Challenge at start of game

I love it when games start out with a challange.  I am not as big of a fan of the 'exponentional growth factor' that many games have where as you put it 'The growth of the empire could be traced back to those first few decisions'.

Sammual

I think I was more saying, that the decision as to which way to explore with my initial units, or whether or not to conquer a certain lair, let to my second city being in a certain location, which influenced which direction things kept growing, rather then anything about the eventual exponential growth factor.  MOM also had a great, gradual mid game.  Where it fell apart, a bit, was the end game, I guess.

Player should start with a rough map of the world & a better one of its immediate vicinity.

The game universe has a past, so it's not uncommun to start with some minimal geography knowledge.

It could be nice to have maps as purchasable items that would give you more precise information about certain area : dongeons, ruins, mines,....

Thats an excellent point. I do agree that there should be some limited regional map knowledge. The idea about being able to by info from merchants about caves and what not is also top notch.

And why not a "cannibal" fog of war until you research good cartography : when you send a unit to an area you see that area. When the unit leave you stil lsee the ground but not units. It's the fog of war. But if the units hasn't the cartography skill then after 10 turns you forget what was there. You lose track of the ground.

The best would be a the world divide in area, for each area you can buy, trade, get maps that increase you understanding of it. You can also explore but some dongeons & secrets should be only available through maps.

The objective would be to discover 100% each area.

Somewhat like sea maps system in the old "High Seas Trader".

NB :

It would be good to have some hard dongeons close to the starting positions that would advert the trope : "Farther you are from home & stronger are the opponents".

How about dungeons becoming stronger over time instead? So you don't get beaten up right at the start because you got placed next to a crazy strong dungeon and got hit by a swarm of demon lords on turn 2, but the dungeons still pose a challenge late-game. So you'd have a "safe" area near your cities, but any area not cleared regularly out gets more and more dangerous over time as you let it grow wild.

It'd be reasonable for new dungeons to pop up over time/old ones to be repopulated (depending on how dungeons are integrated into the map). Perhaps a bit less near major concentrations of cities, (hard to set up a lair if there are soldiers swarming everywhere all the time, and it prevents you from having to play wack-a-mole every couple of turns) but there should still be the odd dragon turning up to ply its trade late-game.

You could even have a pretty good justification for new lairs starting out stonger than the original ones: "as more magic gets thrown around (fighting channelers and their minions) more terrible creatures are woken up/created by the fallout."

Of course there'd still be varying dungeon strengths right at the start. In MoM, for instance, magic nodes were always far more heavily guarded than ye olde ruin.

@Zaisha

Ok, as long we avoid the Oblivion uber armored thiefs in the late game.

How do you feel about uber armoured Death Knights? (who ate the thieves)

Yeah, in MoM, there were varying levels of "badness", with some variability, so you had an idea what you were getting into.

I think folks who are really into the Single-Player experience really like a strong Player-vs-Environment component to a game (this was where MoM excelled, IMHO) while folks who are really into the Multi-Player experience really like a strong Player-vs-Player component and worry that "random neutrals" can unbalance that and/or pose real problems for the AI. 

I think most modern games really whimp out on the P-vs-E component.

One dynamic that was always a bit problematic to solve, whether in MoMor Warlords or Age of Wonderswas how you explored a spot like that without potentially losing your whole army.  MoM spawned the "obligatory sacrificial militia" tactic to figure out what was in there, but that always seemed kind of lame to me. 

On the other hand, spending 40 of your first 60 turns' production to build an army which then gets annihilated by an Efreet Lord wasn't much fun, either.  There's got to be an innovative way to crack that nut (maybe something as simple as requiring units with a "Scout" ability to spend a number of turns scouting out a location, w/a greater chance that they'll be discovered before completing their mission if they check out a particularly dreadful locale?).

You could always just give scout units the ability to immediately see a challenge rating for a specific group of troops/mobs but take time to see how many and of what type they are depending on skill level of the scout.

That way you wont go into a fight entirley blind.

I never had problems in MoM, knowing what the hardest unit in a location was allowed me to build an army to take that location.  Sorcery Nodes with invis units could be a real challange so I always avoided them until I could take them off an AI or I had a god stack ready.

Sammual

I like the idea of a 'scout' ability or unit. simple but effective

I think that there's a scout hero in one of the screenshots, and I'd love to be able to design a unit early with a horse and not much else to go scouting using a ton of moves (and the high LoS that should go with it).

Making the early part of the game challenging is, generally, a good thing. Heck, making any and every part of the game challenging is a good thing; both statements are so straightforward it shouldn't be possible to object to them.

What I would mention as a caution, however, is that early-game challenges should not take the form of the possibility of random and unforeseeable death. If I lose early, I need to feel like it was because I made a sub-optimal decision, not because the game engine just decided to be prissy that night. So, for example, an acceptable early game loss would occur if I failed to build any defensive units and after ten or twenty turns a roving bandit came by and sacked my only keep. An unacceptable early game loss would occur if I moved my first unit into an unexplored tile and a random number generator told me I disturbed an ancient burial ground (that I didn't know was there), spawned a half dozen very powerful units and then proceeded to have them obliterate my unit and city in two turns.

I agree that the early stages of a 4X game - exploration and expansion - should be about more than just those two things. Even though the player likely hasn't encountered the main opponent yet, the game world itself should present challenges, such as roving beasts or neutral/hostile towns. Those challenges, though, should be as carefully playtested and balanced as the enemy AI to ensure that the player can win if they're smart, not just lucky.

- Ash