TW has a dream, and that's being as e-sport friendly on a massive scale as Dota, CS:GO and LoL
No they don't. It's impossible to compete with those games.
Considering that Bannerlord EA failed to garner much attention and is currently losing players, the only route to competitive success is to slowly turn the game around and figure out how to make multiplayer compelling to a broader audience by EA launch.
Examples of games that have grown from humble origins and no major monetary backing are Paladins and Brawlhalla. They've slowly grown into some esports prominence. Paladins is the 40th highest prize awarding game with over $2million in prizes; Brawlhalla is at #64 and continues to climb with over $1million. (
https://www.esportsearnings.com/games)
The class system is horrific to balance yet simultaneously offers players low strategic depth. The most useful balance data is win-rate and pick-rate (and ban-rate but not in BL). This is categorically true in virtually all esports.
This is not true for Bannerlord. It's ridiculously complex. Let's consider some of the factors going into the Sturgia Brigands pick-rate:
- If a player dies in a round, they could pick Brigands again and raise the pick-rate. However, since they died the first life on Brigand, the Brigand is hypothetically weaker despite having a higher pickrate, which would classically suggest that the unit is stronger.
- A team with higher economy may run 3 Brigands while on low economy they run 2 Brigands. This increases the Brigand's pickrate in a decision based on an external factor (econ) instead of unit value.
- In another external factor, players may choose the Brigand simply because they have 110 gold left. However, this is only relevant if Sturgia is the played faction and you have 110 gold leftover. If you were playing Vlandia, you would of had to taken a different route to have 110 gold remaining, altering the frequency when you have 110g remaining.
In the end, you're left with something like a Brigand pickrate of 8.29 per round, which is only completely accurate when comparing Brigands to Brigands, and no other unit, so it's kinda useless. But it's all you have, so, keep it I guess?
Winrate isn't any better. You take the 8.29 average Brigands per round and weight the winrate based off it. You end up with a flat winrate that represents many complex interactions and unique decisions that went into choosing the Brigand.
The core problem here is that due to gold, some troops are played when down in econ (losing) and some troops are played more when up in econ (winning). Nerfs and buffs to units that are chosen for external reasons -- losing teams buying cheaper units and winning teams buying more expensive units -- isn't balancing based on unit impact to winning but instead removing the impact of econ in general.
Let's say that Brigands are used by winning teams and Varyags are used by losing teams. This is because the Brigand allows three spawns at 330g and the Varyag fits 300g. (My mistake if these aren't the correct values.) This is NOT because the Brigand is overpowered or because the Varyag is underpowered. This is simply because they fit their respective team's gold optimally. If you were to nerf Brigands because of their winrate and nerf Varyags because of their lossrate, you wouldn't be balancing the classes. You'd be making 300g closer in value to 330g.
Let's compare this to League, who are transparent in their data and balance framework.
Yeah. It's that simple.
League is thousands of times more complex (to put it lightly) than Bannerlord Skirmish, yet offers devs a clear balance direction.
(If worth noting, Riot has a secondary system to check and balance this framework. It evaluates additional "game health issues" to ensure champion diversity between levels of play and that champions don't get bunched up in some places and have low variety elsewhere.)
Unfortunately, the best method to balance Bannerlord is probably an amalgamation of winrate/pickrate -- and all the nightmares that entails -- and KDA/R -- which doesn't account for massive-impact low-KDA behaviors such as zone of threat, support potential (bowspeed/reload/damage may dramatically alter this factor due to light stuns on melee targets), cav bumps, bodyblocking (some horses do this much better than others), kicks that a teammate follows up on (movement speed, armor/shields impact success), and so on.
For example, in some cases stuns > damage. Such as a fast bow that constantly stuns with low damage compared to a slower bow with higher net DPS. Warband comes to mind where the Strong Bow had the highest DPS potential but other bows were favored due to speed. WARBAND BAD. Okay, sorry for mentioning it.
Due to all of the problems with KDA used in balancing you'd want to avoid it, but in Bannerlord, winrate/pickrate are both confuddled and poor numeric representatives of gamestate, but hopefully it's somewhere between the two.
Which gets back to the original statement: The class system is horrific to balance yet simultaneously offers players low strategic depth.
This ended up being really long for a reply. I might refine it and make it into a thread.