Chromatic orbs are bullshit.
I think color predictability can be maintained without the odds of SIX ON-COLOR being 30%.
lol there's bias...and there's WTF BIAS. right now we have the latter. edit: " It is correct that the system rolls affix types first, and then the range. However, they are experimentally known to be weighted, which complicates things greatly. Unfortunately, I doubt anybody has really nailed down the weights. Last edited by tikitaki#3010 on Oct 3, 2013, 7:40:12 PM
|
![]() |
" Well now I feel like an idiot, didn't know you were thread maker x) True that off colours are not always better, but it's somewhat similar with links. Links are always better, but not "needed". Some skills only need 1-2 links, and same goes with colours. Some only need 1-2 certain colours to function properly, but in order to get something "good" out of it, you need to roll/link against the odds. For example, some skills can run just fine on 4l, or even 5. 6l is just a luxury; the same case applies to colours. Some pieces of armor can have 4 of the colours you want to roll, but you want the luxury of having 6 of the colours you want. You will have to work harder to make it work the same way that links do. PS: I understand what you meant by saying colours are colours on any piece of gear, but that isn't a realistic notion. It's like rolling a 100% Spell damage 1h axe or something and saying "100% spell damage is 100% spell damage." It has to be within context, and in this case, it's rolling the right colours on the right item. Last edited by Dinivus#3418 on Oct 3, 2013, 7:45:45 PM
|
![]() |
I made a tool so that you could see what effect different Chromatic ratios have on the probability of various socket combinations, useful for both suggesting new Chromatic ratios and for estimated Chromatic crafting cost.
https://docs.google.com/spreadsheet/ccc?key=0AlQ8leXHqxMxdENCVEZlSWFoSVJkdkVEVHpHUEloTUE&usp=sharing edit 1:56am GMT: Fixed major math error in spreadsheet. When Stephen Colbert was killed by HYDRA's Project Insight in 2014, the comedy world lost a hero. Since his life model decoy isn't up to the task, please do not mistake my performance as political discussion. I'm just doing what Steve would have wanted. Last edited by ScrotieMcB#2697 on Oct 3, 2013, 8:58:14 PM
|
![]() |
" Thanks! I found this spreadsheet very useful. The default color bias it uses for pure armor (8:1:1) gives odds close to what I've seen in rolling Chromes. One point of interest: there appear to be inherent "break-even" ratios in balancing the relative odds of getting the most likely combos: * Most likely Pure combos: 6-0-0 and 5-1-0 * Most likely Hybrid combos: 3-3-0 and 3-2-1 Pure armor: 6:1:1 gives equal odds (8.8%) to both 6-0-0 and 5-1-0 combos. Hybrid armor: 3:3:1 gives equal odds (12.4%) to both 3-3-0 and 3-2-1 combos. * With color bias more extreme than 6:1:1 and 3:3:1, odds of getting zero off-colors (6-0-0 and 3-3-0) are greater than getting one off-color (5-1-0 and 3-2-1). * With color bias less extreme than 6:1:1 and 3:3:1, odds of getting one off-color (5-1-0 and 3-2-1) are greater than getting zero off-colors (6-0-0 and 3-3-0). Last edited by RogueMage#7621 on Oct 3, 2013, 10:37:54 PM
|
![]() |
"Yep. Here's what I think the ratios should be: Mono: (4 to 6):1:1 Hybrid: (3 to 4):(same):1 Conservatively, I'm leaning more towards 6 for mono and 4 for hybrid. When Stephen Colbert was killed by HYDRA's Project Insight in 2014, the comedy world lost a hero. Since his life model decoy isn't up to the task, please do not mistake my performance as political discussion. I'm just doing what Steve would have wanted.
|
![]() |
More statistical information:
Item used: Vaal Regalia (pure ES chest armor), 6 sockets, 5 linked Desired combination (on 5 link): BGGGR or BBGGG Chromatics spent so far (a bit more, but didnt count results for the 1st 100-150 chroms, so exluding them): 820 Results: 6x Blue: 287 (35% !!!) 5x blue: 193 (23,5%) 4x Blue: 211 (25%) 3x Blue: 129 (15%) 3x Green/Red: 0!!!!!!! Sorry, but I strongly agree with the OP, the chromatic orbs are bullshit. The game is all about build diversity, I can understand 6 socketing an item, or 6 linking an item to be hard, that's fine. But getting the desired colors in no way as powerful as making a good item 6 socketed or 6 linked, and thus it shouldn't be this hard. In my specific case I'm doing a physical wander build, a build which is by itself a bit hard to make when it comes to the right items, getting a decent wand is nearly impossible considering all the different combinations of stats a wand can have. On top of that wand attacks are counted as attacks, not spells, hence a faster attacks support gem is required and essential to the build (along with chain/lmp). With the current chances when it comes to chromatics its kinda a "get a hybrid armor or GTFO" policy. In my opinion the current chromatic orb coloring calculation is purely there to artificially remove more chromatic orbs (which are a low value currency anyway) from the game for no apparent reason, other than artificially inflating the value of them. With no other options to get the desired combinations of gem colors onto gear (considering there is no way to "lock" a sockets color, nor a way to boost the chance to get a specific socket color, nor there is an orb that truly randomizes the colors) its like playing a rigged poker game. I really hope this thread, and the statistics in it will find its way to the devs and make them improve the odds in our favor a bit (tho it is a bit disappointing to see a 9 page thread with no dev reply, especially one with some many constructive ideas and factual data). Last edited by Palaryel#6900 on Nov 11, 2013, 7:54:53 PM
|
![]() |
Bumping to inform the geniuses (see incompetent game designers) at GGG that I've quit the game after losing every last currency to this. Congratulations for making the one orb that opens totally new build possibilities carry with it such an aggressive weight on rolls that those possibilities are gone. You claim your game is open to customization then code this system in. Fusings were garbage enough, but at least we could blame that on RNG. This is straight up coded to be oppressive. Well I've had enough.
My biggest regret was I shelled out $35 in micro transactions to learn the true shitty nature of this game. Well I'll be sure to inform everyone I speak to who shows any interest in this game of this nonsense. Your devs are failures and you should feel bad. Peace, farckers. Last edited by RandomBullocksGenerator#0712 on Nov 26, 2013, 8:35:06 PM
|
![]() |
I give this my +1.
I have had to go into a hybrid gear even though I am a pure evade build just so I can get some colors that work. Using some 500 chroms didnt do it on my pure evade chest... not even close. And if you close your eyes, does it feel like your exalt almost hit this time~
|
![]() |
Reminds me of my Deerstalker boots, spent 2 chaos on them, trying to get BBBG on Evasion boots. I can't use it on chest or weapon those are taken this makes my build or breaks it, ended up spending around 847 Chromatics on those cheap boots which were essential to my build.
Never doing that again. |
![]() |
"Trying to recreate this data spread using my Google Drive spreadsheet tool (link earlier in this thread), this makes it seem like bias is somewhere from 10:1:1 to 20:1:1. I will grant you enthusiastically this much: if the bias truly is something like 15:1:1, then the bias is insanely too strong and should be toned down to something more like 6:1:1. However, the sample size isn't particularly large, and perhaps you just got particularly unlucky. In which case no adjustment would actually be required. Perhaps a community Chromatic Orb log is in order to more accurately pin down the current color bias, rather than relying on rumor and speculation. When Stephen Colbert was killed by HYDRA's Project Insight in 2014, the comedy world lost a hero. Since his life model decoy isn't up to the task, please do not mistake my performance as political discussion. I'm just doing what Steve would have wanted. Last edited by ScrotieMcB#2697 on Nov 26, 2013, 11:19:38 PM
|
![]() |