Forums → Games → Gemcraft CS: semi-optimal combining method for mana gems and amps

Just read the "end game guide" created by psorek (GCS important links, sticky). There are graphics, too.

@cronos51 Which formula should be used for mana gems. When i use the one from the "end-game guide" i get the gem on right from the image bellow which looks a lot weaker for mana leeching that the gem on the left which i get when i use formula from 1st page of this topic. It is mainly because "end-game-guide" formula involves combining black gems into orange ones which makes black ones main gems in formulas. So which one is right?

The second one is right. It's all formulas by psorek, so the actual one is in the guide that he wrote. While the mana is lower, the bloodbound is better. If you would compare the same gems at higher G's and with many Hits (AND with orange amplifiers), the second (right) one will be better.

Possibly the first has less black in it as it was created at the time of 1.0.15 - when black could still be amplified. While now only orange amplifiers are possible.

Right i forgot about amplifiers. Without them the gem on the left is much better. Gonna test it now . So that means that formulas from high-end guide are up to date for current version?
Ye i just checked and ampfs makes all the difference. ._. It also means i was manafarming with wrong formulas all the time ._.5% diff on lvl 20 alone and i bet it would go even higher on higher
grades ._.

I'm also trying to break endurance with currently GC:CS 1.19. The only "abuse" left to do is supergemming and permafreezing to generate enough Mana and poolbound hits to last longer. ATM the real trouble are the pesky "Flying / Spire" spawns that have 5-10 orders of magnitude higher hit-points than monsters (1E60 @ endurance wave ~700, long way to go till wave 999).

I optimized the 64-combine method, to avoid mistakes. Could also be used as a template for macroing-tools like "autohotkey", etc:

legend:
1 -- starting gem (e.g. Grade 1 O, or 32-spec mana)
2 - 32 = number of fused starting gems that make up this particular one
slots A...J -- order in which gems appeare when hitting duplicate hotkey "d"
"d" -- duplicate Gem
-> -- combine Gems with hotkey "g"
"u" -- upgrade Gem



1) start with Gem 1 in lower right slot (A)
2) and 3)
Mouse stays in A, hit d - d - d - d - u - d - d - d - d - u - d - u
4) start with 1 in B and combine with 2 in F for a 3
5) hit d to crate a copy of 3, then repeat dragging and combining
C -> F - d; D -> F - d; E -> F; I -> F
6) triplicate 8-value grade 3 with d - d
7) combine ingredients, order is very important here:
J -> I - d; 12

H -> E; 12
G -> E; 15
B -> E; 19
C -> E; 24
D -> E; 32

J -> A; 20
8) I -> E; 44
E -> A; 64

The numbers also correspond to the graphical combination scheme published here. Only minute mouse movements are necessary and 32 presses of the hotkeys "d,u,g". In all quite manageable with a little practice.

Hope the 64-combine method does not feel too cumbersome to you any more. But for the 1024-combine method, anything but a keyboard-macro tool still seems suicidal to me ;-)

@themait: Great Work!
... but there seems to be a minute mistake. The final combining scheme is missing a line and should read:

H -> E; 10 (not 12!)
G -> E; 12
B -> E; 15
C -> E; 19
D -> E; 24
F -> E; 32 (!!!!!)

J -> A; 20
etc.

That's actually a lot less cumbersome than the way I'd been doing it.

I have a schema for a lvl 10 orange/black which is better... no supergem, just normal same-level combinations. I describe it:

LVL 1.
1=Black
2=Orange
LVL 2. 1-1 1-2 2-2
LVL 3: 1-1 1-2 2-3 3-3
LVL 4: 1-2 2-2 2-3 3-3 3-4 4-4
LVL 5: 1-2 1-3 1-4 3-5 4-5 5-6 6-6
LVL 6: 1-3 2-4 3-4 4-5 4-6 6-7 7-7
LVL 7: 1-4 2-5 3-5 5-6 6-7
LVL 8: 1-3 2-4 4-5
LVL 9: 1-2 2-3
LVL 10: 1-2

Just make the combinations at each lvl, as the number for each new lvl correspond at previous.

You will have 3 different gems lvl 2, which you will combine to get 4 gems lvl 3, and so on.

http://i.imgur.com/puFzbxU.png

I'm not entirely sure how you would classify that as "better".

If you follow psorek's manafarm gem recipe, you wind up with something with some red in it (necessary for manafarming, to leech mana from all monsters that travel across the trap). Just doing a straight U upgrade to the same cost (but Grade 9 as opposed to Grade 10 of your gem) shows an apparently reduced black and orange component of the gem, yes, but more than made up for by hitting multiple monsters per shot.

If you look back and realize that it was the Orange (or Yellow, in the killgem) that was replaced with the single G1 Red, then you would know that making psorek's gems into dual color gems without the red is a simple matter. And doing the same thing (making that dual color gem, then using U to upgrade to Grade 9, the same cost as your Grade 10) finds a gem that would eventually wind up better than yours :
148.97 mana/hit, x2.00 bloodbound, +1.003/hit level, 0 hits

Of course, taking that modified psorek recipe, and doing the 16 gem upgrade also gets you to the same cost (but Grade 8 instead of Grade 9 or Grade 10) :
154.77 mana/hit, x2.05 bloodbound, +1.049/hit level, 0 hits
Of course, with this one, the idea is that you would continue to do the 16 gem upgrade rather than just using U.

The important thing here, though, is that we are looking at these gems for the long term. The more you supergem (and weaving, aka spec'ing, whichever term you want to use, which it looks like forms the basis for your gem creation, is a form of supergemming; it just does not decrease the gem's Grade any) the stronger the gem becomes compared to the straight up U. Concentrating on Orange provides an apparently great mana leeching gem. Concentrating on the Black provides a tremendous boost to whatever Orange (or Yellow, or whatever), as well as the slight bit of Red you need, making them that much more effective, especially with a few hit levels (say, 30 or 40 hit levels). The Black in the gem just becomes even more effective when you boost the Orange (or Yellow) with amplifiers because that apparently is done before the Bloodbound takes effect.

I applaud your efforts. I wouldn't be able to come up with something like that myself. I just use and try to follow the reasoning of the work others have done. Psorek's gems were designed for the long term, not for just creating the strongest initial gem and then accepting a lower growth rate of that power. Using White instead of Black would certainly have accomplished that goal!

Figure out where to stick some red in that gem of yours. I don't believe that it will remain stronger than psorek's gems when all is said and done (higher hit level, in a trap with pure orange amps) but despite what I'd said above, I don't believe that your gem should just be straight up ignored, either. It bears testing, just as the use of White instead of Black for the starting gems bore testing.

I meant to ask, though - what upgrade plan do you use with that starter gem?

Yep, its fair.

In fact, i've made up a program inspired in your quest for the best gem. I figured out that there is a real indicator we can take to mesure the efficience of a given gem. It would be, for orange and black:

Efficience(orange)= (Component / (1.38^(grade-1)))/0.75
Efficience(black)= (Component / (1.09^(grade-1)))/0.75

We have 0.75 for the penalty of having a dual gem.
Component is the value of a lvl 1 gem, multiplied by 2, because we have a differente multiplier applying at lvl 7 (if we suppose we have all skills).

The efficience of just combining and hitting U, is of 63.7% for both components. And for classic gem weaving it comes to 75.59% at lvl 22 and later.

Its noticeable that once we got a given efficience, the normal U hitting will keep this value, then it is a one time work.

I think it is important to check the results of psorek, which are very interesting, and look for this caracteristic. I havent tried yet a program with gem combining with different level gems. But maybe i'll work in that the following days.

The gem i showed it has:

73.26% black
93.86% orange

I explain how to get this result with the resulting gem:
In black i have +0.818 which i ll call B.
In orange i have 172.69 mana per hit.
Then CB(comp. B)=(B * 1.45887/ 0.75) / (1.09^(grade-1)) = 73.26%
Then CO(comp. O)=(O / 7.4256 / 0.75/(multiplier black)) / (1.38^(grade-1)) = 93.86%

I've tried only same lvl combining, as i think it is the best way to increase the efficience, but i have to confirm it. For other gems, grade is equal to the equivalent gem in price.

To finally answer your question, my plan is to get a decent gem in both orange and black, and then hitting U as it will no longer matter. For the red or blue component, we can always make a initial gem with 4 components and then add the created gem upgraded at each level. This is both simple and efficient to me.

Well, thats it!

For the white or black discussion, when I prepare for a long endurance run, I make the first gems in white and I get 12e6 at first mana available (for I use all traits lvl 7) and I immediatly change to black, as I think a lvl 18 gem is strong enough to start leveling up the hit counter, and for the second time the mana is available I get 630e9 of mana, which is pretty nice .

Two mana farm traps and one kill trap in a one entrance map to prevent lag.
I have 1500-3000 monsters at screen at any given time.

I've yet to see a gem that did NOT benefit from the 16, or better yet 64, gem upgrades that psorek came up with.

Just using U is horribly inefficient by comparison.

The 16 gem upgrade is very easy to do. The 64 gem upgrade is a little more complicated, but has MUCH better results - two iterations of it get you the same cost upgrade as 3 iterations of the 16 gem plan, yet those two iterations make a stronger gem. The difference only grows wider as the upgrades get applied.

I forget the value that psorek gave, but according to him, he calculated that the 16 gem plan is some 20-30 percent stronger than just U by grade 75 or so (that's the cost of a grade 75 gem, not an actual displayed grade 75). The 64 gem upgrade was comsiderably better, at some 50 or 60 percent better, I think. I forget, its been a while, but it was somewhere around there.

I had asked because I figured that maybe you had come up with something better, that's all. Psorek has a 1024 gem upgrade plan (that he made as a sort of dig at me, since I was the one who really started using his 64 gem plan and convinced others to use it, getting the popularity ball on that plan rolling) but it is really time-consuming to do, and not worth it overall.

This gem isn't better thant a Psorek 32, upgraded with 16 once.
The mangain is a little less, but the bloodboundmultiplier - and most important the step per hitlevel - is way better.

Psoreks caclulations included the fact that one can only amplify with pure O.

Just put these two gems in a trap, amplified by 6 x G10 pure O. Already there the Psorek-Version will turn out better - and that is with zero hits while having better bloodbound-stats.

Can you show the resultats of the test you talk about?? It would be interesting to see the diference at bloodbound to see if the hit levels needed is little or big to get an noticeable gain.