The first season in awhile that won't open with a 6-craft agenda... Right?
#Season 16
vapid valley
VHs
Highs
Avgs
Granary
G: 12x 
Garnet (VH) >> 12x 
Alyssum (H) >>> 12x 
Spruce (A)
VH: 
H: 
A: 
4hs: VH - 
; H - 
Notes:
• I love that both of Radish's 4h links are VH but Radish itself is only Avg.
• Granary isn't great and our pops are all over the place.
• Expect a LOT of Strong Peaks due to Curated Lists assumption.
vapid valley
G: 12x <:OC_Garnet:1035379581772906496> Garnet (VH) >> 12x <:OC_Alyssum:10353795...
@summer kernel What do we even do with this?
summer kernel
I’m a fan of Jam
But uh
Die mostly
I think
There are like 2 8 hour VHs
It’s all 6s all day baybeee
vapid valley
It's pretty bad but curated tells us something strong will happen a lot
vapid valley
Oop, forgot stuff, I'm dumb
vapid valley
Recommendations for Season 16 (Dec 6 - 12)
@ Roles go here.
• C5: 3738 
(4590 24 Groove)
∟ Earrings (4h) - Butter (4h) - Sheepfluff Rug (6h) - Butter (4h) - Sheepfluff Rug (6h)
▬▬▬▬▬
• C6: 4113 (5151 35 Groove)
∟ Sauerkraut (4h) - Corn Flakes (4h) - Sauerkraut (4h) - Corn Flakes (4h) - Sauerkraut (4h) - Corn Flakes (4h)
||No this isn't a joke, another 6-craft that uses an item 3x times is actually the best rec for C6. That weak Crook really sucked.||
▬▬▬▬▬
• C7: 4443 (5694 35 Groove)
∟ 
Firesand (4h) - 
Growth Formula (8h) - Firesand (4h) - Growth Formula (8h)
||We checked if Potion is better since Firesand will be higher supply tiers and it wasn't. Firesand is still recommended even after C4's agenda.||
•• Season 16's Total: 23,448 (our third worst season to date)
C4 Update: Firesand is in fact a C7S, so unfortunately C4's value did not get the bonus.
✨ The current cycle's agenda and alternative recommendations are in [#🤖︱bot-recs](/guild/1034534280757522442/channel/1046906637614583839/)
Be sure to check your Material Allocation button after setting your agendas! 🦜
vapid valley
Sorry I'm gonna be a few late; wanted to get an Expert done right at reset and the tank DC'd at the first pack :') Currently post-second boss.
mint vine
silly tanks
vapid valley
Just finished, transferred back to Dynamis, trying to login now. RecsBot might beat me to the island
mint vine
well, BC says rest so far lol
vapid valley
Fun
mint vine
waiting to see what bot says but the alts in plugin are...bad lol
vapid valley
I'm gonna pretend everything is Strong peaking
and see if I get something
CG STILL SAYS REST AHAHAHA
mint vine
wow
vapid valley
and RecsBot seems to be shy
mint vine
hmm, bots late
vapid valley
again
@summer kernel RecsBot is late again, not in the dev channel
It's not even recommending things from today
Bed isn't today
mint vine
interesting, that doesnt match mine at all
mint vine
vapid valley
Okay yeah here's what I have now
NOW it's working
and it actually recommends making something
oh geez
if Shark Oil is strong, it's slightly better
This is... the best it can do?
3114?????
vapid valley
I wish to Veto this agenda.
We're now playing Secret Hitler, 5/6 Fascist Policies are down. I'm playing the Veto.
vapid valley
RecsBot seems to have issues with C2
last week it hid, this week, it's just not posting
mint vine
it saw how bad today was and was like...fuck this, I'm going back to bed
vapid valley
At this point, I'm calling for a tentative rest.
while we wait for RecsBot/figure out Shark Oil's peak
Cuz rope becomes better with Oil
mint vine
yeah for sure. I'm doing some craziness to see what the best 6-item craft is but I don't have much hope
vapid valley
Shark-Rope-Tunic-Rope 
mint vine
oof, my sheet wont even give a 6-item in the top 30 options
top 50 either, lol. that doesnt bode well
vapid valley
@summer kernel When you see this, fix #🤖︱bot-recs plzkthx ❤️
summer kernel
When we know about shark oil, tell the bot in #🤖︱bot-recs , and that should get it back running. Tbh, I have no idea what went wrong. It seemed to think the data from TC was invalid, but I ran it locally and it set it fine. I don't get it.
" Peaks were invalid. Rescheduling" dutifully for every 15 minutes since 3:10am
summer kernel
Bot says to rest
summer kernel
Also the issue with the bot has been fixed.
As always, the problem with code is that it does exactly what you tell it to do.
vapid valley
Sorry, overslept
vapid valley
Bot says to rest
Does it?
Oh the tentative rec I see
why is it waiting for shark oil?
laughing hysterically
summer kernel
Please scroll down. XP
vapid valley
TENTATIVE REC, HECK.
Also why rest when that's worth over 3.5k with bonus? Is the week THAT bad?
I can't
I'm so... damn
summer kernel
Silver Ear Cuffs has a lot of words in it
vapid valley
sure does.
summer kernel
Also, I changed it in the future to look like this
So it'll make the bonus more obvious
But no, today the rec is not tentative and it is rest
vapid valley
That's why I'm so confused why it pinged me asking for Shark Oil
summer kernel
Because it was wrong and I fixed it. XP
summer kernel
what about shark rope tunic rope?
vapid valley
I was curious about potentially doing a 5/6 craft and hoping for some C2 bonuses but I wasn't sure w/o the bot, hence why I pushed a tentative rec with Rest
summer kernel
But yeah, I changed some info around where it rests and also fixed a bug where it demanded info when the best rec changed but the rec didn't matter because it was still rest
summer kernel
I fixed the bug
And this fix was way less dramatic than the fix I made last week
Like, last week I basically re-architected the structure of the bot
To ensure I could never make the mistake of leaving the bot in testing mode
And that rearchitect was kinda dangerous and did end up having a bug that happened today
But the fix to the rearchitect is pretty chill
vapid valley
You're gonna have a field day when we switch channels again aren't you? 
vapid valley
I hope people are okay with going from deleted posts to them having an endless archive from the bot
summer kernel
If people complain in sufficient numbers, we can change it
vapid valley
I was thinking to just leave up the season and delete when the next season kicks in so people don't even need to check other channels, but if there's not enough feedback asking for old seasons to be deleted, we can just let the entire archive go through to them
(from whatever season we start full auto)
summer kernel
I mean, I still think the archives in their current format is really, really useful
I wouldn't want to scroll through a whole channel for that
vapid valley
Oh no, I just meant that if people want to, they can go back and see earlier but posts after we go on full automation. I'm still planning on keeping the archive up-to-date.
Like if we start on season 18 with full auto, people will have the archive in their own servers starting on season 18.
Of course, I don't know if people would want that many posts or not.
summer kernel
I mean, we don't get to have fewer posts
We just get to choose if those posts have content in them or not
Our recs channel stays clean but anyone who follows it gets one post per post, no ifs, ands, or buts
mint vine
I mean, I still think the archives in their current format is really, really use...
couldn't the bot just spit them out into Archives in that format though?
summer kernel
It could
bluhhhhhhhh, features
I have already done more work on the bot than I should given all the packing I should be doing
mint vine
i am absolutely not advocating for it happening at the same time manual recs stop! but less manual labor is always good
summer kernel
It's true
vapid valley
It's almost no effort for me
When I stop doing recs completely, even my manual ~12-2p posts? A bit more effort and we can talk then.
mint vine
yeah I meant like by March, when everything else is automated, that should be an easy automation too
vapid valley
Uh... @summer kernel???
TWO Ropes dropped it from Skyrocketing to Increasing???
Is that... accurate?
Rope last season was C4S so it's definitely strong, but would TWO ropes drop it to Increasing? I thought on C1 it sustained more than that.
This just has me concerned for the future when testing stuff out.
summer kernel
....kupo
Wait
Oh, it was strong last time
Okay
So
0 -> -6 is -6, which should definitely be Skyrocketing
I wonder if it's doing something rude and stupid the same way that 2->0 is -2 but also None
But that doesn't makes sense because it only does that to start the day and after you craft something it follows the rules
Theoretically
Instead of a strong peak being -8 and then -7, it could be -7 and then -8? I swear I tested this, but the data is kind of buried
I know the peak is at -15
Because if you make 7 it goes up to Insuff
But if it were at -7 for a strong pre-peak and weak peaked last week, 2 -> -7 is -9, meaning crafting 1-3 would keep it at Skyrocketing but 4 would put it at Increasing
I guess it would be really useful if tomorrow you could craft exactly 7 of something that's strong peaking on C3??
It's entirely possible that I have the peak and pre-peak deltas backwards.....I'd hate it, but it's possible.
I found my data that making 7 of a strong peak pushes it from NE to Insuff
So, again, the strong peak being -15 is not in question
I'm just poring through my data looking for times I crafted something on a strong prepeak
Every time I can find is on C2 and it was a weak peak previously D:
So the ends of weeks where I did my big research were 9/12 and 9/19
Oh here's some data
On a strong peak, nothing crafted before
NE - Skyrocketing
Crafting 1 and 2, it stays Skyrocketing
Crafting 3 it goes to NE - Increasing
3-5 are Increasing
6 is unknown
7 is Insuff - None
So if we're going from -8 to -15, we start at -7, Skyrocketing
Making 2 should put us at -5, Increasing
So yeah
It has to be backwards
It's -7 on the pre-peak and -15 on the peak
Because that'd mean we do -8 on the peak, and -8+2 is -6, still Skyrocketing
Making 6 also should be NE - Increasing in this case though...
Right?
Because if we're at -8 on the day and we add 6, that's -2
Which is Increasing
It also puts us at -9 total, which is still NE. Making 7 should put us at -8, Insuff - None
But no, when we make 6 of something on a strong peak, it says NE - None
We do that all the time
Which means it's within 1 of where it started, which means it has to have started at -8 on the day
help
vapid valley
@summer kernel I guess... we'll know tomorrow?
vapid valley
I was just really confused because it dropped to Increasing
Which means in the past, I may have falsely claimed something as a weak when it was actually strong
I'll make sure it's always an inefficient craft from now on to be safe
summer kernel
Was Rope strong last week??
vapid valley
(I always keep the previous week's sheet until C2)
vapid valley
I wasn't intentionally.
summer kernel
But we weren't waiting on you to give Rope's peak
vapid valley
What if it's a true strong
but unknown
and then I craft it efficiently
and it drops to inc
but it's really C2S?
Unless a known strong is a "weaker" skyrocketing on C1?
(I literally used Rope just because it linked to Shark Oil and it's strong peaking so I was like, yeah why not take advantage?)
summer kernel
That's different though
Because that wouldn't happen for something that was weak last week
Known strong is a weaker skyrocketing
Because it goes down 2 less than it would have if it was a weak peak last week
That's the whole reason weak peaks last and this week show up as Skyrocketing, because they have a bigger difference from where they ended the week and where they are now
Strong peaks last week have a smaller difference and so that's why they don't lie to us
I'm just so confused
Because I crafted a thing on its strong peak
And it went
1-non-sky, 2-non-sky, 3-non-inc, 4 non-inc, 5-non-inc, 6 non-none, 7-insuff-none
If strong peaks go from -8 to -15, then adding 2 should be Nonexistent - Increasing
If strong peaks go from -7 to -15, then adding 6 should be Nonexistent - Increasing
OR
Increasing is just -2 to -4
and -5+ is Skyrocketing
Which would be
a lot more disruptive
to the definitions
Like, that just can't be right
Too much is built on that
And is demonstrably correct
summer kernel
For the future, I’d like to do a fine-grained look at crafting something up to 7 times the day before a strong peak and up to 5 times the day of a (different) strong peak
But even a rough version of just crafting 7 of something on the day before it strong peaks would be excellent. If you want to and there’s anything that strong peaks C3 that you haven’t crafted any of so far
Hypothesis: what if none is +1 to -2?
I would hate that a lot!
But I wanna look into if it’s possible when I’m awake at a computer
It would explain the C1 thing of things going from 2 to 0 and saying None
Without an override
And it would make my strong peak data make sense with -7 -> -15
Which would explain what Polar’s seeing right now.
So it makes all the data right in front of my face work but there has to be data that contradicts it
In sum, I want to know this info:
What happens when you craft 1, 2*, 3, 4, 5, 6*, or 7* of something on the prepeak of a strong peaking item and when you craft 1, 2*, 3, or 4* on the pre peak of a weak peaking item. The *s are ones I particularly care about.
This format is perfect
mint vine
if we need additional workshops for testing, i dont give a damn about ruining my week
hell, i would have forgotten to go set rest if you hadnt started typing here lol
summer kernel
Idk when you sleep but I’m giving schedule setup advice in #💵︱island-chat if you want to help do science. XP
If you want to re-forget to set up C2 rest
mint vine
i fell asleep at like 6pm and i fear i may not be able to sleep again tonight 😦
i shall go read though
vapid valley
In sum, I want to know this info:
What happens when you craft 1, 2\*, 3, 4, 5, 6...
Remind me next week before you go to bed Monday night
summer kernel
I hope recs bot behaves today
vapid valley
.............................
Is... this week curated?
I mean that's not terrible for CC1
summer kernel
It’s just above 3500 with bonus
vapid valley
CG wants to do Ear - Knife - Brush - Crook - Brush, but Crook is a C6/7
so... uh... no?
summer kernel
I’m pretty sure crook is in our top 10 this week
At least top 15 which is what I think I reserve
vapid valley
G: 12x Garnet (VH) >> 12x Alyssum (H) >>> 12x Spruce (A)
VH:
H:
A:
4hs: VH - ; H -
Crook is like our top 2.
We can't waste it
summer kernel
Oh yeah
vapid valley
Garnet > Crook > Alyssum > everything else pretty much
summer kernel
Right. One of 2 VH 8 hours
vapid valley
We might as well use Jam while we have it, plus potential bonusu
summer kernel
Idk. I just like it
vapid valley
Watch Butter be a bonusu
Butter: C4S
I'll laugh
Butter Strong Bonusu btw
Now that's ~3762
Not baaad
Weak bonusu would be ~3642
Waiting for RecsBot to tell me how dumb and smelly this is before doing it
Oh wait
Ciena mentioned a 6 craft
(thank you Ciena)
Yeah that's 3108 before Groove, so ~3708
So that's even better, and Knife can be a potential bonusu
summer kernel
Let’s go bot
summer kernel
Bot refuses to abandon Jam
vapid valley
It should though cuz this is better
summer kernel
There are 18 brushes in there
vapid valley
It's still better though
Is the bot like hard-refusing to do 3 of a craft in a day?
summer kernel
The chains I’m using are from Leiton
vapid valley
uh?
summer kernel
And I think he filtered out anything that used something more than twice
vapid valley
Probably yeah
summer kernel
I have a list of all efficient schedules that I got from him ages ago
That’s what the bot is brute forcing
vapid valley
I mean using anything thrice is only ever going to be in a 46464 or 444444, cuz otherwise you're gonna be inefficient somewhere
Like you literally can't fit three of the same 8s or 6s, but it should be looking at a triple 4 for those 2 setups
You do agree that the 6-craft is better though, right?
3108 +6 groove vs 3222 +3 groove, so the math is definitely on the side of the 6-craft, plus both have the bonusu chance and if Butter procs and Knife doesn't, the 5-craft is like, BARELY better.
3108+6 vs 3462+3 if Butter procs Strong and Knife is C5
summer kernel
Brush is 100% going to come up later in the week
vapid valley
Oh right for Crook
summer kernel
But I guess who cares. It’s strong peaking now
vapid valley
But it'll be at Suff yeah
vapid valley
yeah
Brush will be unusable after C3
What else links with Crook hmmm
Garnet is 4/5, Crook is 6/7
Bah
It'll end up being necklace.
vapid valley
For the 6-craft?
summer kernel
Yeah
vapid valley
Just to confirm, this is gucci?
Before I pull any triggers, I just want your approval (also thank you for being awake today)
We're just leaving Crook up in the air I guess.
summer kernel
Yeah that looks good
vapid valley
alright
summer kernel
Idk, crook pairs with garnet half the time anyway
vapid valley
Garnet is 4/5 though
vapid valley
You're fine
Garnet 4/5, Crook 6/7
so it's like
Necklace, Brush, and Garnet are all that links to it
inb4 Garnet Strong 5, Crook Strong 6, we just do Garnet - Crook - Garnet on 5 and do whatever on 6
I can't wait for that
summer kernel
Honestly, it doesn’t matter what crook is on 6
We’d only use it once and it’d be Insuff
But man
It’d suck to use both our good crafts in one day 🙃
I’ll have to either poke Leiton or generate my own
mint vine
And I think he filtered out anything that used something more than twice
he did, do you want a list of just all efficient?
mint vine
62,535
mint vine
vapid valley
We’d only use it once and it’d be Insuff
summer kernel
Do you know if Leiton filtered out anything else?
vapid valley
If it's strong and we reuse it on 6 after using once on 5, it'd be strong for the first on 6 and insuf on the second on 6
vapid valley
Yeah
5: Garnet/Crook/Garnet (6x Crook at Insuf)
6: 4848 with 8s being Crook (first 8s are NE, second 8s are Insuf)
summer kernel
I cannot believe the groove mover lives
vapid valley
Yeah I'm annoyed.
mint vine
Do you know if Leiton filtered out anything else?
I dont know. how many items was your previous one?
vapid valley
Because we consistently keep having shit C2s.
Yeah
By "shit C2s", I mean "the non-4hs are so fucking bad that we have to Groove Move"
summer kernel
I dont know. how many items was your previous one?
40 something k. The second digit could be 3 or 8 for all I know
vapid valley
Also first time we're main-reccing a triple craft in the same day
summer kernel
Can you send me your beachcomber csv? I might see what it thinks
vapid valley
Mine?
vapid valley
summer kernel
Dance
vapid valley
I have it set for Rest atm
mint vine
just cutting out the triples, it would be 59927 possibilities, so he definitely trimmed something else out
i broke Excel twice trying to do countifs on that many cells at once, lol
62,535*50 is a lot apparently
lol, Mienna look what my default Chrome tab still is:
mint vine
nope
summer kernel
I remember Leiton talking about what he took out of the list
And I was like “hey, I might want you to add the 3 same craft schedules back in because my solver can account for that”
And now I suppose CG can too
But the rest I agreed with being gone
And I forget what they were or why
mint vine
yeah, i only remember the triples, but...all on the old server
summer kernel
Big day for me wanting info from the beforetimes
Anyway
I’m going to bed
This is exactly what I was not supposed to be doing when I took time off to move. >.<
mint vine
haha, sleep well!
signal fog
I only removed triple-crafts and inefficient combos
I think
Don't have time to fully theorycraft the rec yet but I'm a lil iffy on going for the 6-craft over jam, when most of the top products are peaking late this week.
vapid valley
Don't have time to fully theorycraft the rec yet but I'm a lil iffy on going for...
I mean it's all a gamble anyway.
3222 vs 3108 +3G
I don't disagree with your logic, but it's hard to toss 3 groove for only 114 . Both also have the 50-50 on Knife or Butter which we won't know which would be truly better until after C3 starts, and in the very unlikely event that Butter hits, and is strong, and Knife doesn't hit at all, it's still equivalent if you look at groove as +100 per point. The bonusu are EXACTLY the same as well, both being 4h leaving crafts at High pop too.
But like Garnet could C4S and boom, the groove was worth. Nothing could proc on C4 and it's wasted. The system wasn't "meant" to be solved like this which is why C2 and C3 are always such gambles
signal fog
The issue is that regardless of how hard Garnet hits, we're guaranteed heavy C6 and C7
So our average of 4300 or whatever it is that we use for estimated value and base our groove bonuses on... doesn't work as well
It should be weighted less because the C4/5 peaks are worse
If we get a strong Garnet on C4, that miiiiiight make the groove worth? But I dunno, I'd have to do the math and I need to cook dinner. Today was another jampacked day and I didn't have any time to work on anything, and I have company to entertain tonight, and aaaaaaaaaaaaa
mint vine
Yeah but we can't and shouldn't use a dynamically changing weighted value for groove.
signal fog
We should if we can build a system to account for it
Can't be that hard, you just look at current peak predictions and weight the values accordingly
mint vine
except those dont exist on the most important day
signal fog
Well we'd use the average for C2
But for C3 and C4, we should absolutely be weighing it based on current predictions
C3 especially
vapid valley
Either way, the value is like, so miniscule.
We're talking a difference of a hundred or two
signal fog
squint
mint vine
i mean the whole point of using the average is that it IS the average
signal fog
Our entire selling point is being able to solve the absolute maximum possible
If people wanted good enough, they'd go elsewhere
vapid valley
Our entire selling point is being able to solve the absolute maximum possible
Yeah and that's not always possible.
signal fog
We say we're the best, we'd better be giving the best, and there's no such thing as "not worth the effort"
vapid valley
Of course, but when we LITERALLY cannot know which is better until it's too late, that's when I'm saying it's close enough to not really matter
signal fog
It's an idea, and there's no reason not to at least try it
vapid valley
No of course
mint vine
its the math thats difficult. coming up with what number to use
signal fog
I've been thinking about it for weeks at this point, I have ideas
vapid valley
It's a really annoying mechanic
mint vine
its a great mechanic, its an annoying thing to try to optimize around
because regardless of what you do its always a gamble
one thought would be simming the worst possible and best possible days for 4/5 and using the average of those
signal fog
Or even more simple
Assign a value to popularity, assign values to the Big 9 and solid 6's, tally them up
Cuz we don't need to know EXACTLY how much groove value we're getting
We just need to know if we'll get our money's worth or not
Late peaks? No, we won't
Early peaks? Sure!
Right now our groove math kind of assumes early peaks, and I don't think that's safe
mint vine
it does not assume early peaks
signal fog
We assume 4300
mint vine
the bot doesnt work that way anymore
signal fog
What did I miss O:
mint vine
a whole fuckload of headache of Mienna and I redoing groove value calculations
vapid valley
Honestly, this is another case of "hey here's an option to rest" like last week, except now it's "hey you can do Jam 5 craft or this 6 craft and the difference is even less than last week"
But you know the people hate options
I did the best I could with the tools I was given and also asked Mienna to double-check and she said yeah. I'm not blaming anyone, I just, I used what I was given
mint vine
1 groove = 1% of the average value of crafts, with groove stripped completely out, for days it will apply to
so 1 groove = 40 cowries for each day that gets a benefit, and then any partial days get split into 10%/40%/10%/40% of that value based on the number of crafts that benefit from it
signal fog
Oh perfect, that second half was also gonna be part of my groove rework
mint vine
trying to find the actual convo if you want to read it but...searching for groove or numbers in this discord is just fucking lunacy 😄
signal fog
Ye sorry D:
signal fog
I wanted to take a full hiatus and rest up but between sheet updates and work suddenly getting INSANELY busy out of nowhere, it's been stupid hard to keep up with anything
mint vine
lol I am juggling work like mad right now so i get it. I don't know why I am even doing this supply demand test today of all days >.<
bad days could definitely be added into the calculation for sure, but finding a way to fit that into the math is the hard part, at least with succinct, accurate math.
what we don't have is an easy way to view grooveless history though and I havent gotten around to making a sheet to do that, cause it sounds tedious lol
signal fog
That's essentially my current issue. I just don't have enough time for anything lately and all of my ideas require days/weeks of work
Peak converter, manual entry bible, java solver
I haven't even made my video tutorial yet x-x
mint vine
honestly I have free time after work but I've been doing this strange thing where I actually play XIV >.>
signal fog
I haven't actually played XIV in months xD
mint vine
also making this relic tracking spreadsheet for myself lol
well heavily modifying someones old one
summer kernel
Peak converter, manual entry bible, java solver
Sorry, what is that last one? XP
summer kernel
Sneaking onto the computer to run numbers
Necklace: Found -15 supply with 0 crafts done today
Necklace is worth 52 with 0 groove at Nonexistent supply (-15) and Average popularity
Brush: Found -15 supply with 0 crafts done today
Brush is worth 240 with 3 groove at Nonexistent supply (-15) and VeryHigh popularity
Culinary Knife: Found 0 supply with 0 crafts done today
Culinary Knife is worth 132 with 6 groove at Sufficient supply (0) and High popularity
Brush: Found -15 supply with 6 crafts done today
Brush is worth 254 with 9 groove at Nonexistent supply (-9) and VeryHigh popularity
Culinary Knife: Found 0 supply with 6 crafts done today
Culinary Knife is worth 140 with 12 groove at Sufficient supply (6) and High popularity
Brush: Found -15 supply with 12 crafts done today
Brush is worth 218 with 15 groove at Insufficient supply (-3) and VeryHigh popularity
Not using main item Spruce Round Shield that hasn't peaked yet, but are using its helper Brush. Adding penalty of 2x 10
Not using main item Spruce Round Shield that hasn't peaked yet, but are using its helper Brush. Adding penalty of 2x 10
Not using main item Spruce Round Shield that hasn't peaked yet, but are using its helper Brush. Adding penalty of 2x 10
Not using main item Crook that hasn't peaked yet, but are using its helper Brush. Adding penalty of 2x 30
Not using main item Crook that hasn't peaked yet, but are using its helper Brush. Adding penalty of 2x 30
Not using main item Crook that hasn't peaked yet, but are using its helper Brush. Adding penalty of 2x 30
Day 3, 0 starting groove. Workshop value: 1036, grooveBonus: 206, material cost: 60 x0.5, helper penalty: 240, prepeak bonus: 20, total: 992```Compared to
Earrings: Found 7 supply with 0 crafts done today
Earrings is worth 72 with 0 groove at Sufficient supply (7) and VeryHigh popularity
Butter: Found 0 supply with 0 crafts done today
Butter is worth 128 with 3 groove at Sufficient supply (0) and High popularity
Jam: Found -8 supply with 0 crafts done today
Jam is worth 360 with 6 groove at Insufficient supply (-8) and VeryHigh popularity
Butter: Found 0 supply with 6 crafts done today
Butter is worth 136 with 9 groove at Sufficient supply (6) and High popularity
Jam: Found -8 supply with 6 crafts done today
Jam is worth 378 with 12 groove at Insufficient supply (-2) and VeryHigh popularity
Day 3, 0 starting groove. Workshop value: 1074, grooveBonus: 105, material cost: 72 x0.5, helper penalty: 0, prepeak bonus: 20, total: 1163```Here's the info on reserved helpers: Reserving helper Firesand to go with main item Garnet Rapier (#1), difference between Firesand and Macuahuitl? 39 cost of stepping down? 16 Penalty: 17 Reserving helper Brush to go with main item Crook (#2), difference between Brush and Necklace? 39 cost of stepping down? 29 Penalty: 30 Reserving helper Firesand to go with main item Growth Formula (#3), difference between Firesand and Potion? 6 cost of stepping down? 16 Penalty: 4 Reserving helper Brush to go with main item Spruce Round Shield (#4), difference between Brush and Necklace? 39 cost of stepping down? 29 Penalty: 10
So the bot, with knowledge of new schedules, really doesn't like using Brush before both Crook and Spruce get a chance to
It looks like with Necklace at Average - Suff and Brush at VH - Overflowing, Brush is still worth more
so 9 to 1 we're going to be using more Brushes later in the week
Brush at Suff instead gets us 118 more per workshop
So this schedule basically guarantees that we lose 354 cowries later in the week
At least. We might use it on both Crook and Spruce
If we use it on both, that's an additional 59 per workshop loss (what we get by having Overflowing instead of Surplus)
So let's say that's a 50-50
Means an average of 148 loss per workshop from tanking Brush before we're forced to use it later
-148 + 100 extra groove
doesn't work out
It's still an overall penalty
And the schedule is worth less at base
Even just -118 + 100 doesn't work out
Crook, Growth Formula, and Spruce are our moneymakers on 67
We have to use at least one of Crook or Spruce on 6 or 7
It's up to you whether we retract the rec, (I know it's very late and people will get mad, so we probably shouldn't)
But this schedule was a mistake and will lose cowries
Unless we end up in a baffling situation where we don't use Brush at all later
And that's discounting Leiton's groove concerns, which are valid
Though it's hard to figure out their mathematical impact
Sorry
Back to packing >.<
After I live in a house, I will be able to be around and active more at rec time
Instead of just giving off the cuff advice
I wonder if a way to do it is to take the average of C4+C5 grooveless estimates, the average of C6+C7 grooveless estimates, get the ratio, and then multiply that by the 4044 average
That makes sense to me
Also it'd be great to derive the 4044 average from something mathematically so we don't have to wait 12 weeks to have a good number once we get new crafts, but one thing at a time
mint vine
Also it'd be great to derive the 4044 average from something mathematically so w...
I kind of want to strip all the groove out of silera's sim and see how those turn out but I have to reverse engineer the popularity first which will take some time
vapid valley
BONUS KNIFE KNIFE C3S, BUTTER NO BONUS 
WE'RE ON THE GOOD TIMELINE EVERYONE
Hmmm, getting different values between CG and BC
It's definitely double-Garnet day though
Oh, Crook weak
Man this week is so fucking cursed.
summer kernel
Oof
mint vine
omg go to bed
summer kernel
It got so close!!
summer kernel
I am not going to bed at 3:07
mint vine
lol
summer kernel
-354 though.
vapid valley
For what? Net?
summer kernel
Spruce is still on the table, and weak crook might still be our best thing
vapid valley
Oh it does not nvm
I was gonna say 4 - Scythe - Crook - Scythe
But that doesn't link sad
Crook going weak is really fucking sad
We also still have Growth
2/3rd's chance strong
summer kernel
Growth will probably be one day but we have 2. XP
vapid valley
Necklace Earring? Huh.
Idk what's up with CG then
It's saying 4131 4719 for Garnet x2
... and Ciena's does too
summer kernel
👀
vapid valley
CG and Ciena say 4719
BC and RB say 4659
So... which is it? 
Like even the grooveless value is different
@summer kernel Any thoughts on the value disparity?
summer kernel
What are the respective supplies and popularities?
summer kernel
Best guess is that If it’s 6/7 high, making it on C4 would bump the supply tier.
On firesand
summer kernel
Yep. And both my tools assume worst case for future peaks
vapid valley
I never even thought of that
mint vine
yeah better to assume low and get bonus
vapid valley
Should I just manually input the lower value for the picture/recs?
summer kernel
Lying! 🎉
vapid valley
**__Recommendations for Season 16 (Dec 6 - 12)__**
@ Roles go here.
**• C5:** 3...
How's this? Lol
Just covering all our tracks.
mint vine
seems fine
summer kernel
Yep
I feel like I should delete the update post. Like, we aren’t doing that, yeah?
vapid valley
In recs? That's up to you
In the future you can always make a post before I update recs for the next day anyway
mint vine
update post?
vapid valley
Also @summer kernel can we keep the C3 rec as is for the bot?
Cuz asking people to swap the 2nd 4h is... unfeasible and I think we'll be okay without brush
I'm not even swapping mine
summer kernel
Yeah, that’s why I asked if I should delete the update post
vapid valley
Ohhhhhhhhh
I thought you meant in #🦜︱recommendations your post
Yeah I'd delete that
summer kernel
It’s worth fewer cowries anyway. The weighted value is just higher because my algo doesn’t want to use brush this early
vapid valley
So people don't get confused
I mean tanking it to OF is... yeah
I understand it may wanna keep it at Surp
Crook being Weak is a 
though
summer kernel
I mean, we’re only not using it if we get something better.
Idk
Cursed week
Rec for today is v clear
Bummer it’s weak tho
But yeah, demonstrably sometimes we use a VH weak peak. XP
Good night
Wish me luck getting all my belongings into transportable shapes
vapid valley
@summer kernel https://twitter.com/WorkshopParrot/status/1600769517696987139
You have me saying the word "Crunchy" now lol
vapid valley
....... ffs.
That's what I MEANT
but typing dylsexia.
(that one was intentional lol)
Also teaching some people stuff like "D6S"
Can't wait til we pull the rug out from Twitter and switch to C and Season.
mint vine
week?! decade.
signal fog
I'm too tired to knife
mint vine
I've been up since my 2 hour nap yesterday. I'm pretty sure I'm getting delusional
vapid valley
@summer kernel How do you feel about CC4 being a 6-craft that TRIPLE-crafts two items?
🥴
Time to uh, change how recs are done, forever.
vapid valley
Fuck. This. Season.
summer kernel
I noticed that some recs for later days were 6 crafts when I ran historical testing with the new chains file
So I’ll get the bot updated once I sort out a couple weird oddities
But damn
This week
vapid valley
So I’ll get the bot updated once I sort out a couple weird oddities
. Also just in CASE, look at Inefficient stuff too if it doesn't like, totally fry the bot.
mint vine
oh god, that list might be gigantic
vapid valley
Alright then yeah, just allow the triple crafts
Also man every time we do a 6 craft, I cry that BC becomes useless without me figuring out exactly how to place the damn thing perfectly and I really don't like messing with my csv
mint vine
just the 6-item chains with inefficient is 16.7 million combinations >.>
summer kernel
It hangs a little on normal days but I think it'd hang substantially on calculating day 5-7
I'm scared D:
vapid valley
Yeah and it's probably not worth anyway
I'm still laughing at C6.
Crook just had to weak.
and it's not even cuz we used Brush on C3 either; even if you didn't touch Brush, it's still worse.
mint vine
even if groove is already capped so you cant get more, i cant picture any case where inefficient is better, because it means your value would need to be able to make more than 2x the quantity
mint vine
even something like making 3 high value 8-hour items in a row for a workshop, with efficient you're guaranteed to make at least 3 of them
vapid valley
I just remember the one time we did a 4488 agenda
and I'm thinking it could be 4848 with the 2nd 4 being inefficient
vapid valley
Cuz like, the 4s don't really matter in those agendas
mint vine
inefficient helpers, shit, yeah that could matter
vapid valley
I feel like it would ONLY be 4848 agendas.
No other agenda type makes sense. Inefficient 6s and 8s uses too much of the cycle for it to just be a 1x
It already feels bad on Confections and Arms days
888 just feels bad.
mint vine
dont we do 46446 sometimes? that second 4 could be inefficient potentially
vapid valley
I guess my question would be "how much would it fry the bot to allow inefficient on one agenda type only?".
vapid valley
dont we do 46446 sometimes? that second 4 could be inefficient potentially
I don't remember 46446
Cuz it's like why not 44646
mint vine
true, but same point, the third 4 there could be inefficient if it helped the right 6
vapid valley
potentially, but, eh?
vapid valley
I feel like it would ***ONLY*** be 4848 agendas.
@summer kernel how bad would this be?
(also is it even possible to implement only for one agenda type?)
mint vine
one type yes, cause its using a csv so you'd just add those possible combinations to the csv. not sure if her code checks for efficiency at each stage though or assumes all are efficient, so depending on that it might take some re-coding
summer kernel
I do a little of both. When actually calculating value, I check efficiency, but when doing things like estimating how much groove a schedule will make I just check its length
So some of my heuristics would need to be adjusted
But the actual calculations would all be correct
But yeah, implementing all permutations of efficient 4-8 pairs can't add that many schedules to the list
vapid valley
But yeah, implementing all permutations of efficient 4-8 pairs can't add _that_ ...
inefficient
vapid valley
Oh
summer kernel
You just cram them together
vapid valley
Gotcha
You know better just wanted to make sure that wasn't a typo. My eyes still can barely see
You know what's worse than sleep apnea? Not sleeping for long enough on top of that
mint vine
But yeah, implementing all permutations of efficient 4-8 pairs can't add _that_ ...
its 625 total, including the ones already there, so not too bad
lol
nope, 576 total
i can math i swear
but thats all 48 efficients, some of which i dont think we have ever made
damn, excluded more, 400 total. anyhow, definitely a small number
summer kernel
Is that still before excluding ones that secretly are efficient/are already on the sheet? Like Potion - Growth Formula - Firesand - Garnet Rapier?
mint vine
yup, before excluding any
there's 20 48 pairs
20 or 21, lol
i used my file for this SF craziness and may have messed one up
mint vine
yeah, i was honestly considering doing it via powershell lol
that sounded like too close to actual work though
like my job work
summer kernel
How are you doing it?
mint vine
the way I did it before was some craziness in excel using VBA to make all possible combinations then go back through and delete inefficient, which is an awful way to do it
for powershell I think I'd have to do 6 embedded loops since they'd be up to 6 long
and dump each success to a file
summer kernel
Mhm
mint vine
yeah I am totally working on making it work in powershell rather than working. lol
summer kernel
Including all inefficient combos of 28 4-8 pairs
Wrote 63071 chains to file.
Love how we all get different numbers
My 4-8 pairs seem to be these plus Rope + Bed
Gotta make sure I keep all those Quartz Ring combos in mind
Oh, and
But I get 62535 fully efficient schedules without those
mint vine
thats what i had before right?
mint vine
now to see what I get from powershell...
lol i get nothing because somewhere a } is missing >.<
well its taking forever to run so thats a good sign
summer kernel
Java was very fast ^^
mint vine
yours is probably a lot better than mine 😄
summer kernel
Adding the inefficient 4-8 partners added 536 schedules
I think I can handle that
(This does account for removing ones that are secretly efficient)
mint vine
I'm pretty sure mine is going to try every possible combination of the 50 items, which is 15 billion combinations, lol
summer kernel
That's basically what mine is doing but with two crucial pieces
mint vine
well, 15,625,000,000
summer kernel
continue;
int time3 = time2 + item3.time;
if(time3 > minScheduleLength && time3 <= maxScheduleLength)
{
addChainToList(Arrays.asList(item1.item, item2.item, item3.item));
continue;
}```i dont know why but I read that as AddChainToilet
yeah i dont have any special functions to check efficiency
i should re-write what I'm doing with efficiency though
like yours is not going all 15 billion
cause if 1 and 2 dont match, its skipping checking any 3-6 for those
summer kernel
Yep
But cool. Quick and easy. Should be no problem to run it again once 6.3 comes out
mint vine
hmm
mines writing row 200kish atm
wtf lol
oh, didnt trim MJICraftworksObject enough
all those ending 0 items with 0 time add a ton of possibilities, lol
summer kernel
Oh boy
mint vine
whats your minScheduleLength?
mint vine
no worries, I moved on to doing something more of a pain now
summer kernel
I'm allergic to magic numbers
Once I've copied and pasted the same constant into a new place in my code, I break out in hives
mint vine
lol
I'm adding another section to have it do it purely from raw data files, without any tweaking, and also grabbing the proper object name from Items.csv
summer kernel
Ahhhh
mint vine
I dont know why, because i'm crazy?
summer kernel
Yeah, my Java solver still relies on hardcoded data
BC reads it from the files
I could just copy it over
but
I don't wanna.
(I should)
mint vine
loading Items.csv sucks
mint vine
and knowing which line you need!
summer kernel
IT's awful
mint vine
I am hesitant to use Item ID -2, but having to search for Item ID in the whole file is rough
