Attempting to solve the scaling problem

OneLastGambit

This thread is for ideas, please feel free to add change or criticize anything posted.

The issue persists, 5* can spoil scaling and some scaling is outrageous (hello EotS) I've devised a very quick and simple model which I hope will solve the issue..here goes...

Scaling would be based upon the average level of your top 3 guys with a modifier added for each cover for those characters.

So base level = average of top 3
Modifer = adds 2 levels for each cover in those top 3.

Some examples...

Top 3 (my roster as it happens)
Base=300 + 274 + 271 = 845/3= 281.66
Mod= 13 +13+13 = 39x2= 78
Total max level for this roster= 359.66

Top 3 (a 2* transitioner with 5*)
Base = 255+255+144 = 654/3=218
Mid. = 1+1+13 = 15×2= 30
Total max for this roster = 238

This to me represents a fair challenge whilst not being impossible for any roster strength and also doesn't punish you so badly for pulling a 5* as a transitioner.

Ideas peeps?

simonsez

All I know is, each opponent shouldn't end up with twice as much health as each of my characters, and the game shouldn't be telling me, "Great, now beat them 4 more times".

Gmax101

Its tough...

Top 3 average means that for rosters with a couple of All Stars (be they 5* star in a 2* star world or just a couple of PVP MVP 4* Championed), the only viable team will quickly become those 3 characters...

At least if you run off un-boosted levels then the boosted characters might be useful.

The more characters you use to generate the average the more of the roster you make usable, but you reward the existence of having 1 or 2 MVP characters who are massively levelled over the average of the roster.

I don't know what the solution is... which is why scaling is an odd choice over a fixed level but tiered Gauntlet style approach. Getting it right for everyone is nigh on impossible.

Having said that, my preference for Scaling if it has to be done (ironically not the best for my actual roster given the relatively equal levels I have used) is to go with as high a number of characters as reasonably possible for the average... as at least then if I choose to aggressively level my favourite character, whoever it might be, he/she will always be an powerhouse... which is likely to make me feel good as a player/fan. As I am using the characters I have invested in, but can still use a relatively high number of characters from my roster during the event.

So I would take the average of the boosted characters... at their boosted level at the start of the event.

People with bigger rosters might have unboosted favourites that will be powerful.

People with bigger rosters (inc 2 & 3*) will have lower scaling than people with narrow high end rosters (4* +) and that enables more of the roster to be viable.

It will emphasise better rosters as the big 4* will be a powerhouse, which is fine, but also avoids the impact of 5* on scaling.

I wouldn't include the Essentials, if you have them then they are always going to be useful given that their boost is greater than the weekly boosts.

Sure the 2* with a 5* star in the roster will likely use the 5* for everything, but hey, that's still cool right?

It arguably penalises those without the 2* and 1* boosted characters, as they wont be pulling the average down, but then chances are you have a mostly high end roster if you have got rid of the 2* and it might even get people to roster the 2* anyway.

example:
Full Roster, max level but not champed, comes out at an average level of 235 vs the Booster 4* level of 360

Just having the 3 & 4* characters the average level is roughly 300. But these guys also probably have some developed 5* to power through, and non-boosted 4* Champs will be relevant/powerful

A 3* transitioner with underlevelled 4* and all 2* will be looking at roughly average of 220ish, meaning that all their boosted characters and any focused characters will always be useful.

As an idea it has some big issues... I have no doubt.

and Id really rather just have a tiered gauntlet style event with fixed difficulties, so you only get to do what your roster lets you...

Bowgentle

The only way to solve the scaling problem is to kill scaling.

Set levels, lock rosters out of content that is too hard, give them something to work for.
Sense of accomplishment when you're finally able to beat the hardest nodes.

Cap levels at 375 so that the 5* guys can have fun without being punished for being lucky / spending a gazillion bucks on LT draws.

No need to further complicate the byzantine labyrinth that scaling has become over the last 2 years.

mpqr7

One of the big problems with scaling is that enemy hit points and attack power jumps up in massive amounts, exponentially, I believe.

First, I'm facing enemies with 8000 hit points, then suddenly it's 12000, then 18000, then 25000... It's mind-boggling.

Attacks go from taking away 2000 to 4000 to 8000 to 140000 and so forth.. it just gets completely ridiculous. Why make a level that basically obliterates your team within a few moves? I'm not going to use healthpacks for pve. I'd rather just stop playing.

And then if you happen to win, you could win a crit boost or 70 iso.

hopper1979

I am about done with a roster power calculator, it takes your entire roster into account, it is similar to what you have here but includes everybody. I am making sure there is a linear relationship between power and what level you face so far it is very good, the biggest issue is data for higher power players. The major issue will be to get the scaling equations from people at different levels once I have presented everything, please keep your eyes posted and participate so we can solve this issue.

OneLastGambit

To resolve the issue perhaps the equation perhaps extending th3 average to top5 and also state that only the covers after the first add additional levels.

X= top 5 character levels /5
Y= total number of covers - 5

X+Y = total max scaling

I'm trying to keep the equation simple because the more factors you include the higher the margin for error.

I'd like to have a gated difficulty setting but I don't think that's the way d3 want to go so id rather try fixing a problem that might actually be taken into consideration.

GMax: id like to use the entire roster but sadly if you do this people will retain a litany of low level characters to decrease their scaling. Top 5 means any unlevelled characters rostered for the purpose of lowering scaling are not included.

Shemhazi

What about something like giving players the option to lock out certain certain characters at the beginning of a new Story mode event? Any characters chosen to be locked would not count toward scaling/matchmaking, but also wouldn't be available for the duration of that event.

Gmax101

OneLastGambit wrote:

GMax: id like to use the entire roster but sadly if you do this people will retain a litany of low level characters to decrease their scaling. Top 5 means any unlevelled characters rostered for the purpose of lowering scaling are not included.

I guess but that would be a fair number of additional roster slots....

like I said, it had flaws as a suggestion

OneLastGambit

Gmax101 wrote:

OneLastGambit wrote:

GMax: id like to use the entire roster but sadly if you do this people will retain a litany of low level characters to decrease their scaling. Top 5 means any unlevelled characters rostered for the purpose of lowering scaling are not included.

I guess but that would be a fair number of additional roster slots....

like I said, it had flaws as a suggestion

I've spoken to a number of whales GMax and funds are irrelevant to them (and by extension so are roster slots)

They would gladly spend x amount on 100 roster slots to decrease their scaling by x amount. When money is no object it is rarely a decisive factor in any decison. I would rather not have whales facing level 100 enemies taking up the t50 in my bracket while I face level 450 enemies

wgasadude

What about allowing players to bench characters - can't use, but won't count against mmr. Those that get 5* can hold onto their pulls and at the same time not be penalized.

revskip

I'd like to see scaling adjusted specifically to the team you take in to the match not the top three characters on your roster page. This way your entire roster is available insomuch as you can decide to mix and match with the boosted characters or bypass them entirely and use unboosted characters or even a team in the 2* or 3* realm for the 4* and 5* rosters. This would encourage wider rosters that are leveled and max covered which means more money spent on roster slots.

It would also alleviate the problem of having characters locked out of a PVE event either because they are present on the other team or because they are different iterations of the same character (ie Phoenix & Teen Jean).

Gmax101

OneLastGambit wrote:

Gmax101 wrote:

OneLastGambit wrote:

GMax: id like to use the entire roster but sadly if you do this people will retain a litany of low level characters to decrease their scaling. Top 5 means any unlevelled characters rostered for the purpose of lowering scaling are not included.

I guess but that would be a fair number of additional roster slots....

like I said, it had flaws as a suggestion

I've spoken to a number of whales GMax and funds are irrelevant to them (and by extension so are roster slots)

They would gladly spend x amount on 100 roster slots to decrease their scaling by x amount. When money is no object it is rarely a decisive factor in any decison. I would rather not have whales facing level 100 enemies taking up the t50 in my bracket while I face level 450 enemies

That is fair, and certainly a worry.

As I suggested in my post, I would probably argue just for including the highest level version of each boosted character. So having 10 level 1 versions of a boosted character means nothing.

Granted, rostering every 1 drop is not going to happen... even the 2 drops are unlikely, for most players.

I dunno to be honest, I wrote a suggestion that my perfect PVE is a tiered gauntlet-esque event. Progression only rewards and locked mission difficulties, ideally even locked options on characters to force roster use.

But if I have to have scaling, then I want a starting point that reflects most practical rosters.

Granted, the major argument is that in a placement based game the scaling logic would be abused to enable whales to burn through the event...

But if placement was not an issue, only progression, that complaint goes away I guess??