#Attachment rolling for gold chips
1 messages · Page 1 of 1 (latest)
The odds for each % stat is weighted heavily towards average stat rolls, you can find each chance here:
I created a simulator to roll 1m tries with these percentages to see on avg how many tries it would take to get to some stat breakpoints. You can try it out here: https://colab.research.google.com/drive/11G6t0H3UMILEE26mEkgpP4Vi8AKy4blh?usp=sharing
A quick summary of results (this is not exact numbers)
+-----------+-----------------+
| Roll Avgs | Avg # of Rolls |
+-----------+--------+--------+
| | 3 Line | 4 Line |
+-----------+--------+--------+
| 140%+ | 15x | 30x |
+-----------+--------+--------+
| 150%+ | 30x | 67x |
+-----------+--------+--------+
| 160%+ | 72x | 207x |
+-----------+--------+--------+
| 170% | 215x | 900x |
+-----------+--------+--------+
This considers all rolls must be at least 100%
related discussion where they make recommendations based on the amount of time to get gold chips (this might change with the new bp?)
https://gf2-bbs.exiliumgf.com/threadInfo?id=57325
ahh, nice they have it all calculated out.
Here is for the 3 line pieces:
Here is for the 4 line muzzle pieces:
glad i was introduced to this thread. i've recently looked at this author's work and had some questions. though because i don't speak mandarin i couldn't contact them to get access to workings and clarification:
- the calibration estimates seem to be taking the probabilities conditional on no roll being under 100% [1], whereas we need the probabilities of certain total calibration and no roll being under 100%. the effect of this is that the required number of calibrations are underestimated, indeed, as compared to [2];
- when farming for 3 line sets, we are also farming muzzles, so by optimising sets and muzzles separately, farming time is overestimated and chip expense is underestimated; and
- highest base stat rolls for ATK and HP are collapsed into the next lower level as per older works, and therefore chip costs are not precise.
with that in mind i've replicated the results of previous works by 无音 [2] and 聊以行国 [3] and extended the results to approximate the approach 啾啾喵 [1] used:
- probabilities excluded cases where any calibration falls below 100%;
- giving ATK 5 base roll scores to better reflect chip cost;
- combined farming time and chip cost for 3 piece sets and muzzles into one table;
- allow modification of chip/stamina/gold gain for future estimates; and
- provide an estimate for full 4 piece sets.
i've included an excel file with full workings, and one with results only for ease of use. it would be wonderful if someone can help check my homework 
edit: i've updated and released the tool as a google sheet here https://discord.com/channels/838924736720797727/1367760392842903614
references:
[1] New tuning chip analysis, 啾啾喵, https://gf2-bbs.exiliumgf.com/threadInfo?id=57325, also https://bbs.nga.cn/read.php?tid=43438074
[2] Calibration, 无音, https://gf2-bbs.exiliumgf.com/threadInfo?id=26756
[3] Substats and farming time, 聊以行国, https://gf2-bbs.exiliumgf.com/threadInfo?id=46322
some further observations:
- at this stage it's hard to recommend a solid number to aim for
- chip income is still uncertain due to other sources not yet known (event, new bp, new game modes). the author used 0.625 chip per day but i approximated report of 15 chips per month with 0.5 chips per day (15 * 12 / 360)
- farming and calibration results vary wildly between accounts
- matching chip income and expense results in ridiculously high level of base stat score and calibration requirement, which is not advisable when you still have other sets to farm (this is alleviated by us having a headstart before new chips hit, but we don't know when)
- no one has done any analysis on the limiting factor now that gold is uncapped. for example, is attachment xp fodder an issue?
- this will need to be tailored to each account's chip income and gold income from other sources and frequently revisited
Considering the head start the recommendation is probably going to be "start with your best pieces and the most shareable sets (i.e. phase strike)". Nevertheless, the math to strike a balance between farming attachments and obtaining chips is useful for new farms. It's also good to see the bigger picture of "how much time can I save on farming/calibration if I'm willing to wait a bit longer for more chips"
I suspect fodder is not going to be a limiter since boundary push also yields a ton of fodders without actively farming attachments with stamina (how many runs of attachment stages that income is equivalent to is TBD). I've not ran out of fodder yet although tbf I have really strict locking plans
The recommendation from the BBS is:
Set : Level 50 - Average fine-tuning level 160% or more
Muzzle : Level 65 - Average fine-tuning level above 160%
However, that is quite a high bar
50+ according to the graph is 146 rolls or nearly 3 million gold per set piece
and 65+ for the muzzle is nearly 290 rolls or close to 6 million
getting above 48+ seems much more achievable as its only 72 rolls
was surprised this was so high when their attachment recommendation was so low:
Level 7+ seems like nothing
The NGA post did have 49/63 for pack whales and 45/60 for newbie/stamina-eco
45/60 is 30 rolls per set attachment and 67 for the muzzle
which is definitely much cheaper 
if u farmed as a f2p for an entire week on gold, including all the different sources, you would get what like 1.5m?
It's about 150k per day on natural stamina only so closer to 1m
its 12 stam per hour so 12 * 24 * 7 = 2016 stamina a week
10660 per run of the gold stage i think
also 50k/week from guild, 25k/week from weekly boss.
dispatch is 6k per day
it seems appropriate at this stage to find all other sources of gold and chip incomes and add them to the estimates. this should not affect farming time estimates but have some significant effect on calibration.
yeahhh
i guess mica can't just release this chip and give everyone 'enough'. they probably reserve some for future game modes/events/packs
which i think is the main deciding factor here
at 15 chips/month we are essentially going to take 4.5 months to get a perfect set as f2p if we go with 150% avg stats
actually more since we need base stats too
which is a long time to farm gold
tbh the present income is already more generous than I was expecting for a something that guarantees improvement. but I guess now we also have the collapse flower thing to worry about 
true 15 gold chips a month is a lot
I think the uncertain, recurring sources for gold chips at this point are
- minor events
- new tarkov
- ?
yeah, tarkov is variable too so hard to judge
does it reward based on difficulty? or milestones?
do you mean the current one we have in cn now or the flower thing they are about to implement?
i only play on global so totally unaware
what's the current one like?
current you can farm up to this amount of chests a week, they each have chance to drop gold in quantities of about 800-2k
not that ive seen
ah that a shame 
thanks
if you happen to remember any other sources of gold not available on global yet please do leave them in the thread i will add them to the sheet
hmmm
i can probably go and check the global ones in the following days
i think if we know it's a recurring source at regular intervals we can add them
the shop also sells 100k a month in gold can buy for 200 tickets 
do you guys have enough in cn to start buying gold in platoon shop yet? 
oh yeah i've heard. 23 days instead of 30 now
it has the gold chips in it too now so some other stuff changed
actually if you click into that question mark it gives you total amounts of gold for each tier per week. Also states that the attachment boxes are the same probability as rolling them in attachment stages
ohh
just found out 
if you're already satisfied we don't really run out of fodders i guess we could just focus on chip and gold for now
28 boxes
aren't they reworking this too?
that's like... 2 attachment runs
yeah the flower thing. they showed in the preview that it has a similar reward structure
uhh theres going to be a new thing i dunno if they getting rid of this one
guess we will see after the update
the flowers are supposed to be another method of upgrades
but we dont know much yet
Also, for players who have been around since tarkov mode first came out everyone should have their characters max affection by now
hopefully we can do something with the extra gifts
is that flower the collapse epyphyllum something related to the remolding pattern upgrade system?
ya
actually now that I see this pic again it makes me think this is in addition to old tarkov... since these rewards are lower. Was wondering if it replaced old tarkov with higher difficulty & more rewards
nah u can see the old tarkov name in the tab in top right
so it seems u can switch between
hmm right
seems strange but for whatever reason they blurred out something in this pic
in the top left
the bonuses on the right really sound like elemental bonuses
iirc they blurred out the 50k gold cost when they previewed the new chips too. probably something not yet finalised 
Someone pointed out b4 that they look like monster hunter buffs
lol
lol
Can the new chips max out any piece regardless of substat roll, or do they just send them to 200% and a non max sub will always suck?
it boosts base stats too
it boosts base stats (3 to 4 steps) to max first before you can start boosting calibration
Nice, thanks
the new pic today has different buff names
literally atk/crit/elemental buffs lol
"throne shatter" sounds like def pen maybe
nice
how many can you get from the battle pass? free vs paid?
theres some more too
where else again
2 from gunsmoke?
yah 2 from gunsmoke
In EN
Weekly:
Platoon Shop - 1
Monthly:
Growth Stack - 3
Per Event:
Event Store - 6
Paid Items:
Paid Items - 1+3+10, a total of up to $65 USD)
In CN Now:
Battle Pass (F2P) - 1
Battle Pass (Paid) - 2
Gunsmoke Frontline Individual Rewards - 1
Phase Breakthrough (Purification Rewards) - 1 per week
i cant tell if its RNG but you definitely get more than 1 total
we also got 2 from the relic recovery but dunno when that resets
ya
Uhhh we’ll chalk that up to random patch events ig not consistent enough
so for CN calcs as of RO patch
Weekly:
3~ estimated from Phase Tarkov
1 from Platoon Shop
Monthly:
Growth Stack Shop Reset - 3
Per Event:
Event Store - 6
Gunsmoke Milestones - 2
Free BP - 1
Paid:
$10 for Paid BP - 2
$5 for Pack A - 1
$10 for Pack B - 3
$50 for Pack C - 10
Monthly income:
(calculated by 8 weeks, 3 events, 2 months divided by 2)
32+ a month (F2P)
35+ a month (BP buyer)
56+ a month (BP + All 3 packs buyer)
Event modes like Roguelike not included

what does level here correspond to? can't figure out how it maps onto stats.
think like every 10% is 1 lvl
set is 3 stats up to 200% ea stat
so a piece with 160%+170%+170% would be 500% total or lvl 50
thats what they recommending you aim for before using a chip 
gotcha, thanks for explaining
level 50 before chip is wild
i guess if you have pieces that are close to max base stats, so way higher than level 7 from the other image, you can effectively do average a level lower level plan though
so i guess my strat will be to farm good base stats and average 150s or something
I think this is the best place to share this.
Made a little site to quickly check an attachment's calibration from a screenshot
https://morphthemoth.github.io/GFL2-Attachments-Calibration-Check

Wow what u using to read the numbers from the screenshot?
ahh it's a GitHub lemme looks
it just searches and counts the pixels from the white and orange lines
it can get it wrong rarely cause the pixels arent perfect and the white line in the middle could mess up the count a little
that's what the number in the parenthesis is for, its what the ratio is before rounding it, if its close to 5% it could be the 10% up or below, but it did work for all the ones i tried
I was wondering if u were using ocr 
this reminds me someone shared another way to check #🧮theorycrafting-en message
still, it's wild that there's no easy and direct way
If u open the gold chips menu for a piece it shows the stats too
Menu only available if its stats are over 100% ea
oh cool. that's not terrible then 
dam, thats ultra cool @distant sandal 🍻
I would suggest adding it to the general guide section. It's an extremely useful tool and a lot more useful than it just being somewhere hidden in the middle of a discussion on gold chips.. but it's your call.
i didnt post it initally because its not really a theorycrafting thing, but actually yea i agree i should make a post about it 
I'm confused about something, can someone clear it to me
how many ticks does flat attack needs to max the base
flat atk and flat hp have 5 possible base rolls. so maximum 4 chips to go from min to max
the rest of the substats have 4 possible base rolls, so 3 chips from min to max
I'm confused because I've seen more than 5 base rolls
that's interesting. it shouldn't be anything outside of these values
https://gf2-bbs.exiliumgf.com/threadInfo?id=46322
I've seen 20 and 22 flat attack
could it have been atk% or another flat stat you saw?
i see, that's interesting
could you take a screenshot of it next time you come across it?
thanks, these are great
i wonder if it's some funny rounding or if they adjusted the base rolls to be all 4 to simplify things
could take a look at the rest and see what possible values there are
what's strange is that these are exceptionally rare
i have 600 attachments with flat atk but only 2 of them got 22 base flat atk
the ideal situation is that it's just a display thing due to rounding, or they are grouped together, ie 19 and 20 both take 1 chip to get to the next break point at 21.
if someone playing on cn client can verify this it would be wonderful. otherwise i will revisit this when global patch drops.
just to confirm using one chip on 19 gives 21 right?
based on this from the preview stream, it takes 4 chips to go from 17 flat atk to max base, so likely tuning would go to the next break point
but ideally this is verified in game
tuning from the rare cases of 20 and 22 atk however will be tricky. will be tough to find an atk cr atk% calibrated to 100%+ with that specific base atk 
unless someone makes the sacrifice for science

i know this is asked alot
but i cant find it anywhere
i get the "target 150% per substat on average"
but what about base stats?
i found out that some needs for example 3 while others need 5 to max a base stat
i think there is a sweet spot between farming for calibration to obtain higher base stats and using that same energy for more Sardis gold to obtain higher avg% rolls
although this seems quite low requirements even for the standard plan
so basically if you had an attachment with base stats:
21 ATK / 3.0% ATK / 4.0% CR
Thats 3 + 2 + 2 or 7 points
7 for pieces and 9 for muzzles
aight , i will target 8 for pieces and 10 for muzzles
till i find my chip getting stacked unused
thanks
it goes from 20 > 22 > 24
so it skipped 23 and ended up being the same number of steps
like
17->19->20/21->22/23->24
oh wonderful that’s great to know thank you 
so for our purpose it’s safe to treat 20 the same as 21 and 22 the same as 23. i was worried we’d have to add more combinations for some small chance of dropping these 
Did it start at 20 for some reason? I thought it was 17,19,21,23,24
wonder if it depends on your client
Seems odd that I wouldn't see one in so many
it's very rare
https://gf2-bbs.exiliumgf.com/threadInfo?id=68763&hash_flag=1 LE GOAT HAS POSTED A NEW GUIDE
Imma translate this probably this weekend
me when I'm using my bachelor's in statistics to its full extent:
So basically, from this guide it seems that the advice is to reroll the piece until it needs x chips to max (2nd numbers for muzzle):
Standard: level 7/12 - 9/16 base stats (5 - 7 chips), level 50/60 - 65/80 rolls (10 - 15 chips) = 16 - 23 total chips needed
Kriptonian(?lol? wat is dis name): level 6/12 - 8/16 base (6 - 8 chips), level 49/60 - 63/80 rolls (11 - 17 chips) = 17 - 25 total
Newbie: level 5/12 - 7/16 (7 - 9 chips), level 45/60 - 60/80 (15 - 20 chips) = 22 - 29 total
I feel like these plans are not "calibrated" well, the 2nd one is way too close to the standard one, and the beginner is too far with too many extra chips lol
For example just now I rolled an AR sight that now needs 20 chips to max, according to these plans this would not fit into any plan, it would be very good compared to standard but very bad for the other two, which doesn't feel right imo
I think the stats to aim for rely heavily on how you are farming. The original guide was created in a time when gold stage was still limited. If you mainly farm gold stage might want to aim for higher avg rolls.
Still 7/12 base stat rolls still feels very low
My strategy is to aim a bit lower on the roll avg and higher on the base stats. i think 9/12-12/16 is pretty reasonable for base stats and 45/60-60/80 minimum for the attachment rolls. That gets you around 18-24 to max pieces.
With the odds of getting a 50/60 piece or a 65/80 piece you could only likely roll 1 attachment with those stats a patch.
And that's with farming gold nonstop the whole patch
anything above 350% is wasting money if you are using boost chips
did the author explain anywhere else the rationale behind dividing the expectations by 2.5 to account for gold cost of calibrating set pieces using boost chips? 
i guess thats for the gold cost, 50/20
mhm, i was thinking about that. but it's 44k for 3 line and 52k for 4 line, over 20k without boost chips for both
the gold cost of using boost chips
right. what i don't get is how he arrived at that 2.5 figure when gold cost using boost chips of 44k divided by normal gold cost of 20k gives 2.2
since i'm effectively illiterate here, if you come across any mention of workings for this bit please do let me know 
Honestly there are so many jargon I might as well be half illterate 
https://gf2-bbs.exiliumgf.com/threadInfo?id=68763&hash_flag=5 the original thread
oh sweet thanks 
For reference I have the breakpoint at around 400-410% from random simming https://docs.google.com/spreadsheets/d/1CVY5A5fT2628D4RdnCZYUmYMaEAwBca0hc9mdm0oFGQ/edit?usp=drivesdk
The difference is probably just from the gold cost difference
yeah same. so i wanted to see some workings. his basis is also slightly different. he evaluated the expected gold chip saving if calibrating once more at each level
which does not take into account the difference in cost/chance to get to that level in the first place, between using/not using boost chips
so not sure if it can be considered a complete comparison between the two approach
then again, this doesn’t contradict the generally accepted norm that boost chips are a scam. so there’s no real rush to investigate further
I am just glad we have a the consensus that we should stop at about 350-400%
when you say 350% - 400% you are talking about just 3 piece?
yeah
mfw spending a million gold to roll 50x per piece
and i dont stop until at least 480%+

wait what do you mean stop at 350-400%, isn't this for boost chips, not gold chips?
yeah i think its fine to sit on 400% for general gear but if you plan to use gold chips that definitely not enough.
A Simple Analysis of the Attachment System in GFL2 Written by 聊以行国 Translated by DiggusDBickus Forward from the translator: It takes an insane man to write an essay-length analysis about a single system in a gacha game, and it takes a doubly insane man to translate said analysis, especially i...
some1 told me this is not good enough for gold chips
#🌄gfl2-screenshots message
bruh momento
it really depends tbh
since gold chips are the bottle neck, it might actually be better to get a better calib
thats my 2nd muzzle (1st one was 4x perfect roll base + 200/190/170/150, dont ask why, Lord just smiled on me), so i just got the base to perfect lvl and left it like this x(
570% for muzzles are the awkward zone tbh
the 400% is really for normal attachments and even that was assuming you are prioritizing gold instead of chips
its so much easier to roll good on 3stat attachments x(
i wish they lifted additional cost on orange chips chipping
mate that means my translation is worthless if they do that 
I just hope they don't over simplify the calib system
its already simplified enough (compared to all other gacha) and dont get me wrong - i am greatful for having gold chips, thats like a fresh breath of air compared to all this hoyo sht
i jsut wish they listed cost on orange chips chipping
i wish other games had stuff like chips. wuwa i'm sitting on more resources than i'll ever use and have almost 0 reason to farm for anything other than to just spend the stam bc dailies tell me to. would be fantastic to have massive resource drain to perfect my equips
gfl got me panhandling for gold for the next attachment roll lmao
thats so true x(
i dont even have enough gold to cap crit on this one x(
I saw the spreadhseet over in the other channel
with their guide the optimal path is rolling for 490% for set pieces and 630% for muzzle
pretty steep tbh
essentially 160% avg stats
which would take about 16 gold chips per piece for the set and 25 for the muzzle on avg
fml
part of the process (is what I cope with)

how bad of an idea is to chip 540% muzzle but it requires 8 chips to get perfect base? x(
8, so your base now is 3 on average
that is an awkward position, yiu want around 600 on calib, and your base stats arnt that good either
shouldve goetten a better muzzle before calib
been farming smg attachments for quite some time, got tons of quite good triple offensive stats scopes/foregrips/underbarels but not luck with muzzles x(
from earlier today for example
no beuno
time to reroll some more
worst thing is that i have enough other parts for both leva and lenna
but only 1 scope
and its el garbanzo
