Would you agree that it is more likely to be a fair match if you match based on the top third of the roster's strength than the bottom third's strength?
Let's use a dating app/site as an example. These companies have algorithms to match people, and a good algorithm should put more weight into more important characteristics/preferences individuals are looking for, like religion/values as opposed to less important traits/preferences, like which toothpaste they use. (Maybe Nexon should consult this industry. )
The same thing is true for war matchmaking. Making sure the top 3rd is evenly matched is more important for a fair match and so should get more weight in the calculation. This does not mean that the algorithm has to ignore the middle third or the bottom third, but just that the top should be more of a deciding factor in pairing alliances for war. The middle third should also prob be weighted more than the bottom third.
It doesn't have to be thirds, it can be broken down differently, but the idea is to match more based on the top because let's be honest, that is what matters. If one side doesn't have nearly as much power as the other side at the top, it is a lopsided match.
See, this would've been true before the time tie breaker was implemented, top players could pull off perfect score and the match would end in a tie. Now a lot more attacks matter if it comes to perfect score, and I would make it approximate that the upper 60-70% of the team has around equal importance, with importance slightly diminishing as we go downwards.
In the case of making parts of the team less important and counted in matchmaking, it would again be a simillar problem as the 'just cut out 40% bottom from matchmaking'. Teams would put as much power as they could in those places where that power would be calculated as less than it really is. More closely-leveled teams that have low variation of levels would destroy mixed teams. Replace one problem with another.
I think this solution is what many players are arguing for, and I don't see why figuring out how to do this and implementing it would take over a year...
There have been other solutions proposed as well but no noticeable actions taken. The problem isn't sandbagging. Sandbagging is a behavior that comes from the real problem which is how Nexon is calculating war match ups.
You cannot calculate the perfect match in a system with 2 attacks, and making 1 attack is too risky if not paired with some exceptions like having limited 2 attempts on one base.
And you cannot stray of the alghoritm of averages (currently used) because you would replace one problem with another, and the current system doesn't place confusing and straining rules for the matchmaker.
The only way to do this while keeping the 2 attack system is not to change the alghoritm, but to limit the things that can compromise the alghoritm, like making it only viable to sandbag with Gunpowder/EA bases with some rules. It will have a less effect that sandbagging with Iron Age bases.