Heroic Difficulty Weighted Against Full Roster

fmftint

It's long been questioned whether or not heroic difficulty is weighted against your full roster or not, I've always though not. But what I am currently seeing, I'm positive that it IS

My available T3 average for this event is 303


However this node that should be rated "Normal" is rated "Easy" despite being higher than my T3 average


My Full roster, 360 OML (locked), 360 Phoenix (locked), 356 Red Hulk (available) averages out to 358, a differential which would rate that node as "Easy"

Conclusion, Heroic difficulty is weighted against your full roster

Soitswritten

I was thinking the exact same thing. Not having my best characters outside of rulk available is making it very difficult. Zero chance of not using health packs after every fight.

Bowgentle

Thank you.
I've been saying this for two years, regardless of the CS mails people were wheeling out that claimed only useable chars are considered for schaling.

y2fitzy

I had always soft capped my 3 star at 94, because I didn't want to get walloped in PVE. Only started again a couple of weeks back after a year off, have pushed a bunch of 3* to 127 and the difficulty of PVE has gone up MASSIVELY.

This heroic is a perfect example, losing multiple characters just unlocking the daily sub! So much for easy and trivial nodes...

Polares

Yeah I have also said the same thing for a while now. For Heroics it uses all your chars for calculating the levels of your opponents, not just the ones you can use in that particular Heroics PvE. Which of course is very unfair.

Slarow

Agreed. EOTS had a similar issue whereby OML was not available in the first 5 days, yet the scaling was influenced by him, causing a roster like mine (360 OML, 275HB, 275JG) to get destroyed.

mpqr7

That bothers me as well. I don't like being scaled to my 405 oml if I can't actually use him. The rest of my characters are in the 270s range or less.

Demiurge_Anthony

Hey everyone, I wanted to clear up a misconception on how difficulty scaling works with events that restrict your roster. Difficulty scaling takes into account which characters you are able to use, not any you are not able to use. I tested this internally.

I imported the roster I play with and here are the top 5 characters:


This is what my roster looks like in the next heroic mission (Dark Avengers: Heroic):


This is what the first mission that is scaled. This is what my roster looks like:


When I remove Old Man Logan from my roster (Selling 5-Stars is much less painful on a development build...), this is what my top 5 look like:


My Heroic roster did not change, and when I go back to the mission I showed earlier, the levels did not change at all:



It looks like something is still weird with the difficulty banner, and I will make sure there is a bug logged for that.

Thanks everyone, I hope this was helpful.

mpqr7

Thanks for showing us this with screenshots. It's very helpful to see examples of various scenarios that concern us! Especially because it's not as easy for us to sell of our high-powered characters to test ourselves

So it does look like the "trivial / easy" banner does change, as you noted. Glad to see that will be tested and fixed.

JVReal

Demiurge_Anthony wrote:

Hey everyone, I wanted to clear up a misconception on how difficulty scaling works with events that restrict your roster. Difficulty scaling takes into account which characters you are able to use, not any you are not able to use. I tested this internally.

I imported the roster I play with and here are the top 5 characters:


This is what my roster looks like in the next heroic mission (Dark Avengers: Heroic):


This is what the first mission that is scaled. This is what my roster looks like:


When I remove Old Man Logan from my roster (Selling 5-Stars is much less painful on a development build...), this is what my top 5 look like:


My Heroic roster did not change, and when I go back to the mission I showed earlier, the levels did not change at all:



It looks like something is still weird with the difficulty banner, and I will make sure there is a bug logged for that.

Thanks everyone, I hope this was helpful.

My only concern with the test you just performed is that it is my understanding that base scaling is determined when you first join, and doesn't change until you begin the next event. Even though you sold OML, you did not join the event new again to adjust the base of your scaling.

Perhaps you could pull that roster into a test environment, sell OML FIRST, then join the event and see how that same node is scaled. That would be helpful.

fmftint

Please understand, I'm not saying the enemies are based on your full roster, but the difficulty assesment is weighed against it.

313 vs 303 should be classified as "Normal"
I do agree this is a bug

The node in your example is displayed "Trivial" with OML on your roster but "Easy" without

Demiurge_Anthony

JVReal wrote:

My only concern with the test you just performed is that it is my understanding that base scaling is determined when you first join, and doesn't change until you begin the next event. Even though you sold OML, you did not join the event new again to adjust the base of your scaling.

Perhaps you could pull that roster into a test environment, sell OML FIRST, then join the event and see how that same node is scaled. That would be helpful.

I made sure that the scaling was recalculated when looking at it again. Special debug commands

DC1972

Top three in my usable roster is 276 Patch, 257 BP, and 254 KK. My Too little Too late node is marked as "Easy" with 243 Ares, 244 Sentry, and 243 Daken.

My alliance mates top three usable is 267 Patch, 250 BP and 250 3* Thor. His Too little Too late node is exactly the same with 243 Ares, 244 Sentry, and 243 Daken but it is marked as "Normal".

I was thinking that the full roster was taken into account because I couldn't see why there would be a difference if our top usable players were only an average of 7 levels apart.

I see from post above that it may be the difficulty assessment that is impacted by full roster.

BlackSheep101

This was super handy. Thanks, Anthony!

revizion

If you really want to test it, also delete your unavailable champed 4*s. Keeping them and only selling OML changes your top 3 avg by only 25pts (299ish to 274ish). Since none of them are available in this Heroic, it would give a much clearer picture I think.

Your available top 3 in the sub come out to an avg of lvl 218 when boosted. Isn't that what the nodes should scale from?

Bowgentle

Oh wow. Thanks for actually testing this out, Anthony.
That sort of interaction is very much appreciated.

Now excuse me, I have some crow to eat.

Polares

Mmm I don't know if this test is very useful Anthony, there have been not enough variation between both rosters.

Removing just OML is not going to alter your scaling that much when your other chars are 275+ In this test he was just 70 levels above all the other chars. For this test you should have sold all 250+ level 4s, not just OML. If after selling your IMHB, your Jean, your deadpool, etc. your scaling is still the same then we can say there is no doubt about it, but this very small change in your roster I don't think we can safely say what you are saying is right (more when scaling is also determined in part by your previous performances in PvE).

Also, are you using the new scaling or the old one? Because if it is the new one, your OML just has 6 covers.


In fact there is at least one difference, the node went from trivial to easy after you sold OML. So basically you proved there is difference and it was using all chars not just the featured ones.

Demiurge_Anthony

Challenge Accepted!

I sold all 4 and 5 Stars:


I then reset the scaling and checked the mission again:


To confirm that when I reset the scaling it updates correctly, I removed 2* Human Torch, 2* Thor, Colossus, Sam Wilson, 2* Magneto and Rocket and Groot and this is my new Heroic Roster:


I went into the mission and the scaling changed (phew!):

Phumade

Polares wrote:

Mmm I don't know if this test is very useful Anthony, there have been not enough variation between both rosters.

Removing just OML is not going to alter your scaling that much when your other chars are 275+ In this test he was just 70 levels above all the other chars. For this test you should have sold all 250+ level 4s, not just OML. If after selling your IMHB, your Jean, your deadpool, etc. your scaling is still the same then we can say there is no doubt about it, but this very small change in your roster I don't think we can safely say what you are saying is right (more when scaling is also determined in part by your previous performances in PvE).

Also, are you using the new scaling or the old one? Because if it is the new one, your OML just has 6 covers.


In fact there is at least one difference, the node went from trivial to easy after you sold OML. So basically you proved there is difference and it was using all chars not just the featured ones.

Sounds like Anthony is saying that the visual banner discription of what is easy trivial and hard may have a bug

But that the actual math that determines the pop levels is calculating correcting.

I'm okay with the situation if it's just a display bug and not a fundamental mistake in how base difficulty is being calculated

chamber44

those 4* and 5*'s you sold. Any chance you want to pass them on to me?

Lemminkäinen

Thank you, Anthony, that was absolutely brilliant!