Quote:

Original post by LorenzoGatti As a player, I would find random variations annoying: tactics that work or fail with some specific units would have different outcomes with other units of the same type. Variations would actively invalidate any experience the player accumulates, degrading the game: instead of applying reliable qualitative and quantitative rules, the player could only gamble on the stats of friendly and enemy units, without much skill. The chance factor increases as described when the stat variations are hidden or not worth inspecting; if variations are evident a lot of micromanagement potential, as already noted by Kaze, is added to the gameplay, and there is a different negative impact on player learning: strength estimates depend on unit variations, not on units, wasting a complexity budget that most games prefer to spend on actually different units. I don't agree with tstrimp on the incentives to the new types of micromanagement (sorting units by stats, getting rid of the weaker ones, etc): every remotely useful behaviour has a place in a winning player's repertoire, the game designer can only hope to make it a small place. In this case, a computer AI completely oblivious to stat variations would reward any amount of micromanagement from the player and set up an interesting decision (should I micromanage these units by stats, or should I attack with a randomly composed force and micromanage something else like their targets?).

I think it would add more strategy to the game because you wouldn't necessarily be able to rely on the old rock, paper, scissors type of game play. Imagine your enemies surprise when his average horsemen don't completely slaughter the group of elite archers you put together. Or when he realizes he is cutting through the infantry you sent at him way to fast because you didn't want to waste your strong units on the diversion.

Traditional stats leave way too much predicability in the game. You know what units counter what units and at that point it just becomes a matter of applying the techniques in the right order.

Quote:

Original post by tstrimp I think it would add more strategy to the game because you wouldn't necessarily be able to rely on the old rock, paper, scissors type of game play. Imagine your enemies surprise when his average horsemen don't completely slaughter the group of elite archers you put together.

Imagine how dumb your opponent would have to be to send his average units in the first place...

Making it so only 30% of your units are worth using is not tactics.
I'm all for variation, but not on the scale of "worth using" versus "useless junk"
Rather, you could give some of them random special skills that would benefit the rest of the squad.

But don't make it so some units aren't worth using.

Quote:

Original post by Spoonbender But don't make it so some units aren't worth using.

Are they ever not worth using? If nothing else they can be fodder to soften the enemy up. Like in X-COM you'd recruit soldiers and most would be average, a few junk and a few exceptional. There was certain units that I would use to peek around corners. Even weak units have their uses.

Unit variation sounds interesting.

Some variants on unit variation. :)

1> You can't tell what the variable stats of a unit are.
2> Units get promotions and decorations as they survive battles.
3> Unit stats vary as they gain experience (but generally get "better").
4> The UI allows you to sort units quickly and easily based on promotions and experience.

So you start out with a bunch of grunts. You put them into a fight.

The ones that survive get decorations and promotions.

You pull the most decorated and promoted units for elite squads. Possibly you spend money upgrading those units to more powerful ones (better weapons/armor/etc), or put them through extra training.

In effect, this duplicates some of the information that generals don't have. They don't know how competent their armies are -- unless they test them.

In a game where you progress from map to map using the same supply of units, this could get rather interesting. =)

Maybe there should be a 2% chance (or a similar low chance) that the unit you build will have that special talent. He's that guy who seems to take things naturaly and who always seems to come out on top. Then there could be random type of bonus this unit is awarded.

He could be better at shooting, have more stamina or be able to take more beating before falling dead to the ground. He might also be that natural leader type who will advance in ranks faster than others.

Do variation, but don't make it gambling. Make the units different, but without making one better than the other. You could have multiple stats like strength, stamina, etc. and have say 10 points randomly distributed between these stats. This would allow you to use your unusually strong warriors to rush in and attack the enemy base, ambush his counterattack with your unusually stealthy ones, all the while defending your base with your warriors that can sustain a lot of damage.

There seems to be mixed reaction to this idea.

For all of the people against units having slight variation to units. It seems that people are concerned with the idea of Junk units and lack of consistent outcomes to confrontation. Doesn't variation in fact add to the intensity of the game by forcing the player to be more creative with their strategies and adapt to changing situations? How does the fact that that you can't be certain a grunt can kill 3 zombies before they can reach him every time make the game unplayable or prevent learning?

There is defiantly plenty of scope for game play in regards to unit variations and unit growth. One idea I had was once unit reaches “level 1” only then do the become unique. Once they become unique they can begin gaining special awards to further distinguish themselves.

I do like the idea of unit stats determining unit appearance. Not sure what the best way to do it at the moment it will depend on the capabilities of the graphics engine I use. Composite models, or modifying unit meshes sound like the most visually appealing option, although just having multiple models for a unit would be simplest to implement.

Unlike traditional RTS game there will be a pre battle preparation phase which players spend resources points on a number of different things. One of which is to include units from your personal command. Any surviving units can be added to your personal command allowing you to use your favorite units in later battles.

It's all fun until players start blaming their losses on a crappy batch of units.