I know I did not post for a long time but this thread intrigues me.
Right now there everybody seems to look for possible exploits of the feature but there may also be chances, which I'd like to point out. I don't know if they are (easily) implementable, but these are my ideas:
- Force Splitting:
One thing which was always unrealistic was units who heal back to full health. I make an example to make my point more clear: Say you nearly wipe out a batallion (500 men) of riflemen (reduce its health to 10%). This rarely means that everybody is badly wounded and if you wait long enough, you get back 500 men fighters again. In real life this will mean that there are 25 standing men plus 25 wounded. You cannot just wait and recover to 500 men. The Size Matters Option could handle that. I propose that if the health falls below 30% there is a chance that its volume is reduced by one; health is multiplied by 3 the same time. The chance could increase when health drops more. Maybe it is 100% if the health falls below 15%. Multiple reductions in the same turn should also be possible.
I've considered this, yes... perhaps at some point I may work on something along these lines. Depends on how well accepted this option becomes in general first. It'd certainly reduce the value of having larger groups though, which right now is one way to keep yourself to potentially lower casualties (though at great risk of a large chunk of casualties taken all at once.)
- Merging:
I suggest that promotions can be kept, with a certain chance if three groups merge. Say one unit has a promotion and the other two units dont. The promotion should be kept by a random chance of 33%.
It can already be kept if you make that the primary unit that starts the merging mission. If you're aware of this and still making the suggestion I'll just say I'll take it into consideration for now.
- Siege: There was the point that one person should not be able to block a whole tile. I see two solutions:
1. maybe it is not possible to block a tile if your GS * GV is below a certain number. Maybe 20 is a good number.
2. or with finer granularity: Maybe a sieging unit can only block some resources, but not all. Maybe you can only block GS*GV/5 resources. (i.e. a "party - volume 2" of "humans - size 5" could steal 2*5/5 = 2 food or trade or shields). Details of what is taken first are a matter of discussion.
By sieging do you mean blockading? You make a point except that one pirate ship could certainly block a lot of trade in an area due to fear of him being out there and if he's only one ship he shouldn't be too tough to take care of. I mean, logically you make sense but the difficulty in coding a solution for this would be... ugh. Blockade in and of itself is already an intricate monster.
- Thieves etc. I propose that certain missions can only be effectively done in a group, say espionage. Either the cost is reduced if the group is bigger or certain missions are possible only in a group. On the other hand a group can be more easily detected, which may affect visibility or the risk of a mission. This way merging and splitting could make some sense for espionage units. Finer granularity on counterespionage may work here as well.
Perhaps long down the road I'll return to adjust the way espionage and such works though this isn't the only reason I've restricted Criminal units from being able to split and merge - the other reason is property value changes. If I didn't limit them from being able to split and merge a rogue could go into an enemy city and split himself down to singular individuals and each would have an equal amount of property change modifier - the way property changes are done is too intricate for me to break it down and modify it on the unit like strength and other factors.
- About the Pigeon: maybe some animals should not be allowed to attack ("can only defend"). Just like scout units.
Just my 0.02$ ...
I'm hoping to get to some more intricate animal AI this cycle which would encompass this without setting simple hard limits.
This feature seems interesting. I wonder what does grouping do to ranged attack power. Does it increase like strength or stay the same ?
Although it feels weird that combining three units that each have three strength creates one unit with a strength of only 3,6.
Depends on the ranged attack power. Many units with bombard ability cannot split or merge because it doesn't interact well with the bombardment. I THINK that archery bombardment does base on the strength comparison between units though so it should modify that appropriately for archers.
It does feel a bit weird the way it calculates but it does actually play out quite well in the combat mechanism. You can see a LOT of differences in the odds between 3 and 3.6 strength! Particularly when that added .6 is further modified by any and all applicable combat modifiers like a 75% defense bonus from the terrain which equates to about another .4 for the battle itself, then another 25% from a promotion that influences the fight and you've got a 6 pt unit in battle in the first example vs a 7.2 in the second. It's multiplicative with any combat modifiers since it's actual base strength. Furthermore, as the strength of the unit goes up the +20% means more and more. It doesn't seem like much on earlier units but on later game units it seems like a ton.
Tell that to my Dead stone thrower and Wanderer then. 
