Official algorithm for scaling

Tromb2ch2
Tromb2ch2 Posts: 301 Mover and Shaker
edited December 2016 in MPQ General Discussion
I'm curious about what the algorithm is. At one point I was told it had to do with the average of your 3 highest boosted character for the event. Later we were told it was changed but not really told how it was changed.

So today I got examples of top 3 boosted character from 4 different player and used the required 4* character node in the D.C. Sub as the reference point.

Player A:
276, 275, 274 for an average of 275
Enemy level 304 for a difference of -29

Player B:
376, 297, 282 for an average of 318 1/3
Enemy level 306 for a difference of +12

Player C:
369, 295, 294 for an average of 319 1/3
Enemy level of 304 for a difference of +15

Player D:
377, 296, 293 for an average of 322
Enemy level 306 for a difference of +16

One thing that I found out was that it doesn't seem like boosted characters really change the scaling. This is kind of odd to me I thought it was supposed to (not that I'm complaining at all).

Also player C had an average of 1 higher than B but player B's enemies were 2 level highers. And player D's average was about 4 levels higher than player B but their enemies are the same level.

All 4 players highest non boosted characters are about the same level (around 274 with player D being around 278)

And I know a couple of levels isn't a big deal but when the enemies scale almost exponentially it gets to be a problem at higher levels.

So I'm curious as to see what the formula is or at least what all things it looks for.
«1

Comments

  • Polares
    Polares Posts: 2,643 Chairperson of the Boards
    Oooohhh scaling my favourite thing in the world !!! icon_razz.gif

    Player E:
    455, 452, 451 for an average of 452 1/3
    Enemy level of 435 for a difference of +17

    If 4 chars
    455, 452, 451, 383 average of 435 1/4
    Enemy level of 435 for a difference of 0

    - One player mentioned in the forum that someone in CS told him that for the average they use 4 chars. But this is totally unconfirmed. I asked him about it, but he didn't answer me. BUT in my particular case it looks accurate.
    - In my experience Buffed characters do affect scaling (I can see different scalings just in Heroics, but depending which 4s are buffed my scaling is different, so it definitely affects scaling).
    - Covers are supposed to affect scaling. I think this might be used as a multiplier. i.e 13 covers level * 1.0, but if just 3 covers then level * 0.95
    - Scaling looks to be a curve, it is much higher in lowest and highest levels.

    PS: If this is true and 4 chars are used, levelling just two 5s, would make scaling still nice and the strength of the team would be almost as good as someone with 4 champed 5s. This is also seems to agree with what people reports in the forum.
  • Suddenreal
    Suddenreal Posts: 92 Match Maker
    Scaling isn't dependant solely on the boosted characters, but takes in account the boosted level instead of the actual level. Also, I believe the scaling ignores your highest three characters (so low level players still have a fighting chance in the event if they have a 5 star.png or if they have a boosted yet unusable character).
  • fmftint
    fmftint Posts: 3,653 Chairperson of the Boards
    edited November 2016
    My T4 450, 435, 420, 420, average 431(T3 435)
    4 star.png essential tops out at 403, -28 (-32)
    Hard command point node tops out at 418, -13 (-17)
  • Tromb2ch2
    Tromb2ch2 Posts: 301 Mover and Shaker
    Well I'm not sure about players b-d but I was a in this example and my next highest is 274 so my average of 4 is 274.75 which still puts me at about 30 levels lower than my enemies (ouch). But yeah my enemies were the same level as people with much higher boosted characters so I'm not sure that factors in (which good, why penalize you for having a boosted character).
  • JVReal
    JVReal Posts: 1,884 Chairperson of the Boards
    edited November 2016
    Tromb2ch2 wrote:
    Player A:
    276, 275, 274 for an average of 275
    Enemy level 304 for a difference of -29

    Player B:
    376, 297, 282 for an average of 318 1/3
    Enemy level 306 for a difference of +12

    Player C:
    369, 295, 294 for an average of 319 1/3
    Enemy level of 304 for a difference of +15

    Player D:
    377, 296, 293 for an average of 322
    Enemy level 306 for a difference of +16
    The thing that bothers me about this is that my roster is closer to Player A... an average of 275, and Player A is receiving the same scaling as Player C despite having an average that is 44 levels less. What your numbers don't specify is whether this is initial scaling when the sub started or these are the levels after 4 clears... were they all on the same clear?

    I have begun to see this month my scaling increase into the high 300's as my 3* characters started to break 200 and I began to have more Champed 4*'s and more unleveled 5*'s. I doubt they have any correlation, but it's the only thing that has changed as regards to levels. To me it has to be a change in the scaling calculations.

    EDIT: Someone in another thread reminded me that the last scaling tweaks took into account number of covers when calculating scaling... this makes sense since my max covered 4*'s and my number of 5* covers have been steadily increasing faster than the levels of my top 3 or 4 characters.
  • Chipster22
    Chipster22 Posts: 299 Mover and Shaker
    On the DC required 4* node my top 4 are 396, 390, 390, and 388. One of the 390's is a 5*, the rest are boosted 4*s.

    The enemies top out at 334.
  • STOPTHIS
    STOPTHIS Posts: 781 Critical Contributor
    I don't think it has anything to do with boosted characters anymore. Could be wrong, but my (admittedly not great) tracking of my scaling only seems to point towards non-boosted levels. I noticed it when I first made the jump to the 4* level. I didn't keep great track of where scaling was for me before it was somewhere around 270-280 for the hardest node. In Unstable ISO-8 and both Fury and IMHB were boosted. Over the course of the event I champed Ant-Man, Fury, IMHB, Cyclops, and Rulk. The first 2 were AM and Fury.

    With 2 champed 4* and 1 was boosted (I only kept track of highest enemy level, sorry):
    Hardest node: 304
    4* required/2nd hardest nonessential: 289
    3* required: 269

    By the end of the event I had 5 champed 4* and 2 were boosted:
    Hardest: 318
    4* required/2nd hardest nonessential: 303
    3* required: 291

    Thing is next event (Which according to the Compendium of PVE Thread was The Hunt) I had 0 boosted 4*s and my scaling was exactly the same. So it appears that my scaling creeped up as my champed 4* pushed the average level of my top 4-5 characters from 255 to 270.

    It kind of sucks on weeks where none of my 4*s are boosted, but it's just motivating me to want champ more 4*s.
  • JamieMadrox
    JamieMadrox Posts: 1,798 Chairperson of the Boards
    Nobody except Demiurge knows how scaling is calculated, but it's safe to say that there are a number of factors that determine what levels you face in an event and it likely changes based on the event, etc. Same with MMR. Nobody has a sweet clue how it works and anyone that says otherwise is just speculating.
  • BassMuffinFIve
    BassMuffinFIve Posts: 231
    Jamie Madrox wrote:
    Nobody except Demiurge knows how scaling is calculated, but it's safe to say that there are a number of factors that determine what levels you face in an event and it likely changes based on the event, etc. Same with MMR. Nobody has a sweet clue how it works and anyone that says otherwise is just speculating.

    Maybe it's time they released that information then? I think 3 years of suffering trying to guess their formula is enough, don't you?? This should be public knowledge. There is not a single legitimate reason why they haven't shared it with the playerbase yet. Not one. Time to get off your high horse DEMI!
  • Ducky
    Ducky Posts: 2,255 Community Moderator
    BassMuffinFIve wrote:
    Jamie Madrox wrote:
    Nobody except Demiurge knows how scaling is calculated, but it's safe to say that there are a number of factors that determine what levels you face in an event and it likely changes based on the event, etc. Same with MMR. Nobody has a sweet clue how it works and anyone that says otherwise is just speculating.

    Maybe it's time they released that information then? I think 3 years of suffering trying to guess their formula is enough, don't you?? This should be public knowledge. There is not a single legitimate reason why they haven't shared it with the playerbase yet. Not one. Time to get off your high horse DEMI!

    They'll never release how scaling or MMR works because then ppl could game the system even more, which is something that they don't want happening.
  • BassMuffinFIve
    BassMuffinFIve Posts: 231
    DuckyV wrote:

    They'll never release how scaling or MMR works because then ppl could game the system even more, which is something that they don't want happening.

    Not releasing code because someone might break it just sounds like an excuse not to fix **** code. "As long as we never release it to the public, no one can tell us it's broken..."

    Yep. Just keep patting yourselves on the back guys.

    I refuse to pay to be their beta testers. Either hire them, or release the code. So. Simple.

    Who said anything about breaking code? I was referring to softcappers. It would allow people to build their rosters in a way to avoid scaling and MMR issues altogether and could have lasting effects on the meta (PvP and PvE) if exact parameters were known.
  • BassMuffinFIve
    BassMuffinFIve Posts: 231
    aesthetocyst wrote:

    Whew. Someone missed the whole "tanking" era. Answering:

    No.
    Stop guessing. Why do you need to know? Just play.
    No.
    Other than maintaining anything resembling an even playing field for all players, whether casual or competitive, well-informed or blissfully unaware, and protecting the game economy, no reason at all.

    Not sure what you mean by tanking era, but I have been here since day 1. I don't know how you play games, but being a gamer to me means discovering the hidden in the game, discovering the unknown, the combos, the moves, the strategies, the formulas; whether discovering them for yourself or delving deeper and doing research to find that knowledge. If this is a mindless slot machine grind for you, then more power to you.

    An even playing field and a stable economy can be maintained by good game design and programming. You seem to be very loosely defending their decisions to purposefully hide information from their playerbase, and they just don't make any sense. Hiding information means they don't have to fix anything. Ever. Because nothing is broken, right?
  • nyck1118
    nyck1118 Posts: 106 Tile Toppler
    I'M PLAYER B
  • Tromb2ch2
    Tromb2ch2 Posts: 301 Mover and Shaker
    Well I was more interested in finding out how scaling is determined because in the examples i provided I was at a disadvantage compared to my 3 other alliance mate. And my scaling in pretty much every event (except heroics) cap out at levels higher than my characters (I see red banners every event).
  • monsieurmojo
    monsieurmojo Posts: 370 Mover and Shaker
    Just for data collection's sake: Hell's Kitchen 4* essential node level after 4 clears: 309

    Highest 4 characters, unboosted: 330 (8/13) Bolt, 273 Jean Grey, 272 Daisy, 271 Hulkbuster. Average level 286.5 (+22.5)

    Highest 4 characters, boosted: 330 (8/13) Bolt, 286 Loki, 276 CMags, 273 JG. Average level 291.25 (+17.75)
  • Tromb2ch2
    Tromb2ch2 Posts: 301 Mover and Shaker
    aesthetocyst wrote:

    Describe the players' approach to / consistency in PVE and their typical placement?

    Also, are their highest characters fully covered / champed?

    Clearly something different about player A. Is he a wizard?
    I usually try to place t10 (except 7 day events) I play in cl 8 and the top 3 are champed flaptain, venom, devil dinosaur. And if I am a wizard I clearly am using my powers wrong lol.
  • beyonderbub
    beyonderbub Posts: 662 Critical Contributor
    edited November 2016
    Data: Washington DC - 4* essential node level after 4 clears: 321

    Highest 4 characters, unboosted: 340 (9/13) OML, 340 Jean Grey, 302 XForce Wolverine, 298 Devil Dinosaur. Average level 320(+1)

    Highest 4 characters, boosted: 373 X-23, 340 (9/13) OML, 340 Jean Grey, 315 CMags. Average level 342 (-21)

    Not really sure if I switched the + or - and I'm not sure if these results are favorable or what it means. Do I want my scaling to be about even with my team? All I know was that I didn't use any health packs to clear 4x bc X23/OML/Carnage cleaned up using X23's green and purple; threw out OML's black now and then. So scaling for my progress was not an issue bc I had the 'right' characters to clear the node. Usually not the case.
  • Uninspired
    Uninspired Posts: 36 Just Dropped In
    I always assumed the essential nodes put special weight on the essential character for scaling. Possibly something like the average of essential level + top two instead of average of your top 3. Does that possibly impact the stats that should be collected, especially where the essential isn't a top character on someone's roster?
  • amusingfoo1
    amusingfoo1 Posts: 597 Critical Contributor
    monsieurmojo wrote:
    Just for data collection's sake: Hell's Kitchen 4* essential node level after 4 clears: 309

    Highest 4 characters, unboosted: 330 (8/13) Bolt, 273 Jean Grey, 272 Daisy, 271 Hulkbuster. Average level 286.5 (+22.5)

    Highest 4 characters, boosted: 330 (8/13) Bolt, 286 Loki, 276 CMags, 273 JG. Average level 291.25 (+17.75)

    For comparison, on that same node: 432

    Highest four characters (boost-independent): 466, 463, 460, 460
  • fmftint
    fmftint Posts: 3,653 Chairperson of the Boards
    Uninspired wrote:
    I always assumed the essential nodes put special weight on the essential character for scaling. Possibly something like the average of essential level + top two instead of average of your top 3. Does that possibly impact the stats that should be collected, especially where the essential isn't a top character on someone's roster?
    Any post release event says, NO!

    Scaling on the 4 star.png essential was the same for Agent Venom as it is for any other

Categories