Ascension math question

tiomono

Has anyone done the math on value per cover for the different star tiers as you ascend characters? Like how much cp, iso, hp, shards, per cover across different tiers comparing ascending vs just champ recycling. I have tried searching and I am struggling to find what I am looking for.

meadowsweet

I pride myself as a mathy person (I was at the forefront of the Champions 1.0 Farming debate way back when), but I haven't bothered to run the numbers for myself yet because my desire to Ascend characters to 5★, Levels 450-550 is so much more powerful than knowing that I might be missing out on ±10% of one resource or another. That's also why I've been Ascending everyone as Max-Min combinations, because I want to minimize roster slots in use and minimize the time / covers to get characters to 550. I'm not even sure this game will still exist in the time it will take me to 550 anyone, but that's why I also don't want to slow it down any.

entrailbucket

@meadowsweet said:
I pride myself as a mathy person (I was at the forefront of the Champions 1.0 Farming debate way back when), but I haven't bothered to run the numbers for myself yet because my desire to Ascend characters to 5★, Levels 450-550 is so much more powerful than knowing that I might be missing out on ±10% of one resource or another. That's also why I've been Ascending everyone as Max-Min combinations, because I want to minimize roster slots in use and minimize the time / covers to get characters to 550. I'm not even sure this game will still exist in the time it will take me to 550 anyone, but that's why I also don't want to slow it down any.

Agreed 100%. Also, a ton of characters are getting +100 levels every week, so if you can get them to 450...well, when they're boosted, they're 550.

dianetics

The math was done. The covers have been counted. The rewards per level have been looked at.
Rewards are greater the higher you go for your characters.

But if you want character specific rewards don't ascend your toons. If you ascended toons at max champ you lose covers and levels...ie higher level rewards.

Jumping a 1 to a 5 or a 2 to a 5 will take thousands of covers over the life...but the rewards get better the higher they get.....ie pull a 1* and get 25 cp type of situation.

The short game is painful the long game is bountiful.

All of this math has been done and posted when ascending became a thing. It is on the forums. go find it.

meadowsweet

@tiomono

I have tried searching and I am struggling to find what I am looking for.

@dianetics

It is on the forums. go find it.

Not super helpful?

bluewolf

I wish I could find a post someone had done but everyone who has dug into it says you come out ahead even when you consider the fact that the number of rewards per cover drop.

The main question would be what you want to get; what thing will make the biggest difference to your roster? and what tier are you focused on?

You have basically asked if anyone has done the math on every base tier and every level they get ascended to and what the ROI per cover, or per 100 or 200 or 300 or 400 covers, is.

One thing you get with ascension - once you get someone into the 3/4/5 tier - is some QOL return when you don't need to farm and free up roster slots for a while as you just add covers/levels. AKA initial roster setup can be a bit higher but then you have less to worry about for a longer period of time while collecting covers for that character vs farming.

Also the lowest tiers can be downright annoying to build when RNG isn't being your friend and you can't shard them and you have a 5/5/2 with 13 saved covers. Ascension reduces the frequency of that circumstance.

tiomono

I guess cp and iso would be the biggest mark on my roster. I'm in 5* land with my highest leveled being 467, and have only maxed one 4*, coulson. I am working on his dupe and trying to fill out my 5* tier to be able to one day champ everyone. I'm not really looking to 550 anyone.

Xception81

This might not be exactly what you’re looking for, but might help out…

https://forums.d3go.com/discussion/89152/ascending-characters-a-champ-reward-analysis/p1

bluewolf

@tiomono said:
I guess cp and iso would be the biggest mark on my roster. I'm in 5* land with my highest leveled being 467, and have only maxed one 4*, coulson. I am working on his dupe and trying to fill out my 5* tier to be able to one day champ everyone. I'm not really looking to 550 anyone.

My math (hopefully right) says ascending 1-3 means you get 65000 iso from 200 1* covers. Vs 20,000 from selling them.

You get no cp from 1s unless you ascend them. From a 1-3 you get 6 LTs and 41 cp from those 200 covers you use to get them to 266.

Hope that helps a little.

GrimSkald

So the big complication in the math on how to optimally ascend characters is that covers are converted into levels at a set rate but are never converted back when you ascend more than the minimum. For example, converting two 2★s to a 3★ if they're both 144 you get an extra 25 levels on the new 3★ -- all well and good for the first one, but then you level them up to 266 and you need to get a new 3★. You take two 144 2★ again and get that 191 3★ and combine it with the 266, instead of the 191 counting as 50 extra covers, it just counts as 25 (as if it was a "native" 3★,) so that 25 gets divided by 3 to make a 278 with one cover left over. That's what they mean by "tax." If you just made a 166 champ, Ascended the max and new 3★, then applied 50 covers, you'd end up with a 286 4★ with two covers left over so you're losing 12 1/2 2★ covers in this deal.
The question is are the extra rewards worth losing those covers? I'd say probably yes unless you really want to level your 2★ up to 4★ level as fast as possible. You also definitely want to get at least the LT from level 167 -- that's a no-brainer. OTOH I wouldn't always combine max-champs either -- it gets worse from there since you'd be spending the 3★ levels at 2:1 and then getting a 3:1 return on it. So you'll basically be spending (at least) 200 covers on getting that new 3★ to max champ, if you do that with the second champ you'll then get 33.3 levels out of 200 covers, which would normally give you 66.6 levels.

JoeHandle

@tiomono said:
Has anyone done the math on value per cover for the different star tiers as you ascend characters? Like how much cp, iso, hp, shards, per cover across different tiers comparing ascending vs just champ recycling. I have tried searching and I am struggling to find what I am looking for.

Hello tiomono, I have exactly this. I built a sprawling model to study various approaches to ascension. It calculates the total of every resource from any level to any other level, and the avg of each per cover invested.

But, at 12,000 cells, it's not easily share ... uh ... share ... uh ... bull.

... and there are multiple versions, for considering the various varieties of feeders and nonfeeders, particularly in the 4* tier.

What I was looking for was an optimal, or near-enough-to-optimal solution that was an offramp short of max-max'ing everything, hopefully as close to min-max'ing as possible. My thought was, that, if you took an approach that had the highest percentage of the coverflow "landing" in the 5* tier, that is, the most covers possible were chasing the highest value rewards ... which sounds a lot like min-max'ing ... that would be the approach that resulted in the highest avg per cover of each resource, or near enough ... highest avg per cover of the resources that mattered most.

And ... the answer WAS ... nope. Max-max everything, from lvl 1 all the way to 550, for the highest avg return per cover.

To me that was a very boring result ... and counter-intuitive! After all, consider a 1*, which takes the most covers to go all the way and so many of them get pumped into low tiers if max-max'ing ... only 300 of 1908 covers are spent at the 5* tier, only 15.7% of all the covers chasing those sweet 5* rewards. If min-max'ing, 400 of 1508 covers go toward 5*s, 26.5% ... how does it not result in higher avgs per cover ... ?

It's all the double-dipping, all those champ level counting toward the next tier and getting paid twice for them. If that wasn't part of the scheme, then only sadistic suboptimal fools would ever max-max anything. But, it is part of the system ... and it ensures that max-max'ing IS the optimal "min-max" (harhar) solution.

_... UNLESS ... _

The above is only true if you take the totals and avgs of each resource separately and at face value. If those numbers were weighted (with weights reflecting your personal situation) and indexed, you might get a different answer.

For instance, if I become ridiculously post-Iso (again), I would give greater weight to HP, CP, and/or LTs, and greatly discount iso. That could result in a mathematical endorsement of min-max ascension.

Adric1123

I posted the following on reddit: https://www.reddit.com/r/MarvelPuzzleQuest/comments/173562p/ascended_farming_analysis/

I wrote it just after ascension rolled out, so some of it is out of date now. Specifically, everything that refers to 2 power 1*s. The ascended champ rewards also turned out to be slightly different than advertised. The basic conclusions are sound though and the overall rewards relationships are basically the same.

Adric1123

@JoeHandle said:

@tiomono said:
Has anyone done the math on value per cover for the different star tiers as you ascend characters? Like how much cp, iso, hp, shards, per cover across different tiers comparing ascending vs just champ recycling. I have tried searching and I am struggling to find what I am looking for.

Hello tiomono, I have exactly this. I built a sprawling model to study various approaches to ascension. It calculates the total of every resource from any level to any other level, and the avg of each per cover invested.

But, at 12,000 cells, it's not easily share ... uh ... share ... uh ... bull.

I would just like to say that I salute you for your efforts. I've done a little of that, mostly to refute arguments for non-max-max ascension strategies, and have a full appreciation for your efforts. I never went as far as you did, but would like to thank you for doing so. It makes me much more confident in the max-max strategy.

... and there are multiple versions, for considering the various varieties of feeders and nonfeeders, particularly in the 4* tier.

What I was looking for was an optimal, or near-enough-to-optimal solution that was an offramp short of max-max'ing everything, hopefully as close to min-max'ing as possible. My thought was, that, if you took an approach that had the highest percentage of the coverflow "landing" in the 5* tier, that is, the most covers possible were chasing the highest value rewards ... which sounds a lot like min-max'ing ... that would be the approach that resulted in the highest avg per cover of each resource, or near enough ... highest avg per cover of the resources that mattered most.

And ... the answer WAS ... nope. Max-max everything, from lvl 1 all the way to 550, for the highest avg return per cover.

To me that was a very boring result ... and counter-intuitive! After all, consider a 1*, which takes the most covers to go all the way and so many of them get pumped into low tiers if max-max'ing ... only 300 of 1908 covers are spent at the 5* tier, only 15.7% of all the covers chasing those sweet 5* rewards. If min-max'ing, 400 of 1508 covers go toward 5*s, 26.5% ... how does it not result in higher avgs per cover ... ?

It's all the double-dipping, all those champ level counting toward the next tier and getting paid twice for them. If that wasn't part of the scheme, then only sadistic suboptimal fools would ever max-max anything. But, it is part of the system ... and it ensures that max-max'ing IS the optimal "min-max" (harhar) solution.

_... UNLESS ... _

The above is only true if you take the totals and avgs of each resource separately and at face value. If those numbers were weighted (with weights reflecting your personal situation) and indexed, you might get a different answer.

For instance, if I become ridiculously post-Iso (again), I would give greater weight to HP, CP, and/or LTs, and greatly discount iso. That could result in a mathematical endorsement of min-max ascension.

I'm pretty sure max-max will still be best even if you don't care about Iso. CP and LT awards all come from champ progression, so moving through as much champ progression as possible, i.e. max-max ascension strategy, will still be the optimal solution.

That said I did do a bit of analysis recently that showed that, at the 3* level the HP rewards for maxing the duplicate are not worth the "cover tax" on a per cover basis. Basically to max the HP/cover value when ascending a 1* or 2* you should max-max at 2*, max-min at 3* and max-max at 4*. The downside is it does cost you on CP and LTs.

tiomono

I appreciate the info and replies from everybody. I think I have a more clear decision in my head now.

Daredevil217

Isn’t max/max the best option if you want the most rewards per cover but min max if you want to get to 550 fastest?

JoeHandle

@Daredevil217 said:
Isn’t max/max the best option if you want the most rewards per cover but min max if you want to get to 550 fastest?

Max+min will get you to 450 the fastest.

Max+max returns the most rewards overall.

If asking "per cover" tho that can get murky. Lots of devil's in lots of details! Generally yes max+max maxes returns but if focused on a particular reward type that may not be true and the answer can also vary based on starting tier.

Scofie

@Daredevil217 said:
Isn’t max/max the best option if you want the most rewards per cover but min max if you want to get to 550 fastest?

I don't think Min Max gets you to 550 any faster. It just gets you a 5* faster. Max-Max should give you a 475 (+100 covers /4) so you still need 300 covers to get you to 550. Either way it's 400 covers beyond your 2nd champ.

KGB

@Scofie said:

@Daredevil217 said:
Isn’t max/max the best option if you want the most rewards per cover but min max if you want to get to 550 fastest?

I don't think Min Max gets you to 550 any faster. It just gets you a 5* faster. Max-Max should give you a 475 (+100 covers /4) so you still need 300 covers to get you to 550. Either way it's 400 covers beyond your 2nd champ.

For 4 stars that's true.

They are talking about 1, 2 and 3 stars where you have to ascend multiple times and when doing that there is the cover tax starting on the 2nd ascension if you go the max/max route.

KGB

Pantera236

Always is not correct. For base level ascending yes. So ascending two Polaris max/max. Ascending two 3* Psylocke max/max. Two 2* Moonstones max/max. There's no cover tax in these situations. And actually any base 4* always max/max. But for instance once you have a 370 and a 270 Psylocke you should just ascend them min/max. In order to take that 270 Psylocke to 370 would take 300 covers and you'd get

112,500 iso
4,000 HP
250 CP
10 LL tokens

From all the 4* rewards and you'd have a level 475.

But if instead you min/maxed them you'd not get any of those 4* rewards but would have a 450 Psylocke and 300 covers now to apply at a 4 covers to 1 level ratio or 75 levels for a 5* and get a 525 Psylocke and...

112,000 iso
350 CP
4,500 HP
12 LL tokens

From all the rewards from 476 to 525

This is a copy paste for technically a different question but still applies here.

meadowsweet

@JoeHandle said:
Max+min will get you to 450 the fastest.

@Scofie said:
I don't think Min Max gets you to 550 any faster. It just gets you a 5* faster. Max-Max should give you a 475 (+100 covers /4) so you still need 300 covers to get you to 550. Either way it's 400 covers beyond your 2nd champ.

@Pantera236 said:
Always is not correct. For base level ascending yes. So ascending two Polaris max/max. Ascending two 3* Psylocke max/max. Two 2* Moonstones max/max. There's no cover tax in these situations.

Number of covers required to get characters to Levels 450 & 550, by original Tier:

(In the time it takes to get a Max-Max character Ascended to 5★ Level 450, the Max-Min equivalent 1★ & 2★ would already be at Level 550, the 3★ at Level 525, and the 4★ at Level 475.)

JoeHandle

One of the fun things about a system like ascension, and the larger topic of roster management in general, is that there is no one right, optimal approach for everyone. Everyone's "optimal" depends on their situation, resources, goals, preferences, blah blah blah. But then everyone can pretend that everone else is asserting that their take is the best take for everyone, which leads to unnecessary misunderstanding and confrontation so the internet can happen.

Usually, if you can stay disengaged and appreciate the potential perspectives of others, you can see where they are coming from and why. Why they arrived at their 'optimal' approach to whatever it is. And thereby see why they're wrong and what flavor of psychoses they are afflicted with, harhar.

But sometimes ... not.

For instance, I see this "max base tier, then max-min, then 5* " asserted as 'best' often, in many places, from many people. but ... uhhh ... it don't make no sense (to me?). What's the focus? What's the rationale? Most approaches are focused either on performance (increasing utility (read: level) of the character as fast as possible), or resource(s) (how can I max the return of resource(s) while ascending [characters] ? ).

But this advice is at odds with itself if it's focused on performance or resources ... is it something else? Or a not so great attempt to split the difference? If trying to get the best resource return from a given # of covers, well, this is not the way.

It seems to be a misfire reaction to the increasing # of covers required to advance in ascended tiers (1/ lvl, 2/lvl, 3/lvl, 4/lvl ... argh).

Without understanding what a person wants to accomplish and why they think their chosen method is a good one for achieving that accomplishing, it's impossible to tell them why they are wrong and why.

Is there a white paper on this "base-tier max, then max-min then 5* " school of thought I can read?

@Pantera236 said:
Always is not correct. For base level ascending yes. So ascending two Polaris max/max. Ascending two 3* Psylocke max/max. Two 2* Moonstones max/max. There's no cover tax in these situations. And actually any base 4* always max/max. But for instance once you have a 370 and a 270 Psylocke you should just ascend them min/max. In order to take that 270 Psylocke to 370 would take 300 covers and you'd get

112,500 iso
4,000 HP
250 CP
10 LL tokens

From all the 4* rewards and you'd have a level 475.

But if instead you min/maxed them you'd not get any of those 4* rewards but would have a 450 Psylocke and 300 covers now to apply at a 4 covers to 1 level ratio or 75 levels for a 5* and get a 525 Psylocke and...

112,000 iso
350 CP
4,500 HP
12 LL tokens

From all the rewards from 476 to 525

This is a copy paste for technically a different question but still applies here.