Disclaimer: This is NOT a guaranteed way to have good upgrades on your relics. This is simply pattern reading. You do NOT have to do this if you think it's too complicated. But this is BETTER than nothing, than just hoping for blind luck on your upgrades. If this works for you, then great. If you feel this isn't worth the trouble, then all the power to ya. Have fun in your own way without ruining other people's fun. Stay strong, fellow trailblazers 💪 Small Note: As the video say in 15:50, the data is still small and new. If you are a casual who wants more precise outcome, do this later in 2.1 when more data/patterns has been gathered.
The more data you have the more your realize that the difference in getting stats is rng. They have a very small sample size, if they had a sample size of 100000 people trying this then then number will get closer to 25%. The numbers are skewed because of the small sample size.
@@gabiblackwe want this honestly…. Hard to convince people to burn materials for data. Most players are stingy or are too lazy to even contribute to what we do. My own discord has around 3000 players. Only around 100 of them contribute to the data. Let that sink in. Now if everyone in the discord contributed to the data it would be a different story. 3000 players doing one pattern string a day would easily get what ridiculous number you want for your statical needs towards the end of the patch. There’s only so much we, who contribute data, can do especially with how our own resources are.
@@BigBoiPnoyQuestion, why does data need to be collected again after every version patch? If we're thinking that any change in the code can affect the patterns, wouldn't it mean we need to restart data collection every time there's a hot fix (which is almost daily)?
@@rinoakirova1548it was honestly what we noticed from our own testing. Our thought was since the server goes down, most likely the pseudo “rng” gets re-randomized when the servers boot back up. A hotfix won’t be what will change that rng system. Unless they directly aim for that part of the game. Now… if the whole game server went down during the middle of a patch…. Then ya data gonna have to reset too… In any case you’ll notice if you go through each patch on the data sheet. No day (let’s say Wednesday as example) is similar to any other Wednesday of a different patch.
@@BigBoiPnoytbh tho, the data has been somewhat consistent every patch, with each line being almost the exact same each patch. The only big difference I’ve seen is the Rng per UID, which makes personal data skew either top-low, middle low or high and low. And From those assortments, go from there. Say example, u start with a 4xxxx, and like we talked before, depending on your server, you could get more likely odds of a certain roll. But I’ve noticed that it also tends to cycle more often the. Not. In 10 relic manipulations, 1/2 of my rolls went to 42xxx, 1/5 44, 1/5 in 41, and 1/10 in 43, which was on a Saturday, when 44 & 43 are high, with most lines from there, tending to be closer to middle-rangers then higher or lower, being 4243x or something like 42441, which would be similar, or almost exact repeats every couple of cycles, or the next cycle after that. 2 of the 42 lines were back to back, with the other 3 being somewhat eacenly spaced apart, happening on the 2,4, and 7 relic rolls respectfully. Idk if this would help, but the prospect that depending on UID, your lines could also cycle on old lines is very interesting, b/c if this is true, you can use both the data known, and your own experience with your UID to determine the most likely line upgrade, especially on a specific day. The only saw these cycles was b/c I had enough relic fragments to do it, (700) and In turn, allowed me to build a a 220 BE Ruan Mei, which isn’t even at maxed relic lvl. I wonder what else we could be missing in the relic manipulations? Either way, keep up the good work!
Even though an RNG factor does still exist, knowing that we can have a bit of control on it is reassuring and keep me from the anxiety of rolling upgrades. Hope you keep getting more data sample on the docs!
Been using this since 2.0 and can confirm that i'v already made alot of godly relics. It didnt work 100% of the time, but for sure targetted some insane pieces and hit the "expected paterns". It works, and im a big fan, even at the point where I'm SCARED to upgrade w/o this spreadsheet info. I just hord my stuff untill I can do a few upgrades but ONLY with these chances in mind. I fucking hate relics, relic farming and the fact you're characters endgame power comes almost exclusively from it.. This helps ALOT! It cannot be underestimated, people are fools if they still think this is fake or "a myth" just roll and be lucky.. Thank you, brother!
That’s what I’m sayin. I’ve been playing for a year and the relic grind is such a turnoff for me. But I like a lot of other stuff about the game. Being held back by relic stats is so discouraging.
I think it is perfectly natural and understandable for this to create a lot of skepticism to those who find this guide for the first time, such as myself. Though the genuinely overwhelming amount of positivity you receive, very much intrigues me. I always say having good relics is far more favourable than having the best characters. So a method like this working for the majority of people gets the heart pumping. I gotta admit that i tried an upgrading session using this method and most of my rolls still ended up completely random. I get that its too early and there's insufficient data currently, for as i was consulting the spreadsheet, the % chances were almost close to one another so any of em could've hit really. Or maybe i got the bad uid day, which would really suck to happen to someone trying this for the VERY first time. Ive saved up some good crit pieces and perhaps i will try this again towards the later half of the patch, as you suggest. I truly wish to become part of the crowd that cheers on how successful this method is, and praise you for finding such an unbelievably invaluable gem through sheer hard work and observation.
Having a background in coding did help me in trusting all of this information since you can't truly randomly generate something with scripts, so although not 100% reliable, it still follows a pattern.
@@zerefuuutvall they would do is make it actually random, and then it would just prove the pattern-recognizers correct because these methods would no longer work
Bernard, you absolute Chad. I started HSR late in 1.4 for the PS5 release, and just today, thanks to your guide with some juicy substat roll, i finally fully 36/36 star MoC for the first time! Following your guide, 9/10 next substat patterns were all correct, we take those all day, the extra Crit Dmg rolls put my Acheron over the top! Thank you so much man 😭😊
As someone who has been using this since last patch, I just wanted to say that I think this method is more favourable to farm a large number of relics, pick the ones you want to upgrade, then use this method to upgrade, instead of upgrading relics one at a time.
thank you for making videos like this explaining relic manipulation, even showing so much evidence with testing. i’ve always had the thought of substat rolls being predetermined and came across your channel and everything just makes sense, thank you!
4:41 the reason you have to be connected to hoyverses servers is because they want the rng to be server side not client side. If it was client side you could disconnect from hoyverses servers and when you reconnected you could tell the servers that you got a specific roll or something. This also extends to the pulling system. All modern games do this so its harder to hack. An example of a game that is client side is cookie clicker. you can just inspect element to make the game think you made a bunch of cookies. Overall I think you have the right idea. Nothing is truly random and every outcome is predetermined. However this proudly wont be able to predict your relic rolls as while yes adventualy you would be able to find a pattern I think hoyverses rng system is designed in a way so that it would take hundreds of years before you could find a pattern. I think whats going on is selection bias where your brain remembers times when this a roll fits your hypothesis and disregardes evidence that contradicts your theory. This also applies to people in the comments. Our brains likes to only remember the highest peak and the lowest rolls. You wouldnt write a comment about failing an artifact right? For me to be able to believe this I would need to do my own testing. As a uneducated observer I think its best just to ignore this. I have tried this before and my results are mixed
That’s totally fine. Gamers should have fun how they want to. But ya it’s all server side pseudo RNG. And for many players, they prefer having this over going full RNG. It’s like at least they know the possible outcomes. I already said this isn’t a guaranteed way also. Mainly cause we aren’t able to fully see what numbers are actually on the server, that’s straight up impossible. But overall, yes I do have some failed pieces. At the same time I have a lot less failed pieces than what I used to get when going full RNG in the past during the beginning of HSR as well as the genshin impact days before I quit that game.
I agree with you. I kept an eye on the data they've collected for last patch and the numbers they are getting are not statistically significant enough to prove it is anything but random, even disregarding the likely selection bias like you mentioned. They also attempt to write off the times the rolls do not line up with their determined patten by labeling it "special uid" which further increases the bias.
Maybe. Honestly we can be wrong about all of this, and it can be totally something different from what we think. The goal overall was to help players roll better. And honestly if we have achieved that we are happy. In the whole factor of the data gathering the sad thing is… a lot of players don’t want to contribute data. My own discord server has around 3000 members out of that 3000 only 100 are contributing data…. It’s a sad reality of players just using what we gather to their advantage. Same time can’t blame them, we burn materials to get data and that’s the only way… we want data to be 1000+ each day. But if only a little amount are contributing, that goal won’t be achieved.
As soon as I finished watching this video, I went to upgrade a piece which was atk% with CritDMG/CritRate, speed, break effect. I forced it to 2 then got another 2, and then again and then crit damage. Basically 3 hits on CritRate 18.1%, 2 rolls on CritDmg 19.4%. Thank you for your blessing!
@@---ie2msyeah because 55 cv is impossible? If you start with a 4 substat relic that has crr at 3.2 and cdmg at 6.4 you can get 5 more substat rolls which if it rolled to the max possible value would only amount to 32 cv(from the 5 substat roll) and 12.8 cv(from starting value) which gives a max of 44.8 cv per relic, and this is the absolute best case scenario, assuming op isn't just flat out lying I'd assume they miswrote the crr value since 3 rolls on it could only be 3.2 x 4 at max possible value which lines up to be 12.8 crr and 3 roll for cdmg would be 6.4 x 3 which would be 19.2 at max possible value, i don't remember if there were any rounding error but it's possible that the cdmg stat is possible while the crr isn't, and surprisingly if you do count the substats that i had mentioned it would be (12.8 x 2) + 19.2 = 44.8 cv as previously mentioned
Dude I’m so pissed I just found you’re channel today, I’ve deff been missing out but the data and proof don’t lie. You’ve earned a sub from me my man keep it up
Just now watching this.. had no idea this existed ty for the info.. I understand exactly how to do it but no computer I will keep in my head but I understand the formula u are using.. much love for this video❤!! It's wild that people don't understand that is the best video I have ever watched of hsr!! This knowledge is gamebreaking
Bernard doing the god’s work again. This is dissertation lvl of data science. You & your team has contributed so much to the HSR community. May life bless you with luck.
I can confirm that I used this doc Saturday and rolled 3 times on crit rate on a crit damage chest piece. Honestly lowering the RNG to a 50/50 has been a life saver. Thank you so much.
Relic manipulation is one of the main reasons why my love for HSR sparked again. Rolling for good relics is a pain already but knowing there's a better way to tackle the most frustrating mechanic of the game keeps me going. Thanks for the vids.
Thanks man. My videos is also directed for players like you. I’ve heard all the story’s. Including my own best friend. “I quit cause I couldn’t find good relics for my characters”
Didn't have any relics left to try it on for the current characters, so I tried to prepare something for Boothill and it went really well. Like 25% BE+cr+cd boots and other few pieces with 0 rolls into bad stats. Could've been just luck, but it definitely didn't hurt to try this method. I'll definitely apply this to future chars, but probably won't refarm for my already ok-ish equipped characters.
Tried the manipulation, and I often got an upgrade on a single crit relic at least 2 times. Meanwhile, going full RNG on double crit relic just ended up as a perfect dodge on crit. Thanks for the guide.
Left with more questions than answers: 1: straight pull up a note pad. What is this? What we use that for? 2: excel are full of jargons and disclaimers without coherent description of what to do as a newbie from beginning to end. Can you please do a 1 to 10 sequential step etc of what to do? Example: Step 1: upgrade a fodder? Step 2: then go to excel column xx field xx do what? Step 3: xx Step 4: xx Step 5: upgrade the ssr gear
So the notepad is just an example of how the pseudo RNG works on the hoyo server. the fodder i upgraded... if you were paying attention to what i was saying. was to force the "upgrade line" to the 2nd line out of the 4 lines( the blue relic) most people will have relics with 3 line purple and gold relics which when upgraded forces your upgrade line to the 4th line out of the 4 substat lines. the 2xxx in my actual example is how i start my pattern cause i upgraded a 1line relic( a relic that has 1 substat) to force the pattern to start as 2xxx. Then using the data recorded by other players from other servers( mind you i was using EU and ASIA data only cause america didn't have data yet for this video so this proves region does not matter) we have and i predicted my upgrades. the xxx is the unknown numbers hence why i type it as 22xx, 22xx, 23xx, 23xx, etc. Each time you ADD as substat your pattern resets. thats why i kept going back and leveling a relic with 1substat. its to force the "upgrade line" back to the 2nd substat line. switching relics doesn't change your pattern string from the server. hence why i'm always switching relics. There's documents made by other players on the first page of the sheet to read as a guide for dummies. you have to read though...
I saw ur last vid on 2.0 and i was so suprised that like 90% of the rolls were correct thats how i got my first 160 speed character and alot more. Thanks so fucking much cuz u legit made my acc so much better i will never look at relics the same cuz like i said 90% was correct for me personally genuenly.
Another banger Video, still waiting for the Data at the end of the patch... Planning to hold all of my relics until the and of the patch (especially we have 3x planar and relic drops in this patch)
I'm so glad UA-cam recommended your video to me! I just CAN'T WAIT to put this into action too, but for now, I'm only in the pre-farming stage lol Thank you and everyone on the Discord team so much, you're doing godly work for us 🥰
This works! Thanks bro. You saved me a bit from my rng hell. 😂 Time for me to wait for good days. Now I can save those good relics and upgrade slowly through weeks. Big plan ahead. Woohoo!!
In some way on of your vids go recommended to me 2 or 3 days ago, i started watching it with a lot of prejudice; ended up watching like 5 or 6 more vids about it, so yesterday i tested it on 10 crappy relics, playing it safe and what not, followed your explanations/examples and..... wtf dude every single sub stat that the spreed sheet told me at it was going to hit, dude it did, mam if only i new about it from the start. Hey props to you and all the ppl that contributed to this project/spreed sheet. Just hope Hoyo doesn't see this and change it all up.
man i remember genshin having groups of substat readers too. we just clutch at whatever might be better than pure randomness lol. hope you guys succeed at this
i've been told that theres a group on bilibili that does something similar to what we do but for genshin. And they've been doing it for awhile on that side. Essentially meaning... CN did it first.
Aw man, seems like I'm on a Special UID day the one chance I had to try this out before the update. :( Appreciate the effort that went into recording all this information despite all that!
@@BigBoiPnoy I was assuming patches roll out with new banners like Adventurine coming up? If it doesn't I'll give it another shot later this week since my Monday RNG was being special and Tuesday didn't look too hot
So I decided to give this a go to test this out.... Seems like I got what you guys call "Shit Days" and kept hitting the lower percentages more often instead. I don't know, the fact that such "exceptions" even exists for such a thing makes me question how reliable this rng manipulation trick is. I can believe that there is a form of "patterned rng" happening with the relics though. But I'm going to mainly keep tabs on the data to see how all of this goes rather than rely on such methods atm since I'm not convince that this is reliable as of now.
Hey man at least you tried it. But ya. Most people stop upgrading from there when their patterns are the low %s. Other players in my own discord adapt and just purposely aim for the low %s and it works out for them I basically told players to do test strings on purple relics that have 3 substats already before they touch their own good relics if they don’t want to mess those up.
As someone who is trying to do this method I want to leave a cool hint I learned: Get as many relics as you can and store them to do all this rng when the day is favorable. My greatest problem right now is the lack of relics to either use them as fodder or to upgrade the right rows. For example, I had a great C.rate relic and the next line would be either 2 or 4 (higher %). They where useless stats on that relic since 1 and 3 😂where the good lines. And I didnt have any other relic with good stats for 2 and 4. Guess what? It rolled a 4. So I wasted a 29% roll. Hoard relics. Even those who looked trash at the start. (I got a 1 roll in C rate and 3 in C dmg in a relic that started with flat hp and flat atk)
First time doing this and I was confused ngl but I went slowly and as you did... Good god it works. It's like a drug that I can't get enough of. Only thing stopping me is getting actual good relics though lmaoooo
Thanks, I helped a friend manipulate a few weeks ago and now they have 4 characters over 161 speed....while my own Sparkle is stuck at 155 lol. (saving exp for a big session in the future) Just commenting for the algo ^^
Doesn’t change the fact that it’s possible that the CN bros wasted their time recording “weighted” substats without considering pseudo rng through the hoyo servers.
Today I made my best piece ever in the game ( 37% crit value overall on set for Acheron... my rolls have been very terrible since the beginning) thanks to this... also based on data I am on "Shitty days or Special UID days" since I get the worst case scenario, but It's possible to manipulate that to my advantage
Im hitting a 16% CR Orb today bois. Thank you bradda. Testing around with some purple relic > realize my today pattern is trash pattern > instantly abuse it. hitting 44 nonstop
this at least minimizes the RNG for substat rolls and reduces the overall RNG to just getting the relics themselves, although I've been seeing people recording their data on the farming stage as well
I guess it would it be better to wait a few weeks to allow data gathering to complete? I jumped the gun a bit and managed to get "42" twice in a row which wasn't even on the spreadsheet lol (Thursday - EU am).
Been waiting. Fortunately or unfortunately I haven't gotten any double crit Atk% orbs and ropes for Acheron so now I have this to help me not get fucked as much. Thanks
If HSR was the Matrix, you would be the Oracle. She knows what she knows but it won’t make sense to many others 😂 Clearly, you nailed the system. Now we need to understand how your art works.
I'll come back near end of 2.1 for this, I currently don't have that many relics to upgrade cuz currently farming char mats. Planning to farm as much relics without upgrading then and try out your methods boss o7
i accidentally use the pattern 2 the other day and score an upgrade for my hand piece LOL i didn't know that wednesday has a high probability to hit the 4th line. i got like 3 cdmg roll on the 2nd line. looking forward to the next piece, cheers bernie 🍹
Yesterday i finished upgrading my acheron traces. I think i'm gonna wait for the double drops first and then start upgrading💪💪 I'll come back to this video at the end of the month😂
Well it’s not 100% guaranteed to work. We would have to be looking at hoyos servers directly for it to be 100% which we don’t have access to. But this is better than going full rng and praying
One thing im confused on that i might be doing wrong So I did the string and got up to, "212" so far as an example and the next possible ones were "2124" and "2121" with a 50% chance on each but I ended up getting "2122"
Data isn’t finished yet. I’m guessing you are trying the Friday data. There’s barely any tests right now cause we only started recording April 1st for this patch.
anyone who thinks this is some other bullshit clearly does not take their time to read the data and use their braincells to understand HOW IT ACTUALLY WORKS, reason why some people are hating on this is bc they're all expecting it to work 100% of the time when clearly the WHOLE PURPOSE of this video is to show people how to narrow down the RNG aspect of rolling for substats. Data never lies, keep it up man! Keep proving them wrong ;)
I tried this and surprisingly some parts were just as i predicted but some weren't, the stats somehow just goes off chart on how i expected, maybe because im understanding this incorrectly or just how bad my luck is lol, either way its a fun thing to try honestly
to be honest I really want to do this, but English is not my 1st language so I have a hard time comprehending everything. I'm so glad you put a lot of explanation on the doc and such. Hope you make more video showcasing you make a bunch of great relic, so I can understand it better. You're awesome.
ive been farming for argenti since his release and have only 3 good pieces, then when the new dot/follow up cavern came i grinded that daily and my kafka, bs, himeko and herta arnt even built yet.... ive been having absolutely dogwater rolls for my relics so ive decided manipulation is the only way, even if its just "this has a higher chance to happen than that" its gonna boost my relic game much more compared to whats going on rn
Just found this lol, I just came back and spent 2000 stamina in farming and upgrading relics without knowing this, now Im waiting on you to finish data gathering and then i will start upgrading haha.
@@BigBoiPnoy Based on the video, sadly I just missed the best day which is Wednesday. I guess I'm going to try this again next Wednesday :D Thank you so much for making this, Brother!
Wait so if I understood correctly. The way this works is 1. We assume that the server predetermines the number or the order of which substat gets upgraded. and 2. That there's a finite number of combinations of the order that substats get upgraded. So it's just bound to repeat somewhere and sometime? So the idea is to guess or figure out where in the predetermined order we are? How did you figure out or guess that that's how the server works? I'm guessing like server time is just the most common variable that games use to create good RNG but surely there's more than just that? Or is this data collection like a way to zk-proof the whole Artifact RNG system?
Correct. And yes there is more to it. But a casual player will not dive that far. I did read you other comment too on the other video. your answer will be found actually on another persons comment on this same video. All rng on gacha games if programed this way cause if it were client side, people can easily hack it. So "pseudo RNG" has to be server side.
@@BigBoiPnoy ahh so it's actually like Zero Knowledge based? We actually don't know what the entropy elements are or why, we just know it must be a server-side calculation and time-sensitive... Imma be honest, it's like a miracle that it does work but also beautiful in a math way?
@@Operation-Overlord correct my man. That’s why most players don’t question it and just try it. It’s either this or going “full rng”. And honestly doesn’t hurt to try cause you don’t really lose anything when trying this method out. The only ones that lose their mind are players that seem to have a masters degree in statistics. Yes I get it the data is low, but how about you try to convince other players to burn materials for this theory. Same time players are greedy and want instant gratification. With the current team I have now there’s once again only so much we can do given our own limited resources as well.
Mhmm. Same time, don’t forget relic set bonus and e1 passive. My Acheron really has 84% crit rate. If I use fuxuan with her it’s 94% crit rate. I’m literally chillin 🤣
It's crazy how lucky you are to get relics with such good stats bro 🤣 I've been using the method since last patch, but the hardest part is finding a good relic to use it on.
Wouldn't it be better to design an app that maps permutations and builds the data model from the permutation map? That way anyone can feed data into the API and you can get bigger sample sizes more easily? Using an excel spreadsheet to do all that manually is going to be an exercise in futility.
Yeah that would be nice. Don’t got time to code one on my end. Maybe someone in my discord can do it. Honestly barely have time to do a lot of things on my end. All I do is show the results of whatever my own discord records. I put it into practice. And stream/record when I can do it. If I have the time to.
Hey, great video, now I really understand how it works. One question: Is the fourth line safe to trust? I know the sample size is lower, but it's matching the data for me. On the other hand, my first two upgrades almost never respect the data, is it an UID problem, or just bad luck? I did two strings in a row that went 412, but 412 has the lowest chance to occur (16.2%) in the data today.
@@BigBoiPnoy Thanks, I just got a GREAT lightning orb for my kafka (9 sps, 3 rolls). 4222 was a pattern that was occurying frequently for me, so I took advantage of it (spd was the 2nd line), after 422, the next one was always 2
Great Video. But that doesnt prevent me from getting shitty Artefacs with useless Substats 😂😂 really great to have a Def% Izumo Orb with Atk%, Speed,Crit% and crit dmg -.-
Hey I just wanted to know if the data is still being updated for 2.2 currently? I dont see any data or any way to change weekday and the percentages that you are seeing on the spreadsheet?
Yep we are working on that. the most up to date sheet is the current one. we are editing it so it grabs data off a different data sheet. Keeping all the main data on the main sheet is making the main sheet lag for everyone else trying to use it. I suggest making a new copy of the current sheet. Cause the current one is the one that lets copies get the current data.
Tried it out myself and jaw dropped when it went pretty much how the string went. albet for the last upgrade where it went to the 2nd choice but still, insane. Props to you and the rest of the folks out. Honestly don't know if I will but just in case, is there a way to contribute data as well or is that your team? Thanks!
first question, does the day affect by where am i currently or where my accoutn region is? secondly, i copied the sheet but the data on thursday is not showed (wednesday or tues.. is normal)
Disclaimer: This is NOT a guaranteed way to have good upgrades on your relics. This is simply pattern reading. You do NOT have to do this if you think it's too complicated. But this is BETTER than nothing, than just hoping for blind luck on your upgrades.
If this works for you, then great. If you feel this isn't worth the trouble, then all the power to ya. Have fun in your own way without ruining other people's fun. Stay strong, fellow trailblazers 💪
Small Note: As the video say in 15:50, the data is still small and new. If you are a casual who wants more precise outcome, do this later in 2.1 when more data/patterns has been gathered.
The more data you have the more your realize that the difference in getting stats is rng. They have a very small sample size, if they had a sample size of 100000 people trying this then then number will get closer to 25%. The numbers are skewed because of the small sample size.
@@gabiblack well yeah he needs a bigger sample size for sure. thats basic statistics.
@@gabiblackwe want this honestly…. Hard to convince people to burn materials for data. Most players are stingy or are too lazy to even contribute to what we do. My own discord has around 3000 players. Only around 100 of them contribute to the data. Let that sink in. Now if everyone in the discord contributed to the data it would be a different story. 3000 players doing one pattern string a day would easily get what ridiculous number you want for your statical needs towards the end of the patch. There’s only so much we, who contribute data, can do especially with how our own resources are.
@@BigBoiPnoyQuestion, why does data need to be collected again after every version patch? If we're thinking that any change in the code can affect the patterns, wouldn't it mean we need to restart data collection every time there's a hot fix (which is almost daily)?
@@rinoakirova1548it was honestly what we noticed from our own testing. Our thought was since the server goes down, most likely the pseudo “rng” gets re-randomized when the servers boot back up. A hotfix won’t be what will change that rng system. Unless they directly aim for that part of the game.
Now… if the whole game server went down during the middle of a patch…. Then ya data gonna have to reset too…
In any case you’ll notice if you go through each patch on the data sheet. No day (let’s say Wednesday as example) is similar to any other Wednesday of a different patch.
my friends: hes probably thirsting over hot anime girls png in a gacha game
Me: learning about relic substat manipulation
Lols it is what it is
For the people who are asking why the data is incomplete, the data recording started Monday April 1st, just wait a little
Thanks man
@@BigBoiPnoytbh tho, the data has been somewhat consistent every patch, with each line being almost the exact same each patch. The only big difference I’ve seen is the Rng per UID, which makes personal data skew either top-low, middle low or high and low. And From those assortments, go from there.
Say example, u start with a 4xxxx, and like we talked before, depending on your server, you could get more likely odds of a certain roll. But I’ve noticed that it also tends to cycle more often the. Not.
In 10 relic manipulations, 1/2 of my rolls went to 42xxx, 1/5 44, 1/5 in 41, and 1/10 in 43, which was on a Saturday, when 44 & 43 are high, with most lines from there, tending to be closer to middle-rangers then higher or lower, being 4243x or something like 42441, which would be similar, or almost exact repeats every couple of cycles, or the next cycle after that.
2 of the 42 lines were back to back, with the other 3 being somewhat eacenly spaced apart, happening on the 2,4, and 7 relic rolls respectfully.
Idk if this would help, but the prospect that depending on UID, your lines could also cycle on old lines is very interesting, b/c if this is true, you can use both the data known, and your own experience with your UID to determine the most likely line upgrade, especially on a specific day.
The only saw these cycles was b/c I had enough relic fragments to do it, (700) and In turn, allowed me to build a a 220 BE Ruan Mei, which isn’t even at maxed relic lvl.
I wonder what else we could be missing in the relic manipulations?
Either way, keep up the good work!
Even though an RNG factor does still exist, knowing that we can have a bit of control on it is reassuring and keep me from the anxiety of rolling upgrades. Hope you keep getting more data sample on the docs!
Thanks. Ya we’ll keep going
Been using this since 2.0 and can confirm that i'v already made alot of godly relics. It didnt work 100% of the time, but for sure targetted some insane pieces and hit the "expected paterns". It works, and im a big fan, even at the point where I'm SCARED to upgrade w/o this spreadsheet info. I just hord my stuff untill I can do a few upgrades but ONLY with these chances in mind.
I fucking hate relics, relic farming and the fact you're characters endgame power comes almost exclusively from it..
This helps ALOT! It cannot be underestimated, people are fools if they still think this is fake or "a myth" just roll and be lucky..
Thank you, brother!
its like eating good food, once you've had a taste, you can't get enough. you want more.
Hii, do you mind explaining the method in an easy way ?
@@awcryonori i don't know how else to dumb it down even more....
@@BigBoiPnoy no problem,, watched it again and got it !
That’s what I’m sayin. I’ve been playing for a year and the relic grind is such a turnoff for me. But I like a lot of other stuff about the game. Being held back by relic stats is so discouraging.
I think it is perfectly natural and understandable for this to create a lot of skepticism to those who find this guide for the first time, such as myself. Though the genuinely overwhelming amount of positivity you receive, very much intrigues me. I always say having good relics is far more favourable than having the best characters. So a method like this working for the majority of people gets the heart pumping.
I gotta admit that i tried an upgrading session using this method and most of my rolls still ended up completely random. I get that its too early and there's insufficient data currently, for as i was consulting the spreadsheet, the % chances were almost close to one another so any of em could've hit really. Or maybe i got the bad uid day, which would really suck to happen to someone trying this for the VERY first time.
Ive saved up some good crit pieces and perhaps i will try this again towards the later half of the patch, as you suggest. I truly wish to become part of the crowd that cheers on how successful this method is, and praise you for finding such an unbelievably invaluable gem through sheer hard work and observation.
Thanks for trying it. And ya most people wait till the end of the patch. It doesn’t hurt to try like overall. We’ll continue to work on the data.
Having a background in coding did help me in trusting all of this information since you can't truly randomly generate something with scripts, so although not 100% reliable, it still follows a pattern.
@@shui804 yep
Petition to call relic manipulation as "Elio Scripting" 🙏🙏
Lols
Yes pls so it doesn't get flagged from hoyo😢 I'm scared they might discover this trick
Yes please
@@zerefuuutvall they would do is make it actually random, and then it would just prove the pattern-recognizers correct because these methods would no longer work
So new patch new pattern? @@Ooweeeooo
Bernard, you absolute Chad.
I started HSR late in 1.4 for the PS5 release, and just today, thanks to your guide with some juicy substat roll, i finally fully 36/36 star MoC for the first time!
Following your guide, 9/10 next substat patterns were all correct, we take those all day, the extra Crit Dmg rolls put my Acheron over the top!
Thank you so much man 😭😊
Thanks man. We are happy as always the data is helping another player
As someone who has been using this since last patch, I just wanted to say that I think this method is more favourable to farm a large number of relics, pick the ones you want to upgrade, then use this method to upgrade, instead of upgrading relics one at a time.
Yep this is a great mindset. This is how I do it.
Got motivated to roll when I watched this, literally got 5 rolls into a Crit Damage Body piece, with full Crit Rate 21.3 when I forced it. tysm!
thank you for making videos like this explaining relic manipulation, even showing so much evidence with testing. i’ve always had the thought of substat rolls being predetermined and came across your channel and everything just makes sense, thank you!
You’re welcome man. It’s what we paid attention to when upgrading.
4:41 the reason you have to be connected to hoyverses servers is because they want the rng to be server side not client side. If it was client side you could disconnect from hoyverses servers and when you reconnected you could tell the servers that you got a specific roll or something. This also extends to the pulling system. All modern games do this so its harder to hack. An example of a game that is client side is cookie clicker. you can just inspect element to make the game think you made a bunch of cookies.
Overall I think you have the right idea. Nothing is truly random and every outcome is predetermined. However this proudly wont be able to predict your relic rolls as while yes adventualy you would be able to find a pattern I think hoyverses rng system is designed in a way so that it would take hundreds of years before you could find a pattern.
I think whats going on is selection bias where your brain remembers times when this a roll fits your hypothesis and disregardes evidence that contradicts your theory. This also applies to people in the comments. Our brains likes to only remember the highest peak and the lowest rolls. You wouldnt write a comment about failing an artifact right? For me to be able to believe this I would need to do my own testing. As a uneducated observer I think its best just to ignore this.
I have tried this before and my results are mixed
That’s totally fine. Gamers should have fun how they want to. But ya it’s all server side pseudo RNG. And for many players, they prefer having this over going full RNG. It’s like at least they know the possible outcomes. I already said this isn’t a guaranteed way also. Mainly cause we aren’t able to fully see what numbers are actually on the server, that’s straight up impossible. But overall, yes I do have some failed pieces. At the same time I have a lot less failed pieces than what I used to get when going full RNG in the past during the beginning of HSR as well as the genshin impact days before I quit that game.
yeah man..
I agree with you. I kept an eye on the data they've collected for last patch and the numbers they are getting are not statistically significant enough to prove it is anything but random, even disregarding the likely selection bias like you mentioned. They also attempt to write off the times the rolls do not line up with their determined patten by labeling it "special uid" which further increases the bias.
@@BigBoiPnoy this is like the placebo of relic rolling lol.
Maybe. Honestly we can be wrong about all of this, and it can be totally something different from what we think. The goal overall was to help players roll better. And honestly if we have achieved that we are happy.
In the whole factor of the data gathering the sad thing is… a lot of players don’t want to contribute data. My own discord server has around 3000 members out of that 3000 only 100 are contributing data…. It’s a sad reality of players just using what we gather to their advantage. Same time can’t blame them, we burn materials to get data and that’s the only way… we want data to be 1000+ each day. But if only a little amount are contributing, that goal won’t be achieved.
As soon as I finished watching this video, I went to upgrade a piece which was atk% with CritDMG/CritRate, speed, break effect. I forced it to 2 then got another 2, and then again and then crit damage. Basically 3 hits on CritRate 18.1%, 2 rolls on CritDmg 19.4%. Thank you for your blessing!
LOL i’m just showing what’s possible bros.
That's crazy. What string did it land on?
@@Kloriaue I started at 2 first, then it went 2,1,1,2,1
@@cIIduru can you post a picture? 55 crit value sounds absolutely disgusting wtf
@@---ie2msyeah because 55 cv is impossible? If you start with a 4 substat relic that has crr at 3.2 and cdmg at 6.4 you can get 5 more substat rolls which if it rolled to the max possible value would only amount to 32 cv(from the 5 substat roll) and 12.8 cv(from starting value) which gives a max of 44.8 cv per relic, and this is the absolute best case scenario, assuming op isn't just flat out lying I'd assume they miswrote the crr value since 3 rolls on it could only be 3.2 x 4 at max possible value which lines up to be 12.8 crr and 3 roll for cdmg would be 6.4 x 3 which would be 19.2 at max possible value, i don't remember if there were any rounding error but it's possible that the cdmg stat is possible while the crr isn't, and surprisingly if you do count the substats that i had mentioned it would be (12.8 x 2) + 19.2 = 44.8 cv as previously mentioned
All thats left is to let the data cook for a bit.
Yep
Dude I’m so pissed I just found you’re channel today, I’ve deff been missing out but the data and proof don’t lie. You’ve earned a sub from me my man keep it up
Thanks
Just now watching this.. had no idea this existed ty for the info.. I understand exactly how to do it but no computer I will keep in my head but I understand the formula u are using.. much love for this video❤!! It's wild that people don't understand that is the best video I have ever watched of hsr!! This knowledge is gamebreaking
We are around just chillin getting top builds lols
Bernard doing the god’s work again. This is dissertation lvl of data science. You & your team has contributed so much to the HSR community. May life bless you with luck.
Thanks man
I can confirm that I used this doc Saturday and rolled 3 times on crit rate on a crit damage chest piece.
Honestly lowering the RNG to a 50/50 has been a life saver. Thank you so much.
Yep. It’s better than overall rng. Most players would agree.
each video of you is better and more intuitive, tysm bernard and all team
Relic manipulation is one of the main reasons why my love for HSR sparked again. Rolling for good relics is a pain already but knowing there's a better way to tackle the most frustrating mechanic of the game keeps me going. Thanks for the vids.
Thanks man. My videos is also directed for players like you. I’ve heard all the story’s. Including my own best friend. “I quit cause I couldn’t find good relics for my characters”
Didn't have any relics left to try it on for the current characters, so I tried to prepare something for Boothill and it went really well. Like 25% BE+cr+cd boots and other few pieces with 0 rolls into bad stats. Could've been just luck, but it definitely didn't hurt to try this method.
I'll definitely apply this to future chars, but probably won't refarm for my already ok-ish equipped characters.
Tried the manipulation, and I often got an upgrade on a single crit relic at least 2 times. Meanwhile, going full RNG on double crit relic just ended up as a perfect dodge on crit. Thanks for the guide.
Glad it’s helping all I can say
Left with more questions than answers:
1: straight pull up a note pad. What is this? What we use that for?
2: excel are full of jargons and disclaimers without coherent description of what to do as a newbie from beginning to end. Can you please do a 1 to 10 sequential step etc of what to do? Example:
Step 1: upgrade a fodder?
Step 2: then go to excel column xx field xx do what?
Step 3: xx
Step 4: xx
Step 5: upgrade the ssr gear
So the notepad is just an example of how the pseudo RNG works on the hoyo server.
the fodder i upgraded... if you were paying attention to what i was saying. was to force the "upgrade line" to the 2nd line out of the 4 lines( the blue relic) most people will have relics with 3 line purple and gold relics which when upgraded forces your upgrade line to the 4th line out of the 4 substat lines.
the 2xxx in my actual example is how i start my pattern cause i upgraded a 1line relic( a relic that has 1 substat) to force the pattern to start as 2xxx. Then using the data recorded by other players from other servers( mind you i was using EU and ASIA data only cause america didn't have data yet for this video so this proves region does not matter) we have and i predicted my upgrades. the xxx is the unknown numbers hence why i type it as 22xx, 22xx, 23xx, 23xx, etc.
Each time you ADD as substat your pattern resets. thats why i kept going back and leveling a relic with 1substat. its to force the "upgrade line" back to the 2nd substat line.
switching relics doesn't change your pattern string from the server. hence why i'm always switching relics.
There's documents made by other players on the first page of the sheet to read as a guide for dummies. you have to read though...
I saw ur last vid on 2.0 and i was so suprised that like 90% of the rolls were correct thats how i got my first 160 speed character and alot more. Thanks so fucking much cuz u legit made my acc so much better i will never look at relics the same cuz like i said 90% was correct for me personally genuenly.
Another banger Video, still waiting for the Data at the end of the patch... Planning to hold all of my relics until the and of the patch (especially we have 3x planar and relic drops in this patch)
I'm so glad UA-cam recommended your video to me! I just CAN'T WAIT to put this into action too, but for now, I'm only in the pre-farming stage lol
Thank you and everyone on the Discord team so much, you're doing godly work for us 🥰
This works! Thanks bro. You saved me a bit from my rng hell. 😂 Time for me to wait for good days. Now I can save those good relics and upgrade slowly through weeks. Big plan ahead. Woohoo!!
Yeeeee 😎
In some way on of your vids go recommended to me 2 or 3 days ago, i started watching it with a lot of prejudice; ended up watching like 5 or 6 more vids about it, so yesterday i tested it on 10 crappy relics, playing it safe and what not, followed your explanations/examples and..... wtf dude every single sub stat that the spreed sheet told me at it was going to hit, dude it did, mam if only i new about it from the start.
Hey props to you and all the ppl that contributed to this project/spreed sheet.
Just hope Hoyo doesn't see this and change it all up.
Only time will tell. Same time the code automatically gets shaken up every patch anyway
man i remember genshin having groups of substat readers too. we just clutch at whatever might be better than pure randomness lol. hope you guys succeed at this
i've been told that theres a group on bilibili that does something similar to what we do but for genshin. And they've been doing it for awhile on that side. Essentially meaning... CN did it first.
Thank you for your hard work at making this happen! I will try it out myself and see if I can get some good stuff.
I hate reading but thanks to Flame14 for that guide, really made it easy to understand
Yep he did a great job.
Aw man, seems like I'm on a Special UID day the one chance I had to try this out before the update. :(
Appreciate the effort that went into recording all this information despite all that!
Hmm? It’s not the end of the patch yet?
@@BigBoiPnoy I was assuming patches roll out with new banners like Adventurine coming up? If it doesn't I'll give it another shot later this week since my Monday RNG was being special and Tuesday didn't look too hot
@@lightsageala till end of the patch. Adventurine is not a new patch. They just switch the banners
@@BigBoiPnoy Oh, good!
Looks like I'll give it another shot on Thursday then, thanks for letting me know!
So I decided to give this a go to test this out....
Seems like I got what you guys call "Shit Days" and kept hitting the lower percentages more often instead.
I don't know, the fact that such "exceptions" even exists for such a thing makes me question how reliable this rng manipulation trick is. I can believe that there is a form of "patterned rng" happening with the relics though. But I'm going to mainly keep tabs on the data to see how all of this goes rather than rely on such methods atm since I'm not convince that this is reliable as of now.
Hey man at least you tried it.
But ya. Most people stop upgrading from there when their patterns are the low %s. Other players in my own discord adapt and just purposely aim for the low %s and it works out for them
I basically told players to do test strings on purple relics that have 3 substats already before they touch their own good relics if they don’t want to mess those up.
As someone who is trying to do this method I want to leave a cool hint I learned: Get as many relics as you can and store them to do all this rng when the day is favorable.
My greatest problem right now is the lack of relics to either use them as fodder or to upgrade the right rows.
For example, I had a great C.rate relic and the next line would be either 2 or 4 (higher %). They where useless stats on that relic since 1 and 3 😂where the good lines. And I didnt have any other relic with good stats for 2 and 4. Guess what? It rolled a 4. So I wasted a 29% roll.
Hoard relics. Even those who looked trash at the start. (I got a 1 roll in C rate and 3 in C dmg in a relic that started with flat hp and flat atk)
yep hoard your relics. use them data is peek
First time doing this and I was confused ngl but I went slowly and as you did... Good god it works. It's like a drug that I can't get enough of. Only thing stopping me is getting actual good relics though lmaoooo
The shit we all deal with. Relic farming lols
Ty for your work and the work of your team as well. It's too useful.
I'm waiting and saving all my good relic's and saving it until the last week of the patch.
Another fire video brother.🔥
thanks man
Thanks, I helped a friend manipulate a few weeks ago and now they have 4 characters over 161 speed....while my own Sparkle is stuck at 155 lol. (saving exp for a big session in the future) Just commenting for the algo ^^
You’re welcome. And ya as weeks go by data just gets more accurate cause of the data we record. It’s an over time thing, more data the better.
Yess finally! I was watching your old videos last night and was waiting for the latest update. The joy I felt when I saw this first on my feed haha
Was skeptical, but tbh after some few vids, the data doesn't lie, will try and hop in to this, nothing to lose either way.
BRO YOU'RE A LEGEND!!!! This method really help me hit all THREE substat for acheron!!!
8:34 weighted stats means its just favoring that particular stat, doesnt mean its impossible, just improbable
Doesn’t change the fact that it’s possible that the CN bros wasted their time recording “weighted” substats without considering pseudo rng through the hoyo servers.
Today I made my best piece ever in the game ( 37% crit value overall on set for Acheron... my rolls have been very terrible since the beginning) thanks to this... also based on data I am on "Shitty days or Special UID days" since I get the worst case scenario, but It's possible to manipulate that to my advantage
Correct. Very nice my man.
Im hitting a 16% CR Orb today bois. Thank you bradda. Testing around with some purple relic > realize my today pattern is trash pattern > instantly abuse it. hitting 44 nonstop
This is why I was testing my own strings before my own upgrades on this video. Great work man
well i deff have to keep following this and possibly try it myself because i have 0 relic luck
this at least minimizes the RNG for substat rolls and reduces the overall RNG to just getting the relics themselves, although I've been seeing people recording their data on the farming stage as well
We’re recording that stuff also. Should check my next video by what I mean. We call it clear time theory.
@@BigBoiPnoy looking forward to that, thanks for everything you guys do
Lols the videos already out 😉 it’s under a different video name though.
Been waiting for this one. You are lifesaver! Thank you!
Hope we are helping you guys lols
I got this video at 1am and I haven't sleep. Look like I wont get any sleep today
bro go sleep. the game will be there when you wake up
@@BigBoiPnoy I tried several times and realized I have bad UID day. Then I go to sleep after knowing that fact 😂👆
I guess it would it be better to wait a few weeks to allow data gathering to complete? I jumped the gun a bit and managed to get "42" twice in a row which wasn't even on the spreadsheet lol (Thursday - EU am).
Yep just wait. We get more data over time. That’s how it is every patch.
Been waiting. Fortunately or unfortunately I haven't gotten any double crit Atk% orbs and ropes for Acheron so now I have this to help me not get fucked as much. Thanks
Anytime
If HSR was the Matrix, you would be the Oracle. She knows what she knows but it won’t make sense to many others 😂
Clearly, you nailed the system. Now we need to understand how your art works.
It’s not so much art. I’m just a dude that records data that no one else would notice. I consider this stuff fun though 😉
you sir, earned everyone's subscription, not i can go 5k-6k attack on kafka
Finally the one with bigger audio volume
🤣🤣🤣🤣
finally
I'll come back near end of 2.1 for this, I currently don't have that many relics to upgrade cuz currently farming char mats. Planning to farm as much relics without upgrading then and try out your methods boss o7
It’s something that grows over the duration of a patch anyway. Meaning data is peek more the last weeks of any patch
welp time to farm relics again, i wasted my relics with blind luck for acheron
It is what it is
i accidentally use the pattern 2 the other day and score an upgrade for my hand piece LOL i didn't know that wednesday has a high probability to hit the 4th line. i got like 3 cdmg roll on the 2nd line. looking forward to the next piece, cheers bernie 🍹
You're probably having a special uid day. I tested the data out earlier with 7 relics and hit 42 5 times and 41 twice
Yep this can happen
I see i just read the left column, good to know
Yep yep
oh i didnt knew that the pattern resets every time you get a substat! anyways insanely helpful tysm
yep your string resets each time you add a substat to any relic that has 3 or less substats
tried it. not 100% but was able to get a 2/2 cr cd piece for acheron despite my trash relic luck
Nice
Aint even know this till it was mention in an fb group lmao 😂😂 man this shits wild... you earned a sub!!
wait what my stuff is being shared on there??? many things i dont know is going on lol... also thanks man and welcome
Yesterday i finished upgrading my acheron traces. I think i'm gonna wait for the double drops first and then start upgrading💪💪
I'll come back to this video at the end of the month😂
Sounds good bro
wow this actually works never had an 35 cv piece befor thank u so much
OMG this is awesome just tried it for the first time and it worked perfectly, why didnt i knew about this sooner, still cant believe it works.
Well it’s not 100% guaranteed to work. We would have to be looking at hoyos servers directly for it to be 100% which we don’t have access to. But this is better than going full rng and praying
@@BigBoiPnoy anyway it still got me best relics in last weeks so its still a win.
One thing im confused on that i might be doing wrong
So I did the string and got up to, "212" so far as an example and the next possible ones were "2124" and "2121" with a 50% chance on each but I ended up getting "2122"
Data isn’t finished yet. I’m guessing you are trying the Friday data. There’s barely any tests right now cause we only started recording April 1st for this patch.
Bernard drops a video, I watch. 🤝
Oooo thank you 🙏
same
We are so back
anyone who thinks this is some other bullshit clearly does not take their time to read the data and use their braincells to understand HOW IT ACTUALLY WORKS, reason why some people are hating on this is bc they're all expecting it to work 100% of the time when clearly the WHOLE PURPOSE of this video is to show people how to narrow down the RNG aspect of rolling for substats. Data never lies, keep it up man! Keep proving them wrong ;)
Thanks. Much love man.
4k subs lets goooo, keep up your great work king 👑
Thanks for all your support bro. Always helps. This one might be the last one for now. We’ll see though 🤣 gotta go more dad mode
I tried this and surprisingly some parts were just as i predicted but some weren't, the stats somehow just goes off chart on how i expected, maybe because im understanding this incorrectly or just how bad my luck is lol, either way its a fun thing to try honestly
We just started getting data this past week. Expect it to be inaccurate for now.
WOOOHOO I'VE BEEN WAITING FOR THIS ONE!! GOAT CONTENT CREATOR JUST DROPPED
But I’m not a content creator. I’m just a bored dad at home with his baby daughter 🤣.
I just started the vid. Can confirm this vid a certified banger!
-This comment has been sent to the Ministry of Truth for further investigation
LOL
to be honest I really want to do this, but English is not my 1st language so I have a hard time comprehending everything. I'm so glad you put a lot of explanation on the doc and such. Hope you make more video showcasing you make a bunch of great relic, so I can understand it better. You're awesome.
I’m trying my best 🫡
That moment where you watch it 6 times to understand coding is a beast
Dude Thank You! Ur a Lifesaver
I was waiting for this one, thanks
No cap I actually tried this and I see the potential in it
ive been farming for argenti since his release and have only 3 good pieces, then when the new dot/follow up cavern came i grinded that daily and my kafka, bs, himeko and herta arnt even built yet.... ive been having absolutely dogwater rolls for my relics so ive decided manipulation is the only way, even if its just "this has a higher chance to happen than that" its gonna boost my relic game much more compared to whats going on rn
Yep. That’s what most players say on my own discord. This is better than going in blindly.
The amount of work this takes is crazy, I appreciate the dedication to help others get good relics.
We try our best to
Just found this lol, I just came back and spent 2000 stamina in farming and upgrading relics without knowing this, now Im waiting on you to finish data gathering and then i will start upgrading haha.
Most people wait till the last weeks of a patch. Cause that’s when data is more accurate
even with the incomplete data I still managed to get cdmg 4 times with this doc
Yep it can happen. Sorry for the low data as yes we just started
interesting video, will definitely try this after gathering more relics to use, thanks!!
Gives us time to get more data
the guide i have been waiting for
but before this i need to farm my traces
Yep. We all do.
bro just came and made the best acheron
Yep. Well one of em.
Kinda late, but I'm glad that I can be finally here~
Never too late man.
@@BigBoiPnoy Based on the video, sadly I just missed the best day which is Wednesday. I guess I'm going to try this again next Wednesday :D
Thank you so much for making this, Brother!
@@zidstarz ya lols... also show it on my new video tooo. the faster speedrun. gotta love rolling on wednesdays
@@BigBoiPnoy I'll be sure to check it out, thanks for the reminder!!
THIS CHANGES EVERYTHING
I tested this and it actually works lmao ty
maybe my luck just isn't lining up with the odds but I keep re-rolling 4 to start and I'm not getting any kind of consistent second pull
Hmmm have you checked the time stamps page? I just added it. Might help you lessin the mixed results
Wait so if I understood correctly. The way this works is
1. We assume that the server predetermines the number or the order of which substat gets upgraded.
and
2. That there's a finite number of combinations of the order that substats get upgraded. So it's just bound to repeat somewhere and sometime?
So the idea is to guess or figure out where in the predetermined order we are? How did you figure out or guess that that's how the server works? I'm guessing like server time is just the most common variable that games use to create good RNG but surely there's more than just that?
Or is this data collection like a way to zk-proof the whole Artifact RNG system?
Correct. And yes there is more to it. But a casual player will not dive that far.
I did read you other comment too on the other video. your answer will be found actually on another persons comment on this same video.
All rng on gacha games if programed this way cause if it were client side, people can easily hack it.
So "pseudo RNG" has to be server side.
@@BigBoiPnoy ahh so it's actually like Zero Knowledge based?
We actually don't know what the entropy elements are or why, we just know it must be a server-side calculation and time-sensitive...
Imma be honest, it's like a miracle that it does work but also beautiful in a math way?
@@Operation-Overlord correct my man. That’s why most players don’t question it and just try it. It’s either this or going “full rng”. And honestly doesn’t hurt to try cause you don’t really lose anything when trying this method out.
The only ones that lose their mind are players that seem to have a masters degree in statistics. Yes I get it the data is low, but how about you try to convince other players to burn materials for this theory. Same time players are greedy and want instant gratification. With the current team I have now there’s once again only so much we can do given our own limited resources as well.
Nice work. 1:5 cr to cd is not ez. With this, can change everything.
Mhmm. Same time, don’t forget relic set bonus and e1 passive. My Acheron really has 84% crit rate. If I use fuxuan with her it’s 94% crit rate. I’m literally chillin 🤣
Now the first step is to get the double crit substat relic drop first right? Which is more harder than you think
Very hard… like bro… drops suck in this game
It's crazy how lucky you are to get relics with such good stats bro 🤣 I've been using the method since last patch, but the hardest part is finding a good relic to use it on.
step 1: find good relics
step 2: use method when data is peek to not waste good god substat relics
@@BigBoiPnoywhat I meant is that the problem is get the relics to the step 1 🤣
the main issue we all fall under
Wouldn't it be better to design an app that maps permutations and builds the data model from the permutation map? That way anyone can feed data into the API and you can get bigger sample sizes more easily? Using an excel spreadsheet to do all that manually is going to be an exercise in futility.
Yeah that would be nice. Don’t got time to code one on my end. Maybe someone in my discord can do it. Honestly barely have time to do a lot of things on my end. All I do is show the results of whatever my own discord records. I put it into practice. And stream/record when I can do it. If I have the time to.
Man honestly I love your work man and appreciate what you're doing
I just wish my dumbass could understand it
Maybe one day
You don’t have to bro. Just enjoy the game how you want to.
That was my mentality i promise you its actually easy they have added the translate pattern for us
Hey, great video, now I really understand how it works. One question: Is the fourth line safe to trust? I know the sample size is lower, but it's matching the data for me. On the other hand, my first two upgrades almost never respect the data, is it an UID problem, or just bad luck? I did two strings in a row that went 412, but 412 has the lowest chance to occur (16.2%) in the data today.
Remember it’s a pattern string. So over time the 3rd and 4th are actually more reliable
@@BigBoiPnoy Thanks, I just got a GREAT lightning orb for my kafka (9 sps, 3 rolls). 4222 was a pattern that was occurying frequently for me, so I took advantage of it (spd was the 2nd line), after 422, the next one was always 2
This was the other goal. Get to the point other players can predict without our help. 😉
I'm more causal but I want better relics so I'll give this a try. I suck at math and i don't get all of this but i think i get the gist of it.
Don’t be afraid to join us on discord
Great Video. But that doesnt prevent me from getting shitty Artefacs with useless Substats 😂😂 really great to have a Def% Izumo Orb with Atk%, Speed,Crit% and crit dmg -.-
I have a couple of those also 😭😭😭😭😭 honestly the 2 relics I have on this video were the lucky ones…..
Why do you swap substats in down left corner after upgrading every blue piece?
Just a bad habit. Does nothing
Now I only need to get the relics in the first place. Great video though.
For getting good relics worth leveling… we are all on the same boat
Hey I just wanted to know if the data is still being updated for 2.2 currently? I dont see any data or any way to change weekday and the percentages that you are seeing on the spreadsheet?
Yep we are working on that. the most up to date sheet is the current one. we are editing it so it grabs data off a different data sheet. Keeping all the main data on the main sheet is making the main sheet lag for everyone else trying to use it. I suggest making a new copy of the current sheet. Cause the current one is the one that lets copies get the current data.
@@BigBoiPnoyhey is it still updating even now? Sorry to brother 😅
@@keylimepie3602 i just updated it. we dont update till 1 week into each patch. We can't instantly give data if there's none.
@@BigBoiPnoy ohhh thanks I see 😁 ❤️
Tried it out myself and jaw dropped when it went pretty much how the string went. albet for the last upgrade where it went to the 2nd choice but still, insane. Props to you and the rest of the folks out. Honestly don't know if I will but just in case, is there a way to contribute data as well or is that your team? Thanks!
Ya man data contribution is done through discord
first question, does the day affect by where am i currently or where my accoutn region is? secondly, i copied the sheet but the data on thursday is not showed (wednesday or tues.. is normal)
Data collection just started I think
We just started collecting data. This video was made same day the data was gained