Now we will discuss the effect of merging the same skill of different-level heroes fighting in one army. This applies to the following sciences from HMM3: necromancy, diplomacy, logistics, pathfinding, resistance, sniper, ... and possibly some other "group" skills.
Various options for finding a "golden mean" have been proposed here:
1. Averaging characteristics (an absurd idea, since 2 heroes are in any case better than each individually).
2. Simple summation (not bad, but everything trips over exceeding the "threshold" of some skills; for example, Necromancy cannot be more than 100%).
3. Layering the effect (my earlier proposal, taken from Probability Theory: if one Necromancer has a skill of 70% and another has 50%, then together they will raise: 0.7 + (1-0.7)*0.5 = 0.85 = 85% corpses).
Until recently, I considered the third option to be the most rational—because no matter how many Necromancers gather in a party, no matter how much they level up this skill and how many bonus artifacts they wear, the total effect will only approach the coveted hundred!
Unfortunately, after some reflection, a negative side was revealed: initially, the total skill grows at a breakneck pace, but then, when the heroes are already authorities in their own right, the effect of moving as a crowd practically vanishes. Why drag two or three freeloaders along for the sake of 95% if you yourself do 80%?
On the other hand, perhaps this is exactly the reason for them to split up and start wandering on their own? In any case, I have another idea, more "honest" in my view.
4. Adding experience
The thing is, in my model, the level is not granted to the hero as a whole, but to his individual skills (this way, gaining levels happens more evenly). Accordingly, each science can be development level 1-10. Twice as much EXP is required to reach each subsequent level.
So, why not merge the experience in a given skill rather than the effect of each single hero? In this scenario, a group of 2 heroes will be able to increase the total skill by no more than 1 level. For a larger party, the maximum would be +2 ... +3. So no significant cheating is foreseen.
A similar scheme will be especially relevant for Diplomacy—so that monsters don't join too easily. In general, Diplomacy is a special topic. It's not for nothing that I have returned to it more than once:
# 34: Diplomacy and Alliances:
http://www.alex-world.nm.ru/ie/gamezone/sub/hmm5/week_51.htm
# 51: Diplomacy and Charm:
http://www.alex-world.nm.ru/ie/gamezone/sub/hmm5/week_34.htm
I wouldn't want to do it again now, so for those interested, I provide a link to my epic on the Nival forum:
http://www.forum.nival.com/rus//topic/28491/
Another point that is not directly related to the topic of this issue, but is very good from the perspective of using the third idea. There was a spell called Resurrection. Moreover, only units raised by a hero with expert Earth Magic skill are resurrected forever.
I have a proposal here: within one spell, initially separate the concepts of Resurrection and Animation (only animated ones live until the end of the battle, resurrected ones—even after it). So, let's set 5% animation and 10% resurrection per skill level. That is, with level 1, 5% of the unit's initial number will be animated, of which 0.05*0.10 = 0.005 = 0.5% will be resurrected forever.
Too little? Well, that's what merging and leveling up are for. Let's calculate together! At level 2, 10% of the total number will be animated, of which 0.10*0.20 = 2% forever. At level 3: 15% and 4.5% respectively. At level 4: 20% and 8%. At level 5: 25% and 12.5%—now we're talking!.. Moreover, this percentage is taken not from the number of those who passed away prematurely, but from the initial size of the army, i.e., it is constant throughout the entire battle (unless a level-up occurred, of course).
The coveted (truly rarely achievable) 10th level will give 50% animation, of which all 100% are forever! Just imagine that with one cast you easily raise half an army! Why not all? Firstly, there will be artifacts that enhance this skill. And secondly, a prize is due for full skill development—passing the exam and receiving a substantial bonus (why not those remaining 50%)!
Merging experience should certainly help especially in this matter. Furthermore, this approximate formula does not take into account the magic power accumulated by the hero. And that, as you know, is worth a lot!