PSA: How to Have Your Cake and Eat It Too: PvE Scaling

Vhailorx

TLCstormz wrote:

Lord. This is all soooooo confusing!

I am too scared to even level up my Loki and my Grocket to level 94, all because I'm a very early transitioner, and the rest of my roster struggles A LOT at their current levels of 94.

Sort of off topic, but remember not to just level up whichever 3* has the most covers, R&G has true healing, so he should tank as much as possible. Loki is squishy and his powers don't scale much with levels, so he is generally better hiding behind someone else.

Unknown

Isn't taking a ton of damage to lower your scaling essentially the same as purposefully losing in PVE? Tanking is faster. I think we also overestimate the skill of the average player, whose C and D teams just aren't going to cut it a lot of nodes. The characters that a transition player can level are probably going to be their best characters and telling them to not use those characters in PVE seems like taking a significant penalty for leveling up the character. Unless you've got a very diverse roster, there's going to be a bit of PVE/PVP team overlap. Further, with 3*s the power tends to come with covers vs levels, so a max cover 3* at 110 vs enemy is probably going to be easier to handle than if you had the same guy at lvl166 vs the same enemy + 50% lvls. I'm pretty certain scaling is percentage based, but someone can correct me if that's wrong.

NorthernPolarity

daibar wrote:

Isn't taking a ton of damage to lower your scaling essentially the same as purposefully losing in PVE? Tanking is faster. I think we also overestimate the skill of the average player, whose C and D teams just aren't going to cut it a lot of nodes. The characters that a transition player can level are probably going to be their best characters and telling them to not use those characters in PVE seems like taking a significant penalty for leveling up the character. Unless you've got a very diverse roster, there's going to be a bit of PVE/PVP team overlap. Further, with 3*s the power tends to come with covers vs levels, so a max cover 3* at 110 vs enemy is probably going to be easier to handle than if you had the same guy at lvl166 vs the same enemy + 50% lvls. I'm pretty certain scaling is percentage based, but someone can correct me if that's wrong.

But if you think about how scaling works in the first place, it shouldn't matter at all. If all of the sudden you stop using your A team and are struggling to beat stuff with your B team, then PvE should eventually scale down and get easier so that you stop wiping with your B team. That's the beauty of this strategy and scaling: it doesn't matter if your B team is weak, because the enemies get weaker to compensate. Scaling works both ways!

TLCstormz

Northern, I have the main eight 2*s at level 94 and only a few 3*s with decent covers.

Deadpool 2 / 2 / 4, can go to level 102, currently.
Grocket 3 / 0 / 2, can go to level 76, currently.
MoStorm 2 / 1 / 4, can go to level 91, currently.

Other than the fact that I should NOT be pouring Iso into them until I have at least 10 covers for them each......would bumping them up screw me over, scaling wise? I've heard a lot of people say not to take 3*s past level 110 when you're transitioning, but I wanna hear it from you.

Trisul

For the record, PvE tanking clearly works.

I've tested this on multiple PvEs suiciding level 40 teams into Normal-classified nodes. It takes time (sometimes 6+ fights) but it has always worked when I've tried it.

What does this say about scaling? I think Phantron is wrong, and missing health still is the primary scaling factor (this is what IceIX re-iterated to NP), and for all matches, not just wins. Otherwise, we would have heard of people running true healers that have crazy reverse scaling.

So if one wanted to keep scaling down and optimize, one would suicide their entire roster into the PvE as often as possible. But it's so time-consuming that I would argue it's not worth it.

All that said, I'm going to test NP's theory and try to refrain from using my main PvP team in PvE to see if it will in fact control scaling.

veneretio

So if damage taken is the primary scaling factor... running 2* Magneto + Storm lineups is probably why 2* players see ridiculous scaling similar to 3* teams at times?

action711

I find the best thing to do is fight with the lowest guys you can to win the fight, and to have a large roster with quite a few slots filled with unlevelled 1*s

action711

Also, fighting with badly damaged heroes seems to work too, or 1 very badly damaged hero and 2 at about 2/3rds

Unknown

Because I've seen this too often: I agree that once you have a functional 2* PVP team, keeping your 2* at anything lower than 94 is overall hurting you. The minor amount of scaling from 60 or 70 to 94 is not at all worth keeping your characters low, and it makes the difficulty of getting to 400 in PVP a lot harder particularly due to incoming attacks.

Tarheelmax

Update one: I took for guys to between 110 and 120. I haven't seen any difference in my pve scaling. I haven't used any of them in pve.

Unknown

I tested this in The Hunt running a team of X Force (270), Iron Fist (127), and Mystique (73). At one point Last Dance was at 250 and now it's at 247. Note that the levels should still be going up overall due to community scaling, as it was easy to observe levels of everything spontaneously going up without doing anything after a period of inactivity. It is impossible to tell how much damage was taken with this team compared to say, Thor + Loki + IF but most of the games ended in 3-4 turns with an infinite combo with Iron Fist combo into Surgical Strike (a lot of DAs have purple as their color for Surgical Strike). I also subbed in Daredevil (129) since these guys are pretty low level versus the level 250 guys I'm usually fighting and attrition damage does pile up quickly.

It is not possible to conclusively say whether the level decreased due to lower level character used or that additional damage taken (if damage taken is measured as a % of health). Still, I find it pretty skeptical damage taken alone can account for a decrease in level because that'd mean there would have to be someone out there who is genuinely very bad at the game that'd have no problem taking an incredible amount of damage and then easily winning due to lower scaling, or at least placing much higher than can be expected. Note that you must win with your lower level team. I also cannot prove whether using a low level team on trivial nodes have any effect, but it seems like a safe thing to try when you're not pressed for time because it's not like it really matters whether you use a 2* or X Force + Thor on those. Also note that it takes a lot of games before this does take an effect. I saw my first decrease in levels by the second pass of the nodes, and I didn't bother getting a team lower than 166X3 for the 3 trivial nodes in Alaska because I was slightly pressed for time even though they could've been easily beaten with my 2*s.

This seems to suggest you want to keep your 4*s at a very low level for PvE and then use them, since they lose very little ability damage as a function of lack of levels. I think it's really backwards to punish you for using higher level characters, but I guess the strategy for PvE is that you can level your characters as long as you don't use them (or have a duplicate copy of them at a lower level).

Trisul

Instead of worrying about taking damage, if you're looking to optimize scaling reduction as much as possible. just tank your entire non-usable roster after each clear. Tanking is proven to work at reducing personal scaling.

(Someone suggested that Trivial nodes affect scaling much less than other nodes, so I would tank Normal/Hard nodes.)

Finishing your nodes with your best team ASAP then tanking will be more time-efficient than worrying about using evenly matched compositions. Tanking takes a lot of time, but purposely taking damage will likely end up taking much more time for less effect.

Of course, it's debatable whether manipulating scaling this way is really worth any time spent at all, but for min-maxers (especially those without usable 4hor) this might be useful.

Unknown

We'll know soon, since I plan on using winfinite a whole lot when I get my hands on some profX covers, and my team will be 94/100/70

Vankysher

NorthernPolarity wrote:

But if you think about how scaling works in the first place, it shouldn't matter at all. If all of the sudden you stop using your A team and are struggling to beat stuff with your B team, then PvE should eventually scale down and get easier so that you stop wiping with your B team. That's the beauty of this strategy and scaling: it doesn't matter if your B team is weak, because the enemies get weaker to compensate. Scaling works both ways!

I tried this in The Hunt nodes where the DA team were 20 levels higher than my non-boosted A team of Ares/OBW/AWolv.
Match ended very badly as they wiped my entire team and I only managed to do 50% damage to Yelena.
My non-boosted A team can handle the "Easy" rated nodes with some difficulty depending on board composition but cannot do a thing above that and the node levels have not gone down despite wiping twice in them. Don't know if the node isn't going down due to community scaling but I suppose scaling people out is what they want to do to help score stratification.

UPDATE: Node LMD Assault was rated "Normal" at 109/110/110 so I went in with Cage/GSBW/Grocket instead who were boosted to 108/130/108 respectively. Wiped the first time and the node levels stayed the same. Used 3 health packs & went in with the same team iso-boosted all colors. Won with Grocket down, GSBW with ~800 health & Cage ~2000 health. Node level now shows "Easy" 89/90/90.

Unknown

I suspect the game is looking for 'quality losses' and 'quality barely wins'. I know there was a time where straight up dying always worked but people stopped doing that for a reason. You used to be able to see very tangible results for dying over and over but this doesn't work reliably anymore which is why people stopped doing it. It may still work in some capacity but I suspect it probably takes about as much time as trying to beat the nodes themselves. Even when tanking worked straight up, it always felt to me that the game valued a close game with lower level characters more than a higher one. That is, if you bring out Ares, do 3 Sunders in a row and barely have any health and win, that's a lot more effective than bringing pre nerf Magneto, kill all but one guy and then just stand there and let the enemy slap you for a ton of damage. Since the game almost certainly cannot analyze the type of game, it seems to me a close call with a lower level character is valued more than a high level one. If scaling is meant to help the little guys, it'd make sense as running with a level 270 and the game never seems to care how badly you got owned on the times you do lose, because you got a level 270! But if your level 94 guy gets beaten up a few times the game seems to note that. I have no shortage of games where X Force got completely obliterated by overscaled enemies and I can't recall a time where this has led to a reduction in scaling.

DCUDCU

What about the 3* required node characters? If you have a lot of 3* and can move them to say 110 but not 166, you'll have to use them for some point for a PVE.

squirrel1120

There also gets to be a certain point where this just isn't the best course of action anymore. I mean, I can remember quite a few nodes as a 2* player that I just simply couldn't beat because I didn't have a fast enough combo (feeder nodes), the correct counters (no viable loki vs c.mags/im35/don), and other instances where things just simply scaled past where they could be beaten due to community shenannigans. I love the idea in the OP, and hope it is indeed true and proves to be confirmed. I just don't think I'd personally want to spend a massive amount of time to try to get my scaling back down to what I saw as a 2* player, only to be having to bring in some of my bigger guns again just to get some of the nodes done. #1 in pve just isn't all that viable if you have even one node you have to skip consistently...

Unknown

Well, keep in mind that community scaling does set a floor on how low things can go. If you've ever walked away from the game and came back and found everything went up by 50 levels, that's certainly not because someone secretly hacked into your account and was winning games for you in the wrong way while you're away. An event like Simulator Hard, Prodigal Sun, or the end of Gauntlet sets the floor way higher than a normal event and no amount of roster clever tricks can get around that. In general strong roster tend to dominate events where the community scaling is set at a very high level (the 3 mentioned above) while weaker roster tend to dominate the other events.

Zen808

Some thoughts:

Phantron wrote:

That is, if you bring out Ares, do 3 Sunders in a row and barely have any health and win, that's a lot more effective than bringing pre nerf Magneto, kill all but one guy and then just stand there and let the enemy slap you for a ton of damage. Since the game almost certainly cannot analyze the type of game, it seems to me a close call with a lower level character is valued more than a high level one.

Re: Simulator questions
Postby IceIX » Fri Mar 28, 2014 9:44 am

jozier wrote:
I see. More so that you end up with collecting a lot of AP (because you start > 0) and probably take less damage and deal more total damage?

Exactly.

My guess would be that if you "just stand there and let the enemy slap you," you're still building AP. When the algo looks at your match results, and sees a lot of damage, but also a lot of banked AP, it probably (and rightly so) assumes sandbagging, and adjusts scaling accordingly.

squirrel1120 wrote:

I just don't think I'd personally want to spend a massive amount of time to try to get my scaling back down to what I saw as a 2* player, only to be having to bring in some of my bigger guns again just to get some of the nodes done. #1 in pve just isn't all that viable if you have even one node you have to skip consistently...

Scaling mitigation is really a more viable strategy to casuals than to hardcore players. Trying to go for #1 in PVE, and also reduce your scaling is like trying to take #1 in PVP while avoiding X-Thor.

Trisul wrote:

Finishing your nodes with your best team ASAP then tanking will be more time-efficient than worrying about using evenly matched compositions. Tanking takes a lot of time, but purposely taking damage will likely end up taking much more time for less effect.

The problem with this is that it also contributes to burnout. 1) you're using the same 3 guys, following the same strategy match in and match out, and compounding that with 2) the dull tedium of waiting for the AI to kill your cannon fodder.

Again, under the assumption that Top 50 in a 3* reward tourney (let's just leave tinykitty 4* PVE's for another rant thread) is your modest goal, if you're mixing/matching between 8-10 heroes, the game becomes a little less of a grind. For my team, Xforce, LThor, and LCap are my best PVE guys. And I almost never use all three on the same team. By mixing in guys like Sentry, Psylocke, Punisher, Daredevil, the boosted 2* guys, etc, you will organically take more damage, your matches will last longer, and you'll take less of a hit to your scaling as a result. An ancillary benefit is that it also puts you less at the mercy of a bad cascade, because instead of burning three health packs, you just load up another team and keep chugging along.

Just my two cents.

Scoregasms

I had high hopes originally and went out to try this out and after 3 weeks, can say using lower level characters instead of your top characters doesn't help with lowering scaling. Maybe it takes more than 3 weeks to accomplish this? But if it does, that seems impractical for most players.

"Normal" level end nodes usually started off at the same level as my top characters even though I only used characters 80+ levels lower the whole time. I have someone else in my alliance that shares similar scaling to me and using lower level only characters never gave me a lower starting level. In fact, I didn't even play one sub the first 24 hours and decided to give it a shot to see what the last node "Reinforced Steel" looked like (Mystique sub in Rocket PVE). After my first pass that node was at 213/214/213 which is the same as my alliance mate after multiple clears, so no help in starting levels and apparently community scaling kicks in even if you didn't play the sub at all (lol), for reference, he said it started a lot lower for enemy levels which I suspect would've been the same for me if I played it from the start.

I normally finish top 2 in most pve's and have now been finishing bottom 100 (or worse) for 3 weeks now. I am honestly just sort of at a loss for how this works. I am interested to see how pve scaling changes (if at all) for those who sold off their 4* and now only have max 3* as their top roster.

Thankfully, Sim is rewarding lazy cap, so I have one more pve to basically tank before I give up on trying to lower my scaling and just go back to using my top level characters.

Anyway, would also be interested if others tried this and got different results than me.