1.6 Update Desert Festival Villager Shop Randomness Details

[PerfectionWhaleFarm]

Based on the wiki page here: https://stardewvalleywiki.com/Desert_Festival#Villager_Shops:

Every day, two villagers are randomly chosen to host their own shops at the festival.

I was led to believe that all villagers have a random and equal chance to appear in any of the three days as a vendor.

I'm curious to know if anyone knows if that interpretation is true, or if anyone has delved deeper into this topic.
Based on my experimentation of multiple day 1 vendors, I can tell (at least from the Switch version) that:
- Alex, Haley, Caroline, Pierre, Sam, Sebastian, Gus have never appeared for me on day 1 (Y8 if that matters)
- Pam, Maru, Leo, Kent, Clint, Demetrius, George, Evelyn, Marnie, Harvey, Penny, Robin, Vincent are among the vendors that I remember appearing on day 1

I have done this testing on 40 unique seeds, so 80 possible vendor outcomes, so having 7/29 chance not happening 80 times is suspicious.
My hypothesis is that villagers who attend the Desert Festival (as a non vendor) will never show up as vendor during that day, and I'm curious to be proven/disproven (or maybe I simply misunderstood the wording from the Wiki page).

This is not an issue for me, I'm more curious from a pure technical standpoint and so I understand the exact behavior (if anyone is aware of it).
I do have to say that I'm sure this is not a bug either, and I truly believe all villagers do show up for the Desert Festival. My suspicions are that not all villagers appear on every day.

 

[PerfectionWhaleFarm]

After experimenting all three days, I can say for a fact that (at least on the Switch), villagers who attend the Desert Festival on specific days as a non vendor will never show up as vendor during that day, is true.

 

[MogBeoulve]

I'm guessing it is done that way so that the villagers who are visitors each day don't need to be changed on the fly. Keep the visitors and their special routes and dialogue intact, then pull two random villagers from those remaining to be the vendors. The posted explanation is probably just a simplification.

 

[PerfectionWhaleFarm]

Yeah I'm definitely not disagreeing on that.
However, the Wiki tends to get crazy internal and technical details very correct so there's definitely ways this wording can be improved for accuracy (this probably won't impact 90% of the playerbase lol).

Especially since they have a "Villager Attendees" on the same page that details who attends the festival on what day (which is how I even figured out the "non-random" part of the randomness in the first place, so integrating this detail could have been as simple as "Every day, two villagers that are not active attendees of the Desert Festival under the "Villager Attendees" section (or provided as a hyperlink) are randomly chosen to host their own shops at the festival" or something to that degree.

This of course excludes Emily and Harvey, who are "doing work" for the festival for all three days and not regular attendees so to speak.

I'm also not sure if that is the behavior on PC as well (which is the primary platform) so hesitant to update the wiki in case my finding is wrong (but also I'm not even sure how to update the wiki), so I can't speak to generalizing this finding throughout.
Just an interesting tidbit for now, in case others have this doubt as well.

 

[FilthyGorilla]

Yeah I'm definitely not disagreeing on that.
However, the Wiki tends to get crazy internal and technical details very correct so there's definitely ways this wording can be improved for accuracy (this probably won't impact 90% of the playerbase lol).

Especially since they have a "Villager Attendees" on the same page that details who attends the festival on what day (which is how I even figured out the "non-random" part of the randomness in the first place, so integrating this detail could have been as simple as "Every day, two villagers that are not active attendees of the Desert Festival under the "Villager Attendees" section (or provided as a hyperlink) are randomly chosen to host their own shops at the festival" or something to that degree.

This of course excludes Emily and Harvey, who are "doing work" for the festival for all three days and not regular attendees so to speak.

I'm also not sure if that is the behavior on PC as well (which is the primary platform) so hesitant to update the wiki in case my finding is wrong (but also I'm not even sure how to update the wiki), so I can't speak to generalizing this finding throughout.
Just an interesting tidbit for now, in case others have this doubt as well.

You can create an account for the wiki and make changes from there but I’d generally avoid putting any experienced-only observations (unless the chance is insanely low and even then ask in a community that can check code or do so yourself).

Especially if they’re platform specific and not backed up with anything from the code to actually suggest so

I belive your observation is correct though the desert fest vendors are a bit finicky. They’re supposed to be seeded and generally function as such but can under some specific circumstances be changed out, as I’ve noticed a couple people I know were testing (which I may ask them for any conclusive results if they have them)

 

[PerfectionWhaleFarm]

Yep I agree, and I have the same hesitation as well. Experiential evidence is never solid enough conclusions (though I'm certain something is definitely up for the Switch version to make the randomization biased, even if it means just reducing the chances for some vendors), and I currently don't have 1.6 source code to check. Would be curious if those people can corroborate my findings!

Especially for consoles, I know the seed generation process may vary in a way that PC isn't. For example, some time ago I found that the Switch seeds never generated certain items in Remixed Mines rewards pre-1.6 (I don't know if the same behavior still happens after update), and someone who knows the RNG better was able to confirm that for me as well. However, that type of issue just did not happen for PC.

 

[VampireCake]

There's definitely something finicky about the vendor choices, even the predictor has a hard time with them. As best I can remember from the ones I've been to, it tends to get Monday right but Tuesday and Wednesday are wrong more often.

I do wish the calico eggs would carry over to the next year. There's a lot of things to buy. Also then I could just collect them as well. (I totally do not have like 47 extra gold tags from the trout festival because they're shiny. Nope.)

 

[PerfectionWhaleFarm]

There's definitely something finicky about the vendor choices, even the predictor has a hard time with them. As best I can remember from the ones I've been to, it tends to get Monday right but Tuesday and Wednesday are wrong more often.

I do wish the calico eggs would carry over to the next year. There's a lot of things to buy. Also then I could just collect them as well. (I totally do not have like 47 extra gold tags from the trout festival because they're shiny. Nope.)

Desert Festival has to be my favorite! It's the only thing I look forward to every year haha. You are sooo right, there's so many things to buy and I love that it gives every villager in town an active role in (capitalism?) the valley. I tend to find that if I tried to get to 25 rating for Gil, that is enough, but I'm still working toward the illusive 55+ rating (I was soooo close one day at 53, but between needing to wait for Emily to open shop, the cacti and exchanging for goodies before the villagers leave for the night, I was always a few short). Hopefully next year treats me right!

 

[stardew_luv]

Based on the wiki page here: https://stardewvalleywiki.com/Desert_Festival#Villager_Shops:

I was led to believe that all villagers have a random and equal chance to appear in any of the three days as a vendor.

I'm curious to know if anyone knows if that interpretation is true, or if anyone has delved deeper into this topic.
Based on my experimentation of multiple day 1 vendors, I can tell (at least from the Switch version) that:
- Alex, Haley, Caroline, Pierre, Sam, Sebastian, Gus have never appeared for me on day 1 (Y8 if that matters)
- Pam, Maru, Leo, Kent, Clint, Demetrius, George, Evelyn, Marnie, Harvey, Penny, Robin, Vincent are among the vendors that I remember appearing on day 1

I have done this testing on 40 unique seeds, so 80 possible vendor outcomes, so having 7/29 chance not happening 80 times is suspicious.
My hypothesis is that villagers who attend the Desert Festival (as a non vendor) will never show up as vendor during that day, and I'm curious to be proven/disproven (or maybe I simply misunderstood the wording from the Wiki page).

This is not an issue for me, I'm more curious from a pure technical standpoint and so I understand the exact behavior (if anyone is aware of it).
I do have to say that I'm sure this is not a bug either, and I truly believe all villagers do show up for the Desert Festival. My suspicions are that not all villagers appear on every day.

Ok you Stardew stats analysts are both insane and insanely smart. I couldn’t begin to think that way let alone do all the research and tests required to prove those theories. You are on a whole different wavelength than I am. @FilthyGorilla is insane and insanely smart about all these stats too!

 

[PerfectionWhaleFarm]

Thanks, but it's really just laziness coupled with curiosity, no smarty pants over here at least :D
I got tired of skipping through the whole year just to get to the Desert Festival again, but I just really wanted a longer Desert Festival where I see different vendors and what Emily can tailor me, so I decided to see if it's possible I see certain folks more often than others - then I realized I could never get certain pals to show up, so then it became a mission to figure out why.

The analysis is really just insurance to save me from working hard . I don't want to start some crazy hunts just to realize due to an in-game bug or misunderstanding, it may be literally impossible...

 

[FilthyGorilla]

Ah ok @PerfectionWhaleFarm, I figured some stuff out.

Yes they are seeded, BUT for some reason the way that the game draws for which npcs should be vendors each day depends on the order in which each location is listed in the save file then the order of npcs listed in that location when last saved afaik (so you’re able to influence what vendors appear depending on when you sleep the day prior).

I would assume this is the same for switch, with some changes perhaps to values in the way each is selected (switch has different rng values). Not super useful as you aren’t able to go back to the day before on consoles.

 

[FilthyGorilla]

Ok you Stardew stats analysts are both insane and insanely smart. I couldn’t begin to think that way let alone do all the research and tests required to prove those theories. You are on a whole different wavelength than I am. @FilthyGorilla is insane and insanely smart about all these stats too!

Haha thanks, have to agree with whalefarm here though too, just some latent curiosity and good connections (and a willingness to admit what you don’t know and ask questions)

 

[PerfectionWhaleFarm]

Ah ok @PerfectionWhaleFarm, I figured some stuff out.

Yes they are seeded, BUT for some reason the way that the game draws for which npcs should be vendors each day depends on the order in which each location is listed in the save file then the order of npcs listed in that location when last saved afaik (so you’re able to influence what vendors appear depending on when you sleep the day prior).

I would assume this is the same for switch, with some changes perhaps to values in the way each is selected (switch has different rng values). Not super useful as you aren’t able to go back to the day before on consoles.

Thanks for getting that sorted! I'm not quite sure I understood what you mean by "order in which each location is listed in the save file", does the order of location saved varies day-to-day? Like based on the physical locations the farmer has visited the day prior? I suppose that's intriguing in and of itself.

So I guess that means then there is a chance that vendors could be visiting villagers for that day of the Desert Festival on the PC (based on your description).

I still find that I have not yet disproved my hypothesis (that visiting villagers won't be vendors) on the Switch based on my trials and I have gone through many seeds after this post too. It is also possible that because I'm hot-swapping seeds (i.e., not loading/letting Desert Festival happen in an organic way) that that may have introduced the quirks I'm observing, and that it may not happen if I just go through Desert Festival(s) without hot-swapping seeds.

I'm curious to try a few things:
- A few organic Desert Festivals (with no shenanigans)
- Clone my save with seed swapping shenanigans and see if taking steps the same way or different way and sleeping at the same time / different time to influence who are vendors the next day (sleeping on the same time should lead to the same vendors and sleeping on different time must produce different vendors)
and see if my hypothesis can be disproved (or if I can corroborate your findings on the Switch).

 

[FilthyGorilla]

Thanks for getting that sorted! I'm not quite sure I understood what you mean by "order in which each location is listed in the save file", does the order of location saved varies day-to-day? Like based on the physical locations the farmer has visited the day prior? I suppose that's intriguing in and of itself.

Locations stay the same, npcs within the locations can differ, I believe depending on where they were when the day was saved but I'm not entirely sure

I still find that I have not yet disproved my hypothesis (that visiting villagers won't be vendors) on the Switch based on my trials and I have gone through many seeds after this post too. It is also possible that because I'm hot-swapping seeds (i.e., not loading/letting Desert Festival happen in an organic way) that that may have introduced the quirks I'm observing, and that it may not happen if I just go through Desert Festival(s) without hot-swapping seeds.

I would assume this would disprove that but I haven't checked for any additional code that would dictate any more criteria than chance then that selection I mentioned before. I would assume switch is no different in core process as stated before.

I'm curious to try a few things:
- A few organic Desert Festivals (with no shenanigans)
- Clone my save with seed swapping shenanigans and see if taking steps the same way or different way and sleeping at the same time / different time to influence who are vendors the next day (sleeping on the same time should lead to the same vendors and sleeping on different time must produce different vendors)
and see if my hypothesis can be disproved (or if I can corroborate your findings on the Switch).

Do keep me updated, I'm curious to see your outcomes, I'll personally try to come to some harder conclusions too, I have some people I can ask

 

[PerfectionWhaleFarm]

Locations stay the same, npcs within the locations can differ, I believe depending on where they were when the day was saved but I'm not entirely sure

Okay, actually that explanation makes a lot of sense. I wasn't sure exactly what this meant before.

I would assume this would disprove that but I haven't checked for any additional code that would dictate any more criteria than chance then that selection I mentioned before. I would assume switch is no different in core process as stated before.

Effectively yes, having code always helps understand the behind-the-scenes. I'm just not 100% there yet as the live observations don't exactly align, and I suspect some Switch shenanigans may be at play here as well.

To disprove this, ideally I would like to see a vendor show up at a day when they would have been visiting the festival on the Switch. I haven't seen any posts about that yet(?) but your explanation helps me set up the scenarios more properly as there are a few flaws in my previous tests as well.

I'll deffo update here if I find anything supporting either!

 

[PerfectionWhaleFarm]

Update:
I have done five saves with a various amount of seeds and sleep times, and if there are any relations between time slept the previous day and next-day vendor, I have not seen it (or it merely boosts the chances of some things happening). And even with this new round of testing (coupled with many dynamic seed swaps previously), I have yet to see any vendors who would have been visiting that same day.

The seeds below are all set up and ready by Sun. 14 Spring Year 10, and which don't further manipulate seeding conditions at that point.

Save 1 (Seed 3) - Sleep in immediately at 6:00AM at Sun. 14 Spring Year 10; Played a full Desert Festival day for Mon. 15 Spring Year 10, Tues. 16 Spring Year 10

1. 15 Spring - Robin, Shane
2. 16 Spring - Sebastian, Leah
3. 17 Spring - Clint, Robin

Save 2 (Seed 1) - Sleep in immediately at 6:00AM at Sun. 14 Spring Year 10; Went to sleep at 12:00PM for Mon. 15 Spring Year 10, Tues. 16 Spring Year 10

1. 15 Spring - Jodi, Vincent
2. 16 Spring - Shane, George
3. 17 Spring - Demetrius, Sam

Save 3 (Seed 2) - Sleep in at 3:00PM at Sun. 14 Spring Year 10; Went to sleep at 12:00PM for Mon. 15 Spring Year 10, Tues. 16 Spring Year 10

1. 15 Spring - Robin, Shane
2. 16 Spring - Sebastian, Leah
3. 17 Spring - Clint, Robin

Save 4 (Seed 1) - Sleep in at 3:00PM at Sun. 14 Spring Year 10; Went to sleep at 11:50AM for Mon. 15 Spring Year 10, Tues. 16 Spring Year 10

1. 15 Spring - Clint, George
2. 16 Spring - Gus, Abigail
3. 17 Spring - Emily, Leah

Save 5 (Seed 1) - Sleep in at 3:00PM at Sun. 14 Spring Year 10; Went to sleep at 11:50AM for Mon. 15 Spring Year 10, Tues. 16 Spring Year 10

1. 15 Spring - Jodi, Kent
2. 16 Spring - Shane, George
3. 17 Spring - Demetrius, Sam

Based on this finding, I cannot disprove my hypothesis (which would have been the result I would like to see from this test, as it can definitively rule out something). But based on this now, I can't say at all whether or not my hypothesis is for sure true or false as the results just aren't strong enough. I still highly suspect that is the case, because to not have that happen this many times in a purely random event is quite unlikely. If it does not remove the villagers from the roster of vendors outright, it must have significantly reduced the likelihood of their appearance.

I am not 100% sure I see a pattern between vendors and previous day sleep times. Certain matches may have been coincidental (e.g., Save 1 and Save 3) or maybe the patterns only guarantee a higher likelihood of certain vendors to show up a certain way.