1. Welcome to the Toram Online Forums. Please make sure to read our rules and be friendly to all our members.
    Click here to read the rules before posting.
    Dismiss Notice

On Raising the Level Caps of Pets

Discussion in 'Pet/Taming' started by Morningwood, Sep 13, 2018.

  1. Morningwood

    Morningwood Well-Known Member

    Joined:
    Aug 16, 2017
    Messages:
    84
    Likes Received:
    72
    IGN:
    Sasafras
    So. An important question to ask is: what is the easiest, or most efficient, or fastest, way to raise the level cap of mainline pet A by fusing with sacrificial pet B to obtain pet A* whose level cap has been raised by D levels?

    The benefit is D, at some cost. The goal is to get the highest D at the least cost (or tedious grinding time, or whatever makes the process of leveling A and B painful).
    It is obvious that there is an approximate "sweet spot" for D, because the extremes give silly situations that normal people would never consider. Raising the cap of a lv62 pet A using a lv66 pet B to get a lv64 pet A* (I will be ignoring the extra 1 level from fusion formula in all that follows) is pain with little gain; D is much too small. Near the other extreme, lv001 A with lv120 B is inefficient. Compare the pain of raising B those last 20 or so levels with the ease of raising A the first 20 levels.

    In order to maximize D/cost, we need some way to express the cost. Expressing cost in terms of grinding time would be nice, but I decided that the least subjective measure would simply be in terms of xp required to get to a certain level. So cost of pet A is Ka, Kb for pet B, and we want the ratio D/(Ka+Kb) to be high; D appears in the cost as half the level gap between A and B. Maximizing this ratio will point us to a sensible value of D, and the gap in levels between A and B. In other words, given the level (max) of A, when is a good time to quit leveling B?

    I should say (before I forget) that there is a certain overhead cost to leveling a pet, such as the time needed to train it. In addition, it takes 2 hours or less to level a pet to 60, so I regard the first 60 levels as pretty much the cost of doing business, and don't worry too much about trying to squeeze extra efficiency for pets under 60.

    I used experience data from a Nightmare Crystal chart, which gave xp for 10 level increments, up to level 90. In the vicinity of level 60, the xp required approximately triples per 10 levels. The xp multiple required trends down to about x2 per 10 levels in the vicinity of level 90. I used this trend in determining the costs Ka and Kb.

    With all of the ifs, ands, and buts, I would recommend that the results I am about to state be used more for insight, rather than a recipe.

    In the target range A*=60, D 9 to 13 are most acceptable (a gap of 20-25 I would say). I bumped the gap up a little; it just seemed prudent to stop grinding late rather than too early. A concrete example would be 50+70->60, a gap of 20, but I would probably do 50+76->63 (a gap of 26), for a mainline pet maxed at 50.

    In the target range A*=90, the benefit/cost curve is much flatter, but the costs involved are much greater, so deviation from the best gap value would be more costly, but we are working with a model (and assumptions and nebulous costs) and the "best" gap value is, well, not exactly carved in stone. D 15 to 20 looks acceptable, or a gap of 30-40. I am inclined to lean towards a gap of 40.

    I went through an exercise, working backwards, to see what a reasonable efficient level cap raising scheme might look like. I chose the last sacrificial pet B to have a level of 130 (obtainable currently) and a gap of 40. I rounded everything to multiples of 5. It resulted in a six step fusion, like so:

    A* <- B + A GAP
    110 <- 130 + 90 40
    90 <- 110 + 70 40
    70 <- 85 + 55 30
    55 <- 65 + 45 20
    45 <- 60 + 30
    30 <- 60 + 1


    No, I am not suggesting that you start with a lv 1 pet :)

    Result is lv110 pet, with some semblance of efficiency. The pet can be leveled past 110 obviously, but until it reaches 110, this might be a good sketch of a strategy. I also considered how to shorten it by a step, and ended up with (using less efficient gaps):

    A* <- B + A GAP
    110 <- 130 + 90 40
    90 <- 110 + 70 40
    70 <- 90 + 50 40
    50 <- 70 + 30 40
    30 <- 60 + 1

    A normal person, not working backwards, would write 1+60>30 +70>50 +90>70 +110>90 +130>110 + pain unending > super pet, lol.

    Anyways, take it all with a grain of salt. Anything that happens below lv60 or 70 is probably going to depend on your own situation and habits. Anything that happens above lv110 is going to remain somewhat mysterious for a while, I think.

    EDIT: I have added some stuff, it is in message#4. AND, thanks for the kind words :)
     
    Last edited: Sep 15, 2018
    • Like Like x 2
    • Creative Creative x 2
    • Winner Winner x 1
  2. Red_Fox

    Red_Fox Master of the Ninja Kittens Super Likable Elite Member Epic Member

    Joined:
    Dec 1, 2015
    Messages:
    5,789
    Likes Received:
    3,185
    IGN:
    Red_Fox
    awesome =(^.^)=
     
  3. Xiavu

    Xiavu Lurking Overlord of LOVE[L.O.L.] Elite Member Epic Member

    Joined:
    Jul 28, 2016
    Messages:
    3,695
    Likes Received:
    1,103
    IGN:
    Xiavu
    Nice thing to consider when you actually start powering up your pets. It makes every fusion an important step to take and not just stacking pets on top of the other. It triggered my pet OCD more, nicely done!
     
  4. Morningwood

    Morningwood Well-Known Member

    Joined:
    Aug 16, 2017
    Messages:
    84
    Likes Received:
    72
    IGN:
    Sasafras
    In the opening post, I was looking at the problem mostly from the perspective of "How would one raise the lv cap of a pet to a certain level". That's nice and all from a theoretical perspective, but has some drawbacks. One is that it is awkward when confronted with the question "OK I have my pet maxed at lv A, now what?". The gap values that are probably the most useful would be those that apply to the pet-in-hand. Secondly, the gap value estimates from the target pet A* perspective are really best suited to answering questions such as "what are the best two levels of A and B that will give me a desired A*?", which isn't all that practical. And in fact the gap value estimates that I took from that analysis were too broad, because for one thing:

    To top it all off, the xp data that I was using was not of the best (or even acceptable) quality. So I redid (redone?) the analysis using better xp data, and in terms of "pet A would get the cheapest gain of levels by fusing with pet B when the lv of pet B is about A+gap". I'll just post the charts (they have some fusion examples written on them, to show how they might be used). It is interesting to see the actual shapes of the curves near the maxima, so that one can decided for themselves how far to "color outside the lines."

    I took the xp data from this thread: Exp Levelling Chart (NC) Lvl 150 Cap

    I have to confess that I did smooth out a glitch in the neighborhood of level 80; I don't know if it is a typo or not. So my accumulated xp values will differ a little from Nincy's. My second confession is that I fit (good enough for government work) an analytical function to the xp data, and took the bold (and possibly unwarranted) step of using that function for xp above 145. If the extrapolated data holds to 165, then the locations of the maxima in the graphs should be unaffected.

    Here is a comparison (somewhat contrived) of two routes that take a mainline pet from 60 to 114. The first uses three lv 120 pets, and the second uses four fodder pets B that were suggested by the curves below. The difference in cost (measured in xp) isn't terribly huge, but it ain't nuthin either.

    [​IMG]

    The example fusions in the charts used gap values (colored) taken near the maxima. There is nothing magical about them; they just mark a ballpark value whose neighborhood is all good (assuming I didn't mess up too badly).
    [​IMG]


    [​IMG]
    To be perfectly honest, I don't want to have to decipher these charts every time I want to plan out a pedigree. Because, practically, I find I have two constraints: the lv cap A of the pet that I have, and the lv cap of the best matching last fodder pet that I have or can reasonably get. That usually means squeezing out a fusion step (larger gaps at lower levels), or possibly adding one. I use pencil and paper. Well, eraser and paper mostly. I need something really simple, that doesn't make my eyes bleed too much. So I summarized the maxima locations and pet-pet relationships:
    [​IMG]
    Note: I included the data at lower levels, but below B=70 or so (my current opinion, lol) I don't think it makes sense to take the baby steps that the chart would suggest. One can grind a fodder pet to 70 fairly quickly (I use Ooze nightmare from 30+). Realistically, deviations from the ideal in the B=70 to 100 range shouldn't be too costly either.
    So this last chart is what I have been attacking with my one-hand pencil for (cough cough) fudging up a level cap-raising chain.
     
    Last edited: Sep 16, 2018
    • Winner Winner x 2
    • Like Like x 1
  5. Morningwood

    Morningwood Well-Known Member

    Joined:
    Aug 16, 2017
    Messages:
    84
    Likes Received:
    72
    IGN:
    Sasafras
    I have a neutral Ruin Rat named "Macaroni" that I've been messing around with, on-and-off, for months. He started out with a level cap of 15. I fused him at lv15 with a lv111 fodder and leveled him to his new max of 64, then with a lv96 pet that had Raging Blow 10. At that point his categories were (Bow/All rounder/Intelligent). I decided to shoot for (BG/Accurate/Sly), so I began a taming spree: bags of pets, fused untrained. I fused the (x/x/x) pets, used the pets with one desirable trait to gamble for two desirables, and gambled those to get (BG/Accurate/Sly) lv 1 pets. I left pets with two desirable traits in their cages to use for level cap raising. I was able to correct his traits with two fusions using (BG/Accurate/Sly), but he ended up with a level cap of 20 (but a lot of power).

    "What the heck do I do now?" That's how the subject of this thread was born.

    I fused Macaroni with a lv68 double (preserving his traits) and raising his cap to 45.
    I have other doubles with which to raise his level; the best atm will max at 130.
    I doped out a couple of fusion chains as an exercise, using the constraints A=45, last B=130. It turns out there is a fairly simple way to do this using the worksheet from my previous post. Starting from the point B=130 on the fuchsia mate line, I walked down, left, down etc, along the orange arrows, and collected pairs of (A&B) which added up to an even number. The last step (which is really the first fusion) gets a left-pointing arrow all the way to A=45 (the fudge step). Result would be a lv114 pet, which is a constraint of fusing the lv 130 with the lv 96 pet that the green line suggests. Underneath these worksheet lines are broad maxima, so we can deviate to our hearts' content, within reason.

    [​IMG]
    I considered aiming for a lv 116 pet, which means the 130 pet would fuse to mainline pet lv 100. The procedure is pretty much the same, and follows the teal arrows.

    What I will actually do is... I have no idea. I'm leveling the last few fodder pets, and will probably do the next couple of fusions with fodder according to the teal scheme, or maybe a little higher than the teal fodder.
     
    • Winner Winner x 1
  6. Xiavu

    Xiavu Lurking Overlord of LOVE[L.O.L.] Elite Member Epic Member

    Joined:
    Jul 28, 2016
    Messages:
    3,695
    Likes Received:
    1,103
    IGN:
    Xiavu
    This is why i developed my ocd in choosing fodder pets! The very least 2 target traits must match that is the general rule since you can only lock one, it eliminate the randomness while pushing the cap level upwards.
     
    • Like Like x 1
  7. Cruithne

    Cruithne Well-Known Member

    Joined:
    Jan 23, 2016
    Messages:
    141
    Likes Received:
    257
    IGN:
    Cruithne
    Hey, I find this is an interesting problem ! After some thinking, I made a calculator to try all possibilities (haven't had time to write down full explanation on how I do that, but I will soon). So here's the calculator: http://coryn.club/pet_fusion_optimizer.php ;)
     
    • Winner Winner x 3

Share This Page