[X] My Assistant

[pumpkinking0192]

In the wake of Erty's investigations into the mechanics of the activity score and the Pokewalker, I found myself wondering exactly how the additional grabs gained by Shelter Passes worked. All I could find was a vague post by Big Bidoof:

It's number of Shelter Passes * floor(((some number * interactions) + (some number * proper interactions))/some number) that get added, yup yup.

But I found myself wondering about the actual numbers, as well as whether the formula given way back in '10 was the one still used. (I believe the quote is from before the activity score was introduced, so I wasn't sure whether the shelter grab formula would have been revised to take the activity score into account. As it turns out, it hasn't changed.) To figure this out, I outlined and performed a little experiment. For the purposes of this investigation, the variables (and the resulting formula) are as follows:

S = Shelter Passes
G = # Grabs
A = Activity Score
I = Interactions (all)
P = Proper interactions

G = (S * floor(((x * I) + (y * P))/z)) + 4S + 6
which simplifies to:
G = (S * floor((Ix + Py)/z)) + 4S + 6

Since I have all 5 Shelter Passes, my daily grabs start at 26 and increase in multiples of 5. For two days, I clicked only eggs; for another two days, I clicked only proper berries. Since I wanted to rule out the possibility that the formula had changed to take activity score into account, the former of each pair of days was spent trying to get a high activity score (hatching eggs, buying items, etc) and the latter was spent doing little on the site except clicking.

Day Three (proper interactions only, high activity score) ended up being kind of botched because, as I discovered, feeding an Exploration Pokemon with a status effect doesn't necessarily count as a proper berry interaction. I believe this only occurs if the Pokemon was obtained for a prior task and then inflicted with a status effect (like the Salamence in the Genesect exploration). The data on from Days One and Two, however, proved that activity score doesn't affect shelter grabs, so Day Four's data should be sufficient.

When I was unsure exactly how many clicks I had made, I either noted the maximum and minimum, or estimated (~) if I wasn't precisely sure of the maximum and minimum. My data is as follows:

DAY ONE - NO PROPER INTERACTIONS, HIGH ACTIVITY SCORE

G \ A \ I \ P

26 \ 0 \ 0 \ 0
31 \ ~3300 \ ~2950 \ 0
36 \ ~8820 \ 5862-90 \ 0
41 \ ~14000 \ 8829 \ 0
46 \ ~17780 \ 11772 \ 0

There was a difference of exactly 2943 between 41 and 46 grabs; 3(2943)=8829 and 4(2943)=11772, so I went into Day Two assuming grabs were earned every 2943 improper interactions. (This also confirmed that activity score doesn't have anything to do with Shelter grabs, but I continued to notate my activity score for posterity's sake.)



DAY TWO - NO PROPER INTERACTIONS, LOW ACTIVITY SCORE

G \ A \ I \ P

26 \ 0 \ 0 \ 0
31 \ 3283 \ 2943 \ 0
36 \ 6226 \ 5886 \ 0
41 \ 9169 \ 8829 \ 0

I confirmed what I assumed on Day One.



DAY THREE - ONLY PROPER INTERACTIONS, HIGH ACTIVITY SCORE (botched)

G \ A \ I \ P

26 \ 0 \ 0 \ 0
31 \ ~13000 \ ~330 \ 306-29
36 \ ~18880 \ 669 \ 653
41 \ ~25520 \ 1004 \ 979
46 \ ~31930 \ ~1335 \ ~1306

I botched this day because I didn't realize Exploration Pokemon might not count as proper interactions. I assumed, based on my data, that the proper interaction rate was 326.5, rounded down. As I learned on Day Four, this was incorrect; I believe the difference was because of the "improper" interactions factoring into the equations.



DAY FOUR - ONLY PROPER INTERACTIONS, LOW ACTIVITY SCORE

G \ A \ I \ P

26 \ 0 \ 0 \ 0
31 \ 5552 \ 327 \ 327
36 \ 10784 \ 654 \ 654
41 \ 16016 \ 981 \ 981

This day confirmed that the rate is actually 327 proper interactions per grab.

Our equation:
G = (S * floor((Ix + Py)/z)) + 4S + 6

now becomes:
1 = (S * ((327x + 327y)/z)) + 4S + 6
and
1 = (S * ((2943x + 0y)/z)) + 4S + 6

Some poking around on the Internet tells me that unless we know at least one of x, y, or z, we can't derive any of them. But all is not lost; even if we don't know the exact equation, we have still discovered exactly how many clicks must be made to get more Shelter grabs.

If one is careful to click only proper berries (and doesn't click any eggs or Pokemon without preferences), one will earn more Shelter grabs every 327 proper berry interactions.

If one clicks no proper berries at all, one will earn more Shelter grabs every 2943 interactions.

In other words, proper berry interactions earn Shelter grabs nine times faster than improper interactions.

A little something for you to think about the next time you run out of grabs.

[BTS Jin]

This is very useful! Thanks for taking the time to investigate this. Right now I have 800 interactions, of which 30 are proper. Going to try this out, counting both normal interactions and proper interactions.

EDIT: Dang, forgot it was fog. Oh well, I'll just use a Pass Power and adjust my own minitest numbers...

EDIT2: All right, confirmed. I just got five more interactions. The following should result in x = 1, to confirm that the number 2943 is in fact the correct total improper interaction number to earn grabs.

My data right now, using different variables:

891 improper interactions --> 891x
228 proper interactions --> 228y

As you stated, 9 improper interactions (x) are worth 1 proper interaction (y). So:

9x = y

And also as you stated, for every 2943 improper interaction values, new grabs are gained. So, using that information with my own personal interaction data:

2943 = 891x + 228y

Substitute 228y for 228(9x) using the previously mentioned equation:

2943 = 891x + 228(9x)

Simplify:

2943 = 891x + 2052x
2943 = 2943x
x = 1

A bit overcomplicated, but just checking for myself that yes, 2943 improper interactions and 327 (or, 2943/3) proper interactions are the right numbers before earning more shelter grabs.

This post has been edited by Klaine: Jan 8 2012, 05:36 PM

[Erty]

okay i tend to read the full data before posting ut i'm in a kind of tl;dr mood today so i scanned it and i dont think you tested by doing solely egg warming to see if shelter grabs get increased, so either wait for me to ecome less lazy and find out if egg grabs (the perfect incorrect berry feeding method) increases shelter grabs or do it yourself?:P i'd love to know exact numers too kinda as this was kind of my next 'project' xD

irrelevant then ;p

This post has been edited by Erty: Jan 8 2012, 10:06 PM

[pumpkinking0192]

okay i tend to read the full data before posting ut i'm in a kind of tl;dr mood today so i scanned it and i dont think you tested by doing solely egg warming to see if shelter grabs get increased, so either wait for me to ecome less lazy and find out if egg grabs (the perfect incorrect berry feeding method) increases shelter grabs or do it yourself?:P i'd love to know exact numers too kinda as this was kind of my next 'project' xD

I did indeed. Days One and Two, I had the online list set to "Eggs" instead of my usual "Pokemon and Eggs". (I assumed, as most people assume, that egg interactions and wrong berry interactions would be lumped together equally as "improper", so I didn't test whether there was a difference between them.)

Since I have all 5 Shelter Passes, my daily grabs start at 26 and increase in multiples of 5. For two days, I clicked only eggs; for another two days, I clicked only proper berries. Since I wanted to rule out the possibility that the formula had changed to take activity score into account, the former of each pair of days was spent trying to get a high activity score (hatching eggs, buying items, etc) and the latter was spent doing little on the site except clicking.

[Tyranisaur]

This basically proves that proper berries count 9 more times, meaning y=8x. Also, if you set x=1, then you get to see that the number you divide by z is interactions + proper * 8. And since the number inside the floor has to be greater than or equal to 1 and less than 2 for you to get one new set of grabs, z seems to be 2943. That is assuming you never got any extra grabs with less than that value over the z.

This post has been edited by Tyranisaur: Jan 10 2012, 02:13 PM

[Erty]

This basically proves that proper berries count 9 more times, meaning y=8x. Also, if you set x=1, then you get to see that the number you divide by z is interactions + proper * 8. And since the number inside the floor has to be greater than or equal to 1 and less than 2 for you to get one new set of grabs, z seems to be 2943. That is assuming you never got any extra grabs with less than that value over the z.

someone should be able to see if this is the exact number by running a test (i'm kinda busy right now) but seeing how gpx have decided their other "secret variables" the number is probably 2900, 2950 or 3000

[BTS Jin]

This basically proves that proper berries count 9 more times, meaning y=8x. Also, if you set x=1, then you get to see that the number you divide by z is interactions + proper * 8. And since the number inside the floor has to be greater than or equal to 1 and less than 2 for you to get one new set of grabs, z seems to be 2943. That is assuming you never got any extra grabs with less than that value over the z.

someone should be able to see if this is the exact number by running a test (i'm kinda busy right now) but seeing how gpx have decided their other "secret variables" the number is probably 2900, 2950 or 3000

In my post above, I did test this number. I carefully used both proper and improper interactions bit by bit until I gained new grabs, and came to the same conclusion that pumpkinking0192 did with his proper- and improper-only tests at the exact number 2943. Which is a bit weird, but it's the right number.

[NTrinnu2]

This is like math class all over again~ xD
NICE!!

[Barney Stinson]

Very helpful
Now, when i'll need another grabs i'll know exactly what to do.

[Cherry Chain]

Wow, nice 0.0

I'd always wondered why I'd randomly get extra shelter grabs sometimes (I've been on a very very long Hiatus >.>; )

[Keldeo Knight]

wow, it 's very useful for me
thanks pumpkinking0192 for ur investigation

[Yamper]

Welp I always assumed it was every 5k activity score or something haha.

But thanks for clearing this up, pumpkin will definitely be handy

This post has been edited by Samoo: Jan 11 2012, 08:03 AM

[Erty]

This basically proves that proper berries count 9 more times, meaning y=8x. Also, if you set x=1, then you get to see that the number you divide by z is interactions + proper * 8. And since the number inside the floor has to be greater than or equal to 1 and less than 2 for you to get one new set of grabs, z seems to be 2943. That is assuming you never got any extra grabs with less than that value over the z.

someone should be able to see if this is the exact number by running a test (i'm kinda busy right now) but seeing how gpx have decided their other "secret variables" the number is probably 2900, 2950 or 3000

In my post above, I did test this number. I carefully used both proper and improper interactions bit by bit until I gained new grabs, and came to the same conclusion that pumpkinking0192 did with his proper- and improper-only tests at the exact number 2943. Which is a bit weird, but it's the right number.

i just did only incorrect berry feeds and fed 1 egg every minute at the end and at 2942 i had 26 shelter grabs while i had 31 at 2943 so 2943 is indeed the correct number ^^

This post has been edited by Erty: Jan 11 2012, 11:13 AM