ForumsGamesGemcraft CS: Crusade for a Grade 100 Gem

496 153428
thunderrider
offline
thunderrider
641 posts
Peasant

Hi, I am level 3142, and I am going to be the first to get a grade 100 gem.

http://i.imgur.com/LPI5mar.jpg

How did I get specs obscenely high? Well, thanks to this dude over here...

One important thing to note is how to create dual gems. Adding two grade one's them spamming U is losing about 33% special power.

Optimal creation is a G1 of both, combine, make a G2 of each, add one pure to each dual, combine the G3s, duplicate, and repeat with pure G4s, duplicate the 6, etc.

here is the difference at g34:
Regular creation:
http://prntscr.com/3dgs2c

Special Mix:
http://prntscr.com/3dgs6k

Note that the hit count is the same, and the cost to duplicate is the same. This is a difference of 32.9%, and this difference keeps rising at higher levels. It's only a 16% difference at grade 6. It takes a decent amount of time, but it is worth it at ultimate grades (45+).


I now have huge specs. It takes me about 20-30 seconds real life time to get +1 grade. So a minute before +2(needed if want extra specs). I estimate my specs to be about 80% higher than normal, at least blood is.

PS- Level 46 amps got the spec treatment too, they are wicked.
  • 496 Replies
psorek
offline
psorek
447 posts
Jester

My g78 equivalent mana supergem. Nearly 10 times stronger than traditional one
https://dl.dropboxusercontent.com/u/64607958/g78%20mana%20supergem.png

Lotoreo
offline
Lotoreo
14 posts
Nomad

Weaving:
1.) Create G2 Dual Gem
2.) Duplicate the Dual Gem
3.) Create a G2 Pure Gem of each type
4.) Combing Each Dual Gem with a Pure Gem
5.) Combine the resulting G3 Gems into a G4
6.) Repeat Steps 2-5 for each successive even Grade

Supergemming:
1.) Create the Gem you want
2.) Duplicate it 4 times
3.) Combine the Gem with 1 Duplicate for G+1
4.) Combine the other 3 Duplicates with the Upgraded Gem for G+0


You did not mention that the pure gems used for weaving can be supergems as well. But that would make weaving a lot stronger, right?
psorek
offline
psorek
447 posts
Jester

ideally you weave 3 times and them supergem. Or supergem separately 2 colors and then weave them.

thunderrider
offline
thunderrider
641 posts
Peasant

"weaving".....blah. If you want to recognize it as a sewing method, be my guest :/

benchbreaker
offline
benchbreaker
2 posts
Shepherd

just took a few hours reading the whole thread and some of the links and played around with spread sheet. There were some solid mechanics already revealed so we could predict better ways of combining gems instead of blindly experimenting.

some very important numbers were revealed in http://www.reddit.com/r/Gemcraft/comments/2385eg/sc_about_gem_combining_mechanics_and_why_we//ppAs

Combining two gems of the same grade:
Mana gain: 0.88A + 0.5B

A is the max stat (e.g. here we are interested in mana gain) of the two gems and B is the min stat. this is a very simple equation and obviously explains why the spec method gives higher stats. for normal upgrade min and max are always the same and also equal to average, since the same gem combines with itself. The spec however method exploits the fact the upgrade equation do not use the average but an unequal weights of min and max. so by increasing the max (3/4o 1/4b) and lowering the min (1/4o 3/4b) the resulting stat is higher.

here is whats interesting, the tier 4 spec dual color is optimal, since the only other 2 ways of getting to t4 (0/4 + 4/4 and 2/4 +2/4) are not as good. but the t6 spec is already not optimal as is shown in the spreadsheet below:

grade 1 pure is assigned special = 1 for simplicity sake

red box shows process of making of t4 spec, t2 o/b dual combine with t2 pure o, t2 o/b dual combine with t2 pure b, resulting in two t3 with different stats. (understand this bit and you'll understand the whole sheet)

green number shows difference between spec t6 and optimal t6)
http://i60.tinypic.com/207w3fq.png

here are ingame screenshots of spec t6 and optimal t6:
http://i58.tinypic.com/oa0tw4.png

as you can see the optimall t6 is made by combining 2x spec t4 with 7/8o and 1/8o t4's.

i also tested some T8 gems, left is normal spec t8, middle is optimal t6 spec'ed into t8, right is (i think optimal t8) where 2x optimal t6 are combined with 15/16o t6 and 1/16o t6.

http://i60.tinypic.com/jskivp.png

However this whole thing is recursive and will get exponentially more complex. e,g, for t10 you will need to combine optimal t8 with a 1/32o t8, which is in terms of grade1 gems is 4o+252b, so you start with 4o+4b t4, then combine with pure t4, t5, t6, t7 o's in order. but hang on 4o+4b tier4 is better the spec method than normal so you will have to start with spec t4. for t12 you will have to start with spec t6 but hang on spec t6 is not optimal so you start with optimal t6. it will get impossible at higher levels unless some sort of algorithm is found (e.g. you duplicate certain optimal lower tiers and save for later) but at least its pretty easy start with a better t10 and that doesn't take so long

also i only played around briefly some could look deeper into it, there are formulas for super gems as well, e.g. for mana gain its

Combining grade N gem and grade N - 1 gem
Mana gain: 0.89A + 0.44B

maybe some1 could do some maths work out the some tipping point between the different methods of combing gems

also this whole post applies for 50/50 gems only, obviously with the arbitrary numbers in the game mechanics 50/50 isnt gonna be the optimal ratio (although its easiest to upgrade, which is why spec is very good for casual/hardcore, but for real hardcore it isnt enough :P)

wow that was a big wall lol

psorek
offline
psorek
447 posts
Jester

Well, my experimental method will probably give us the same results, I'll try to proove it within next 6 hours or something (till end of the day here in Poland).

Ronin316
offline
Ronin316
7 posts
Nomad

benchbreaker, is your spreadsheet available for sharing, or is it local on your computer? I'd like to see how you calculated your optimal gem creation method. Did you guess and refine, or did you develop a formula for it? Right now, I'm working on a spreadsheet that is currently capable of calculating pure gem specials and am working toward integrating multi-gems and damage calculations.

My ultimate goal is to calculate the tipping point for the different combining methods, and be able to correlate it with optimal amplification. Currently the optimal known method is squishing red and black, and amping with pure black. I'd like to delve into the math to try to improve these figures. For mana gems, I suspect we may end up amping with a mixture of pure black and supergemmed black/orange, while kill gems will be amped with a mix of pure black and woven black/yellow.

Also, I've found that mixing gem colors causes some ugly calculations to crop up, because order of operations has a significant effect, especially when dealing with 3+ components.

Lotoreo: I did not mention that for 2 reasons: 1, it had not occurred to me. 2, I was just trying to confirm that I had the basic method down for each form of gem-building.

cronos51
offline
cronos51
108 posts
Nomad

Hmmmm, the new mathematically approved scheme, enriched by the 'start-to-supergem-from-G8-only'-magic in action..... xD

http://i58.tinypic.com/wumbtl.jpg

e-Mana at G86 (hitlevel 33 only, amps average is G84,5), which makes me laugh...crazy stuff! One wonders where our 'crazy mathematicians' will be able to push this ...

psorek
offline
psorek
447 posts
Jester

Unfortunately not much more. You were using my 16x supergem scheme, weren't you? Optimum method is no better than 2x mine at grade 100 and about 1.4x at grade 50. I'm also working on a good method of initially combining orange/black at start what will give boost like 1.4 or something.

cronos51
offline
cronos51
108 posts
Nomad

Yes, of course I was using your method, just beginning a little later at G9.
Without it I wouldn't have gotten close to this, of course. No way too, to find a scheme like this within the limitations of empirical observing like I did it. So thanks again for sharing

I would have thought that there exists even more sleeping potential, as the possibilities to combine the gems before beginning to supergem seemed to be countless. On the other hand 1,4 would change things dramatically, too.

thunderrider
offline
thunderrider
641 posts
Peasant

Hmm, for the 16-combine method, if beyond the grade 4, when you supergemmed, if you used pure O supergems for that(besides if main gem is G7 and super is 6, ofc) then....would that give a boost?

Also, I think the higher the grade 6 or less mana, the higher grade 7 bonus be, so grade 1-6 should focus on maximum mana increasing, right?

OH GOD DARN IT. V1.16. Broke game. Blood can't be amped anymore. DARN!!!!! I was going to start a new run, now that's ruined! OMG I HATE PETER!!!!!!!!!!! I'm sick and tired and will no longer stand up for him and his darn annoying "updates" which conceal huge nerfs to fun! OH MY GOD I WILL NEVER PLAY AGAIN. THERE'S NO POINT. THANKS PETER

Also - Venom fixed. No level 3.5K+ anymore...-_- I'm seriously pissed.

cronos51
offline
cronos51
108 posts
Nomad

Uuuaah! Thats not funny. In fact I do not at all understand that. It makes the most active and passionate players here turn their back to the game rather sooner than later. I can't see the need to close this playground.

This thread will thereby sadly die, too.
Bad news!

psorek
offline
psorek
447 posts
Jester

don't worry, he is a **** but we still can beat him firing speed should now work as pretended (no cap at 30 hits per second)

ThomasSpeedrunner
offline
ThomasSpeedrunner
155 posts
Nomad

*sigh*

Welp, I think I would do well to never reload the game. I've still got V1.15 loaded. Thus, I can still amp black.

Also, an edit: Thunderrider down there (below me) has a point. Peter, thou hast ruined thy game. If you can't undo the update, 'fraid you're gonna lose everything you had when this game came out. You already came dangerously close to it within the past few updates, but the nerf of bloodbound sealed it for you.

This game is doomed to die off if the latest update isn't undone.

thunderrider
offline
thunderrider
641 posts
Peasant

IHATEUPETERIHATEUPETERIHATEUPETER.

WHYWHYWHYWHYWHYWHYWHYWHY DIDIDIDIDIDIDID YOUOUOUOUOU HAVEHAVEHAVE TOTOTOTO RUINRUINRUINRUINRUIN THISTHISTHIS EPICEPICEPIC GAMEGAMEGAME! YOUOUOU DESERVE TOTOTO GOGOGO TO JAILJAIL!

OMG I CAN HARDLY RESTRAIN MYSELF FROM CUSSING. YOU NO GOOD BRATTY DEV! I SUPPORTED YOU! I BOUGHT PREMIUM FOR LABY BOTH ON KONG ON TWO! ACCOUNTS AND HERE. I BOUGHT PREMIUM FOR GC2 HERE AND WAS GOING TO ON KONG AND TIP YOU 500 KREDS THERE TOO BECAUSE GC2 WAS AWESOME.

BUT YOU RUINED IT. COMPLETELY. I WILL NEVER SPEND A DA*M PENNY ON YOU EVER, EVER AGAIN. HAVE FUN WITH 80% OF YOUR MOST LOYAL PLAYERBASE GONE. ALL I CAN SAY IS, YOU BETTER NOT RUIN IT EVEN MORE FOR THOSE WHO REMAIN.

I GIVE UP. I'M DONE. I QUIT. JUST DON'T MAKE EVEN MORE PEOPLE FOLLOW IN MY FOOTSTEPS. I will only come back if you undo this update. Please, Peter. Don't do this to your game. Don't. You're killing it.


-EndRant-

Showing 301-315 of 496