Suffice to say Battle Realms as it is, is currently inclined to a single dominant strategy, that of massed ranged units that can shoot far and provide high Damage Per Volley.
Often you'd see the mass Musketeer/Cannoneer, mass Archer/Chemist, mass Warlock/Unclean One and mass Ballistaman/Hurler as the de facto army composition strategy because they're the ones that are able to provide the necessary and survivable DPV. Engagements in BR generally lasts 10 to 20 seconds, and after that both armies usually disengage to consolidate their forces. If an engagement lasts longer than that, it is usually because one army has a clearly advantageous position over the other and is attempting to exploit that advantage by chasing the enemy army down and/or conduct a siege to the enemy's town.
Firstly, if we're to look into the effectiveness of these units in terms of ratios, then we need to consider DPV per attack range (DPV+AR) to understand why these units perform as the basis, or the meat of most armies. DPV+AR is the reason why the utility of these units scale up so well into the late game. Ranged units have the advantage of being able to contribute to the engagement over a variety of distances and terrain features. A Berserker/Ronin/Samurai (any melee unit) can only provide limited DPV+AR simply because they're melee units (we'll ignore secondary missile attacks in this example), and this by extension requires considerably more micro from the player, more stamina burn from the unit, less positional options and considerably more exposure to the enemy DPV+AR than one who massed the dominant range units mentioned above. When we look at this scenario by visualising the overlap of DPV+AR, the range-dominant army would lay down significantly more DPV into the front line compared to a melee-dominant or mixed-army unit. And because stamina plays a key role in how a unit performs (movement, BG utilisation), the melee-dominant or mixed army unit cannot readily disengage from the engagement to the extent that range units can.
Secondly, and this is related to the first factor, Exposure Time to Engagement Ratio (ET:ER). Exposure is when a unit is actively being damaged by an enemy attack. By its nature, melee units suffer from higher ET:ER compared to its ranged couterpart. This is why you'd often see ranged units surviving an engagement far more often.
All this results in the current meta, which I would argue is not as dynamic as I hope it would be. When I first started playing BR all the way back in the 2000s, this was the game that really, REALLY hammered home its rock-paper-scissors concept. You can't just build a tier-3 unit and win, you need the right unit for the right situation. Fast forward to 2020 and things have settled down to these strategies which have become exceptionally dominant for a significant portion of the gameplay.
Admittedly the game does have a built-in method of limiting these types of strategies: horses and trees. Horses not only provide units with extra toughness, but it provides the much needed mobility the melee unit needs in order to engage range units without suffering too much within their DPV+AR area. Even so, this is highly dependent on the horse population of the map, and it could also be argued that a range unit on a horse can negate the mobility advantage melee units gain. Trees can disrupt the dominant strategy by not only blocking the line of sight necessary for range units to attack, but it also disrupts projectile pathing which effectively negates accuracy and damage. Trees unfortunately have limited use in how they disrupt the dominant strategy. Most maps in BR tend to limit their forests, and even then a single tree has relatively few health points (250) that it is easy to clear out a forest section once sufficient range units are available. What this means is that should the competitive multiplayer crowd be forced into playing the forest-heavy maps, they are more likely to cut down the forest than changing up their strategy meta from a ranged-focused army to a mixed/melee-focused army.
IF the game were to change the current meta of its dominant strategy, then I think an update to some of maps is necessary, mainly by adding more trees up to the point that at least 50% of its walkable pathways are covered in trees (we can safely leave Freedom alone, that I think is the de facto open-field map). The other change needed would be to considerably increase the HP of each tree up to the point that a concentrated effort by a player to clear out a forest would take a considerable enough time that his enemy would have the initiative for a flanking attack. I propose 2500 HP for each tree as a starting point. Also, forest alerts needs to be utilised more in relation to the addition of trees to most maps. I believe these changes would be a happy medium ground since it does not act as a universal change within the game (i.e. changing unit stats).
Where do I think all of this would lead to? A closer parity between melee and ranged units. Currently the dominant strategy in army composition is roughly 15:85 (melee:ranged) so imagine there's 5 Spearmen and 25 Archers in a medium population skirmish. With the changes I proposed, a more evenly mixed army unit like 40:60 (14 Spearman and 16 Archers) would be more viable as an overall average.
It should be noted that this theory does not factor in the early-game scenario, where there are too few military units to make the DPV+AR factor viable. I was only focusing on the mid to late game scenario with this piece, where building tech and economy have been developed enough.
There are also melee units that are the exception in these open-field maps like the Wolf Pack Master. They are able to subvert the meta because they are the only melee unit that, depending on the map and horse population, can truly outnumber an army that is focused on ranged DPV+AR. This is because each Pack Master can bring up to 3 wolf aides, and both wolf and Pack Master can be upgraded. Theoretically and potentially this would mean a 25-strong Archer/Musketeer/Warlock would have to face up to 25-strong Pack Masters and 75 (!) wolves. It needs to be stressed that this scenario depends heavily on the horse population of the map. Also, since a single unit can be attacked by 4 melee attackers, then the wolves alone can tie up 19 enemy ranged units, while the rest of the Pack Masters can deal with the leftover ranged units. Even so, the wolves are vulnerable against a cavalry trample attack.
Other exceptions include the Lotus Staff Adept and the Serpent Swordsmen, with their ability to partially negate the enemy missile attacks, hence negating their DPV+AR; they act more as a support unit than anything else.
Beyond these, melee units provide little strategical DPV value in an open-field map once the mid and late game has been reached (building tech, multiple buildings, many peasant huts, horses for harvesting), so it would be quite rare to see them being used en masse by then.
This is not to say that a melee or mixed-unit army cannot be viable. I would not want to take away the fact that some players just have that exceptional micro skill that allows BR's melee mechanic to shine. I would however argue that most BR players don't have the skills necessary to keep up the melee-micro. They might get it to work in the first engagement or the second, but can he keep it up until the end of the game? The player whose army is predominantly range units on the other hand would have considerably less micro to worry about during an engagement. The melee-focused player would have to micro every single unit he has in order to "tie up" the enemy units so that they cannot utilise their weapon and enforce a strong DPV+AR in the fight.
An example of a RTS that succesfully utilises forests/hiding spots/LOS blockers into its meta is Wargame. The "melee" units there would be the infantry, while their ranged units that provides the medium-range DPS would be the tanks. Interaction between the two types are given parity by the existence of garrisonable buildings and forests, where the tanks cannot effectively engage infantries.
Finally, I succeeded in making a new Innate for him. I named it "Frenzy", which makes him gaining 10% attack damage for every hit landed. This stacks multiplicatively up to 10 times, meaning that at 10 stacks he will not merely do 100% + 10%*10 = 200%, but actually 110% ^ 10 = 259% more dmg.
That'd not make him overly powerful though, due to his inherently slow attack animations. Unless the fight drags on long enough for him & he's allowed to attack, he can still be taken down by stunning, blinding & his weakness against Blunt would still be a problem.
But overall that'd make the Berserker a stronger fighter in his own right (will kill the Samurai & Ronin 1on1, without this he'll always lose), rather than just a meat-shield.
Here how I make it:
In Data_Abilities, I chose the unsused No. 9 AbilityType_Berserker_Transform & set its parameters as follow:
Great job dude, I think this will be a good guide to make new innate abilities for other units. I'm doing something similar, experimenting with the "omnipresent" tag to make micro-able BGs automatic. So far I've been successful with the Dragon Geisha's Sacrifice and Fire Shield ability.