Which 5* to choose with my roster?

mpiter
mpiter Posts: 77 Match Maker
Hello everyone,

I read many things about entering the 5* world but I still hesitate between three strategies to champion my first 5* characters.

I mainly play in PVE for progression in SCL 8 though I can play in SCL 9.  I sometimes play for placement and I can reach the top 10 of the event, but most of the times I just play what is requested to earn the last reward of the event to just get the CP.  I also play DDQ.  I rarely play PVP.

My current 5* roster:
  • Black Bolt (433, numbers of covers in the 3 powers)- Gambit (422)
  • The Hulk BB (412)
  • Jean Grey (312)
  • Daredevil (042)
  • Wasp (212)
  • Black Widow (202)
  • Spider Man PP (220)
  • Black Panther (121)
  • Captain America PA (121)
  • Loki (211)
  • Hawkey, Doctor Strange, Ghost Rider, Silver Surfer, Wolverine, Captain Marvel, Iron Man Mx, and Star-Lord: 3 covers.
  • Thanos, Jessica Jones, Thor, Archangel, Green Goblin, Kingpin, Cable, Captain America IW: 2 covers
  • The rest with one cover including Kitty Pride (010) and Okoye (100), except Beta Ray Bill, Carnage, and Storm who have no covers.

The three strategies I can think from what I read from the forums are:
  1. The 3 champions from the Recent Legends pack when at least two good ones are available;
  2. Black Bolt (433, numbers of covers in the 3 powers that I have) maybe with Gambit5 (422);
  3. Kitty Pride (010) or Okoye (100).

I wrote a Python program to estimate by means of simulations the number of pulls needed to fully cover a chosen 5* character from the Recent or Classic Legends vault.

This paragraph details the simulations.  You can skip to the next paragraph if your are not interested in the details.  The program requests the number of simulations to estimate the statistic results, the number of covers already owned in each power, the number of characters in the pool (3 for Recent Legends and 36 for Classic Legends at the time of writing this message), and the number of pulls I want to do without giving the shards to the character target.  The program assumes that all shards provided by the pulls are given to the character target except the first pulls indicated by the last program parameters.  The piece of software assumes that all shards given to the character will be saved until it is be possible to fully cover the character with the shards.  It contrasts with the idea that we could immediately give 500 hundreds shards as soon as they are available for a cover to improve quickly the character but that increases the likelihood to have more than 5 covers in one power before the character is fully covered.  It also assumes that D3 respects the claimed probabilities: 1 chance out of 7 to pull a 5* cover, and all covers are evenly distributed.

Nothing else than using the shards and pulling covers from the Recent or Classic Legends vault is taken into account.  For examples, we could exchange 250 CP against a character cover if it becomes available in "Heroes to hire" (or a similar name), or we could earn a cover in a special event, or we could produce covers feeding the right 4* character...  So one should be able to champion a character in a little bit less of time than what the Python program estimates.

The statistics that will follow were each estimaded with 10000 trials assuming 36 heroes in the Classic Legends pack, although more heroes will slowly join the pack over time.

I will give my current thoughts about the three strategies in three other posts instead of a long one.  This way, anyone will be able to easily quote one strategy independently of the others to provide feed the discussion about strategies.

I thank you in advance for any piece of advise.
«1

Comments

  • mpiter
    mpiter Posts: 77 Match Maker

    1. Recent Legends pack


    To fully cover a character target with only 2 competitors and without a cover at the beginning, we would need around:

        273 pulls on average with 1.2 saved covers on average (due to covering a power more than 5 times).
        The 3/4 one-tail confidence intervall had its upper limit equal to 332.
        The 95-% one-tail confidence intervall had its upper limit equal to 391.
        The 99-% one-tail confidence intervall had its upper limit equal to 469.

    With around 300-350 pulls, the three characters should be at least usable and one or two might be fully covered or close to be.  A combination of Legend Tokens and CP can be used to reach quicker the 300-350 pulls.

    Knowing that I earn around 80 CP/week and 2 Legendary Tokens/week with the time I currently invest in the game (mostly PVE and DDQ), it would take me around 15 months to gather the resources for 338 pulls.

    The drawback is to be forced to save tokens during 15 months without feeling significant progresses in my 4* and 5* rosters.  That might be a bit boring to mainly stick to the 4* land for such a long time.
  • mpiter
    mpiter Posts: 77 Match Maker

    2. Black Bolt + Gambit5


    BB is currently 433 (powers) at level 405. I use him very often with champed American Chavez and either the requested third character for essential nodes, or champed The Hood for other nodes to gather tiles faster for AC's passive and to use faster BB's green with 5 tiles in each color to end the fight.  This team rolls easily on the harder nodes of PVE SCL8.

    Gambit5 is only 422 (powers) at initial level 255.  I have never used him.

    To fully cover a character:

    BB:

        325 pulls on average with 0.1 saved covers on average.
           
        The 3/4 one-tail confidence intervall had its upper limit equal to 347.
        The 95 and 99 % one-tail confidence intervalls had their upper limit equal to 500.

    Hence, gathering enough CP to champion BB would likely require between 300 and 350 pulls, maybe a bit more if I am unlucky with the RNG.

    Gambit5 is often considered as a good BB partner.  Assuming that 325 pulls would be done first to champion BB with all shards given to BB during that time, then the number of SUPPLEMENTARY pulls needed to champion Gambit5 (he would earned a few covers while championing BB) would be:

        424 pulls on average with 0.3 saved covers on average.

        The 3/4 one-tail confidence intervall had its upper limit equal to 500.
        The 95-% one-tail confidence intervall had its upper limit equal to 667.
        The 99-% one-tail confidence intervall had its upper limit equal to 806.

    So I should expect to need around 425 pulls on top of the first 300-350 to champion both guys.  That amounts to around 750 pulls to champion both if I am not too unlucky with the RNG.

    It would be faster and funnier to keep BB as he is now and to concentrate first on Gambit5.  This way, Gambit would require around:

        536 pulls on average with 0.4 saved covers on average.

        The 3/4 one-tail confident intervall had its upper limit equal to 667.
        The 95-% one-tail confident intervall had its upper limit equal to 766.
        The 99-% one-tail confident intervall had its upper limit equal to 834.

    So I should need around 536 pulls to champion Gambit, and BB should need around 100 more pulls (estimated by my program) to be fully covered for a total of 600 to 650 pulls to cover both characters.  That would be faster than championing BB first.  Furthermore, I would have a pair of usable 5* sooner than championing BB first because BB is already usable now.

    With the 80 CP I earn every week => 4 pulls/week in the Classic Legends vault => 200 pulls/year, I should cover the 600 to 650 pulls in a little bit more than 3 years.  This might mean between 2.5 and 3 years using other means to gain 5* covers.

    One advantage of covering them is that there would be no reasons to save CP as it is requested with the Recent Legends vault.  As a consequence I would always see progresses in my 5* roster as well as in my 4*.

    I could also either save the Legendary tokens to gather 100 tokens/year, hence 300 ones in about 3 years to use the first strategy in 3 years on top of BB and Gambit, or immediately use the tokens to earn as fast as possible the new Legendary characters to access the 5* essential node every time in PVE.

  • mpiter
    mpiter Posts: 77 Match Maker

    3. Kitty Pride or Okoye


    With currently only one cover, I should expect:

        1225 pulls on average with 0.0 saved covers on average.

        The 3/4 one-tail confidence interval had its upper limit equal to 1334.
        The 95-% one-tail confidence interval had its upper limit equal to 1589.
        The 99-% one-tail confidence interval had its upper limit equal to 1670.

    Even putting all shards in the target, BB would be fully covered after 700 pulls (3.5 years) and Gambit would also be covered soon after the target.  DD, Jean Grey, and the Hulk BB would be usable at that time too.

    Reaching my goal would request between 6 and 7 years.  This could be reduced to 5 to 6 years using other means to earn covers such as leveling the right 4* character, e.g., Emma Frost for Kitty.  More 5* character dilution over time would not deeply change the statistics at the current character creation rate by D3.

    On the dark side, five years seems a bit long.  On the bright side, I read that Kitty with only two yellow can do wonderful things with Grocket4.  So I could spend 720 CP (9 weeks of play) for a second cover as soon as I have my first yellow (Emma Frost is only 8 levels from the first Kitty yellow).  That means starting using her in a close future while seeing my roster constantly improving over time because there would be no need to save CP except for the second Kitty yellow.

  • mpiter
    mpiter Posts: 77 Match Maker

    4. Summary

    1. Recent Legends: 15 months with almost no improvements of my 4* and 5* rosters but about 3 5* characters to be champions or close to be.  No guarantee that there will be great 5* characters in the vault in 15 months.
    2. BB alone: between 1.5 and 2 years to champion him.  The 4* and the 5* rosters will evolve over time, which is pleasant.
    3. Gambit + BB: betweey 2.5 and 3 years to champion both with a 5* roster evolving over time as well as the 4*.
    4. Kitty or Okoye: about 5 years to champion the target but Kitty usable in a less than a year if I choose her.  BB would be champion in 3 years and Gambit a little bit after the target.  DD, Jean Grey, and the Hulk BB would also be usable.  My 4* and 5* rosters will constantly evolve over time

    What do you think would be the best way?

    Thanks for any piece of advise.
  • Rod5
    Rod5 Posts: 584 Critical Contributor
    My advice would be to try not to over-think it....

    Seriously though, Bolt is kind of semi-useful but ultimately nowhere near meta. For me your best bet is to keep on hoarding and wait till you get a store you like the look of. Gambit is niche at best, really not a useful character any more. Thanos is king for PvE, especially in CL8 or lower. Thor is amazing, Okoye/Kitty both top-drawer too. 

    My main advice would be to maximise CP/LT generation, and that means holding your nose and playing PvP. Plenty of tokens, covers and CP available for a short burst to 575/800/900 depending on your appetite. If you have surplus HP drop a shield if it’s worth your while. Champ rewards mount up.
  • mpiter
    mpiter Posts: 77 Match Maker
    Thank you Rod5.  I will not focus anymore on Bolt and Gambit.  I will hoard LT and CP.

    I had 450 CP last week for the last store including Kitty, Gambit, and I think Cable was the third.  I took my chance with the 22 pulls I could do.  The probability to receive at least one Kitty cover was around 2/3, but a dark fate put me in the other third.  I only got two Gambit covers and 20 4* ones.  I will wait for the next interesting store.

    I have two questions:
    1. If I receive a Kitty yellow cover, would it be wise to spend 720 CP on a second yellow cover because some people claim that she can be useful with only two yellow covers?
    2. I often read that PVE is an easier means to gather resources than PVP.  You advise me to start playing PVP speaking of a "short burst".  Does it mean that PVP is a good way to gather very quickly resources, and then we can leave the event?
    I will definitely read some posts about PVP strategies and wait for a store with Kitty, Okoye, Thor, or Thanos.  I thank you again for your insight.
  • DAZ0273
    DAZ0273 Posts: 9,569 Chairperson of the Boards
    It is not too great a strain to gain 10 CP per PvP event. You can achieve this with an hour's play per event and as your experience grows of what times are good to play, even less time. As there are somewhere of the region of 10 PvP per season that is 100 CP just for getting to 575. 
  • mpiter
    mpiter Posts: 77 Match Maker
    OK, I will try this.  Thank you.

    I can access SCL 7, 8, and 9.  If I am just interested in the 10 CP at the beginning to understand how to play in PVP, is there a reason to play above 7?
  • helix72
    helix72 Posts: 991 Critical Contributor
    The extra 4* covers you get for hitting 900 points in PvP are nothing to sneeze at. They either help round out your 4* roster which is hard to do in this age of dilution or give you champ level rewards which get more valuable as the characters get higher and higher. Plus they provide some "cover" as you raise the level of your 5*s--a level 310 champed 4* is often stronger than a level 310 partially covered 5*, but MMR looks at them both the same way.
  • DAZ0273
    DAZ0273 Posts: 9,569 Chairperson of the Boards
    mpiter said:
    OK, I will try this.  Thank you.

    I can access SCL 7, 8, and 9.  If I am just interested in the 10 CP at the beginning to understand how to play in PVP, is there a reason to play above 7?
    Actually it sort of works in reverse to PvE. There is no difficulty difference between playing 7 or 9 as this doesn't affect who you are matched against, only who you compete for placement with. So it actually makes sense to always play the highest that you can. You will not be getting top placement with a score of 575 but find a slacker brackets and you can place quite high with minimal effort. Hitting 900 regularly is also a matter of practice but you may get frustrated at first getting there.
  • ThaRoadWarrior
    ThaRoadWarrior Posts: 9,125 Chairperson of the Boards
    edited January 2020
    Once upon a time, Gam-Bolt was a serious team, but I think those days are pretty long gone. I'm confident there are WAY worse duos you could transition with, but the gap between the great 5*s and the good 5* is a wide, wide gulf. 

    The good news is that you can shard your way to more Bolts via Squirrel-Girl -> Medusa -> Bolt, and then you can use his yellow to raise the effective level of any other character you want to pair him with. Now that you can Gambit ->Rogue->Gambit some shards there, you could chase the other side of that pair, but it's been my experience that the force-multiplying 5*s like Okoye, Kitty, and Gladiator Thor are so much better than most of the other 5s that you're almost better off punting on someone who is close and just chasing one of them directly.

    I did a little math on timelines for 5* covers using shards, and playing SCL9 to progression in PVE, PVP to at least 575/900, and with the bulk of the 4* tier champed, I have an average legendary pull rate of ~2.6 pulls per day. this means I'm earning a 4*/week via shards pulling as I go, and a 5* cover's worth of shards takes around 63 days to acquire. that time shortens up depending on whether or not I'm "focusing fire" on that 5* character's feeder tree and where the 3/4*s are in the rewards payout. So if you can figure out your pulls/day, you can figure out your shards/day, and that will tell you what fraction of your lifespan you'll be spending before you get the character you're after. The good news about getting shards from feeders though is that you could for instance choose to stack your first 5 covers on Gladiathor on his green, and he'll do what he needs to do (with Okoye) just as well as you probably want him to. Or build your Kitty with 5 yellow then work on the rest, or Okoye can start with Wakanda Forever! and be quite a functional character even pre-champion.
  • mpiter
    mpiter Posts: 77 Match Maker
    edited January 2020
    helix72 said:
    The extra 4* covers you get for hitting 900 points in PvP are nothing to sneeze at. 
    OK, I will try the 900.  Thanks.
  • mpiter
    mpiter Posts: 77 Match Maker
    DAZ0273 said:
    Actually it sort of works in reverse to PvE. There is no difficulty difference between playing 7 or 9 as this doesn't affect who you are matched against, only who you compete for placement with. So it actually makes sense to always play the highest that you can. You will not be getting top placement with a score of 575 but find a slacker brackets and you can place quite high with minimal effort. Hitting 900 regularly is also a matter of practice but you may get frustrated at first getting there.
    Thanks, I will play at SCL9 because I will not target placement, just the progression rewards.
  • mpiter
    mpiter Posts: 77 Match Maker
    Once upon a time, Gam-Bolt was a serious team, but I think those days are pretty long gone. I'm confident there are WAY worse duos you could transition with, but the gap between the great 5*s and the good 5* is a wide, wide gulf.
    OK, you finished to convinced me to focus on Kitty, Okoye, and Thor, but mostly Kitty.  I will not bother giving shards to Squirrel Girl for Bolt.  I am right now raising Jug4 to complement Grocket with Gamora4, then I will feed Cyc3 and Emma Frost for Kitty.
    I did a little math on timelines for 5* covers using shards, and playing SCL9 to progression in PVE, PVP to at least 575/900, and with the bulk of the 4* tier champed, I have an average legendary pull rate of ~2.6 pulls per day...
    I have probably between 0.5 and 1.0 pulls per day.  It might increase following everyone's pieces of advise.  So if I focus on Kitty, Okoye, and Thor, do you think it would be better to pull Classic Legends as soon as I have 20 CP to produce shards and get as fast as possible 1 Kitty yellow cover, or should I hoard CP waiting for a 3-character store containing one of the three targets provided that a store with Kitty just ended a few days ago?

    Is there a known schedule of the next 3-character stores?

    Also, after getting my first Kitty yellow cover, I am tempted to save 720 CP to buy a second one instead of pulling 36 times loosing 108 5* shards, 540 4* shards and 35 pulls (5 5* covers on average) in the process.  Is this a good idea?

    Thanks for you insight.
  • helix72
    helix72 Posts: 991 Critical Contributor
    Hoard. Three months ago I had no fully covered 5*s, now thanks to hoarding and 2 special vaults I have Kitty at 5/5/3 and Okoye at 5/5/3. Prioritize either of those 2 over Thor-you need Okoye 12 or so levels higher than Thor anyway.
  • mpiter
    mpiter Posts: 77 Match Maker
    edited January 2020
    Yes, that what I figured out.  Thank you.  I started to hoard after reading the pieces of advise of this thread.  I also started PVP where I could easily reach 575 as it was suggested.  Big improvement.

    When the special vault with Kitty was out I could pull 23 tokens if I remember well.  I was unlucky and got no Kitty covers but two Gambit5.  I therefore raised Gambit (422) to level 375 to pair him with BB (433).  Big mistake.  I should have kept BB only with AC.  Now reaching 575 in PVP has become very difficult and takes too much time and health packs.  It is discouraging and I gave up waiting to have a better roster.

    I was more lucky with Okoye.  I burned 8400 HP for 500 shards in the Code Red vault.  It was not a problem, I had more than 25000 HP.  I had use all my CP with the Kitty vault so I could only draw 11 tokens from the new Okoye vault.  I got two Okoye covers out of 11 pulls.  I plan to spend all future CP this week in Okoye vault, then hoard again for the next special vault involving Kitty, Okoye, or Thor.

    I still have an unusable Kitty (010) and now Okoye (210) with 500 shards to spend.  I do not know how to use Okoye, so I will read about her after her vault will be over.  I will decide then how to spend the 500 shards if this will made Okoye usable with 4 covers, or maybe more if I am lucky.

    Thank again for your suggestion.
  • ThaRoadWarrior
    ThaRoadWarrior Posts: 9,125 Chairperson of the Boards
    Okoye you need to collect team up tiles and let other characters do the heavy lifting. She adds a bunch of damage per team up in the bank to all non-tile destruction damage (except for some notable exceptions I have listed in the thread about how Wakanda Forever! Works). Medusa, XFDP, and Chavez are great play up 4*s for her. Put all your shards into first black then yellow
  • mpiter
    mpiter Posts: 77 Match Maker
    Thank you.  That confirms what I read after posting my previous comment.  I took a couple of hours to read about Okoye and I reached some conclusions.  Your points confirmed things I had understood, and the thread you initiated and mentionned taught me many helpful things (https://forums.d3go.com/discussion/79221/wakanda-forever-how-does-it-even-work#latest).

    I have Okoye (210) at level 300, and she is already very fun to play with.  She cannot compete at that level with Grocket+Gam4+Jug4 or BB+Gambit5, but I play a lot with her for fun when speed is not a concern.  I have already tried her with Strange3+Thanos3 against goons, or AC+Medusa, or just Medusa+imposed character against moving-node characters.  I will also try with AC+Mockingbird and XFDP.

    When her special vault is over I will put the 500 shards I have in black and I will spend ISO to have her (310) at a higher level than 300.  Very fun.

    Thanks for your help.
  • ThaRoadWarrior
    ThaRoadWarrior Posts: 9,125 Chairperson of the Boards
    Okoye/Medusa/3* Thanos is a very fast PVE team, you can sub out medusa for Strange if you have goons on the opposing team. You’ll use more health than JuggerGOTS does, but it’s less dependent on anything staying alive on the board when tile movers are in play.
  • mpiter
    mpiter Posts: 77 Match Maker
    I have already tested that team.  As you point out, it is less dependent on Strike-tile surviving against tile movers but it depends on Teamup tiles.  With Okoye at 210 (two blacks and one yellow), it is not that fast because I need at least 6 or 9 TU to really speed up things.

    With R&G, Gam4, and Jug4 tanking 5 colors out of 7 thanks to its Advantage, and R&G making a special tile out of a Strike one (more than one turn out of 6 on average with its Advantage and Gamora's), I have the feeling that the fights are done quicker than with Okoye+Str3/Med+Than3.  At the beginning of the fight, I deal more than 8000 damage (4000 to the target and 2000 to the other two) per Jug4 alignment.  With Okoye, I often need at least 3 or 4 alignments to gather 6 TU to only start striking moderately hard.  At that moment, I usually have already dealt something close to 30000 damage spread more or less evenly on the opponent team or on a character if I want to take him out quickly with my Grocket team.  When Okoye is 310 at a higher level than 300 in a few days, teams based on her should be significantly faster.  I will check it again.  But right now, I have the feeling to have more speed with R&G+Gam4+Jug4 with good Advantages to tank more colors with Jug and make special tiles out of the Strikes one.

    But even as Okoye is right now for me, I use her a lot for fun, including sometimes Okoye+Str3/Medusa+Than3 when I am close to end my game session, and health loss does not matter.

    For long fights such as waves with high-level opponents, I prefer BB (433) + Gambit5 (422) + AC champion.  It requires a little bit of setup as with Okoye-based teams, but the setup is worth the speed gain over time for long fights, as with Okoye.

    I like your suggestions, and I will follow your road map: everything in Black, then Yellow for Okoye.  Thanks.