I don’t think it’s ‘punishing’ wins per se. The idea (as I understand it) behind the DDA is that it adjusts itself to your level of play. So if you’re doing well, it makes the game more challenging for you, and if you’re doing badly, it makes the game easier.
As a concept, that’s a fine idea. Problem is, the way it’s counting wins & losses at the moment takes no account of the amount of times a player gives up on a Mission and starts again. So if you Restart 5 times until you get it ‘right’, it’s registering that as a win on a single playthrough and adjusting accordingly. In other words a player who takes 5 playthroughs to register a win is being treated as if he has a record of 1 win and 0 losses rather than 1 win and 4 losses.
I don’t think we ever got confirmation that such piece of code is actually used by the game. But even if it does, I agree with you, it would work in something like an Ironman run but not in cases when you load to the beginning to try to do better.
I personally would prefer some reference difficulty that goes up or down based on the strategic events, so people have at least some form of reference to the difficulty and can share their tactics/strategy otherwise why even have difficulty level choice? Just tune everything is accordance to have player is playing. It’s just this is not too far from having rubber banding and suddenly crabs being less precise and etc. I mean, how far does one wants to stretch this?
Dynamic difficulty can become a hell for modders too, as how do you judge if content that you make is suitable for mass of players, while only few would be able to test it for you.
It’s not that difficulty scaling is something new, on start difficulty scaling share some aspect with auto scaling. Both are closely linked to difficulty management.
So difficulty scaling shown it needs options:
No scale down, is an absolute required option.
No scale up limit is a good option.
If I believe have found a way to manipulate the auto scaling in first 1.5 month and it really seems so. But watching Odd let’s play longer makes me wonder on what the auto scaling is based. I can only admit he is quite better, for now he lost only one soldier and that is already hard to believe, and he is still not getting huge spikes I could have seen soon in Normal and Easy attempts.
At this point I’m starting to believe that it’s a lot about lost time or rush.
Auto scaling looks like a very difficulty problem, some players post quoted that it is supposed to target the problem of difficulty decrease along a campaign. But it clearly cause other problems even if we forget some current tuning problems, like probably too step scale up, wrong parameters, system not enough adapted to each difficulty level.
Have a good play that doesn’t degenerate into an easy rush, this system is probably perfectly adapted if well tuned. The system will scale up and down but will ensure scaling up enough at least sometimes. Then the problem is relying on parameters and difficulty feeling, and manipulating the system.
Have challenge play is perhaps doable with tuning or options on scaling down, but it’s still not a reference play.
Have a reference play, never scale down seems fair enough, but then manipulating the system will still be an important part. Moreover never scale down means a higher tuning.
That’s it or a campaign without any auto scaling which means more tuning to do.