land leading west and north. The closest settlement point was 42 squares away, it literally just shut down, didn't produce any units and when its sovereign lost a fight and got sent home it never sent him out again.

AI Expansion

I think the water setting could directly help prevent fubar starting locations, after playing alot with watching the ai there definetly was an optimal number of opponents for a given map size, the more breathing room an ai has the better it does in the settling phase. On the large maps the best performance was with 6 players including myself.

NOTE: "end turn tactic" I mean the ability to end turn while having a unit selected and then move that unit immediatly on the start of the next turn to avoid it being attacked by monsters, I assume this is not intended gameplay and an exploit.

I also played a few games on challenging where I would not allow myself to reload or use the end turn tactic and after 50-75 ish turns I would reveal the map and see how I was doing against the ai in amount of cities. Pretty consistently I was ahead by 1-2 cities and in some cases 3-4. Except of course against Tarth.

A customized player with master scouts is an i win button I can outsettle the ai 4 to 1 (if not more). The crazy thing is how well the AI does as Tarth with the same ability, not as well as the player but nearly. A game like this comes down to number of turns and a players ability to make the most actions per turn. Master Scouts just reducing the movement penalty in forests is probably the best trait in the game because of the sheer number of moves it saves you across your units, but that on a scout and most insanely on a pioneer it elevates to an "I win" ability.

Scrying pools and Enchanted hammers also a bit op, which means an ai without earth magic or getting its first hero without earth magic is crippling it against a player who with earth magic will save a ridiculous number of turns. So nerf it, move it to tier 3 earth, give an equivalent tier 1 to air/water/fire ( 1 could be moving the +gildar enchantment and teach the ai to rush civics and buy pioneers,1 could specifically produce settlers at a faster rate/make the cost less, 1 could produce settlers with a movement speed bonus and/or stealthy/immune to monster attacks.

How to improve the ai settling problem.

1) Add a water coverage option to the map generation

2) Make a table of suggested num of opponents based on map size and water coverage (since water essentially changes the map size). I found 6 players on a large map (including me) to be optimal.

3) Add a check to the starting position function that makes sure from their is a second settlement position within at least 20 squares.

4) Teach the ai the 11 square city radius and how to get 2 cities into a single settleable area if there is a 5 square gap

5) The ai does better with good scouting and it should put stealthy on its early game scouts to increase their suriveability.

6) If you don't want to do number 5 you could flaten survivabilty balance in the < 100 turn game, make 80-90% of roaming monsters counterable by an early game unit, I.E be able to give a scout the ability to survive most early game encounters then sit stationary and heal, same for a pioneer with an escort. If you have played a little Warlock: Master of the arcane the survive one encounter and rest mechanic works well and could be adapted.

7) Trading should be almost a straight research line for everyone, pushing settlers to your champions or escort units with roads saves an enormous amount of turns as well as gets units past roaming monsters and camps. It would also be nice if your movement line indicator was segmented by turns (you could see if a pioneer moving on a road was going to end a turn next to a monster camp and adjust its move (helpful for ai as well)

Other stuff

1) The green outline of the unit portrait in the tree does help to see which units have movement left but after a unit moves you need to cick it again for the outline to go away, it would be nice if you had the "at a glance" type indicator like the zzz for cities to see which units have movement remaining. I turned off auto advance to next unit because I hated the camera snapping when I wanted to see the map layout and cycling through all units and looking at movement left is painful

2) Tooltip for cities in the left tree doesn't include essence / num of enchantments. You should have a nice flashy zzz type of icon notification when you have a city with a free enchantment slot.

3) In all places where food is referenced it should show in parens number of turns at current growth rate til you are at max pop. A city within 5 or so turns from max pop should have its icon change color or get a nice zzz type of notification.

4) Train Henchmen for Altair is too cool im torn as to whether this should be made available to every faction or unique to altair, its kinda a base game thing that is painful to play without.

5) Early game mages, do they even exist? There need to be more direct damage ranged spells early on for mages, or more frequent staff drops. There is an early game problem with the amount of mana and actually using your mages, as mages in tactical battles.

6) Would be nice to have the classic roles played out more in tactical battles, mainly because the ai always seems to attack the closest unit with damage and its easy to position your defense guy to take that damage, would be nice to see monsters go for the squishies in the back and add taunting abilities to defenders