How many messages does @griffpatch get daily?

ZZC12345

DifferentDance8 wrote:
ZZC12345 wrote:
DifferentDance8 wrote:
Jeffalo wrote:
DifferentDance8 wrote:
you would be doing some advanced math anyway
((message count today) - (message count yesterday))
is hopefully not advanced math…
I didn't mean advanced math to get the daily number. I meant advanced math to predict the stats for tomorrow today.
…or just use Desmos regressions.
?

https://www.desmos.com/calculator/fibjkjobej

chatroomy

Idk

Sonic_No_FNF

(messages of (Griffpatch) is (123459))

Do the math! [Maybe 32791 Learn some math!

Last edited by Sonic_No_FNF (Oct. 31, 2022 22:33:44)

Sonic_No_FNF

DifferentDance8 wrote:
I know that.[/quote wrote:

ISTILLMAKESTUFF

Chiroyce wrote:
Step 1. Go here https://api.scratch.mit.edu/users/griffpatch/messages/count - and note down the number somewhere, and call it number A. (for example: A might be 200)
Step 2. Wait 24 hours
Step 3. Go here https://api.scratch.mit.edu/users/griffpatch/messages/count - and note down the number somewhere, and call it number B. (for example: A might be 250)

Subtract number A from B (with the examples, that means 250-200=50), that should be a pretty good average on how many messages he gets in 24 hours. You can continue doing this for a week or more, then take the average for a more accurate number too.

Ya what if he checks it once a day?

SpeedBee411

Wow! This forum has become really popular!

And I had 7 messages today!

Last edited by SpeedBee411 (Sept. 28, 2022 22:03:38)

qloakonscratch

Let's assume a is the amount that he has today and the amount he will have tomorrow without checking his messages, or at least an approximation, is b. x and y are the values of his messages yesterday and today. The equation would be: b = a + (y-x)

Alright, so, for example, imagine yesterday I had 28 messages (x=28) and today, I have 98 messages (y=98). Now, 98-28=70. That's how many he got in 24 hours. Now, add that to a, which is 98, which is 98+70. That's approximately 168 messages tomorrow. Now you do the math to find out the answer to your question.

Sonic_No_FNF

qloakonscratch wrote:
Let's assume a is the amount that he has today and the amount he will have tomorrow without checking his messages, or at least an approximation, is b. x and y are the values of his messages yesterday and today. The equation would be: b = a + (y-x)

Alright, so, for example, imagine yesterday I had 28 messages (x=28) and today, I have 98 messages (y=98). Now, 98-28=70. That's how many he got in 24 hours. Now, add that to a, which is 98, which is 98+70. That's approximately 168 messages tomorrow. Now you do the math to find out the answer to your question.

I don't think it's that, But I'll give you that.

_nix

I'm kind of annoyed the so-far final answer for this post was defined based on a sample size of only two minutes, rather than a whole day let alone multiple days

—
Edit because I should have explained instead of just being negative! What I mean is, you're doing pretty much exactly what I'm suggesting—a thing called “extrapolating”. That means finding out what happened in a small period of time (like getting 9 messages in 2 minutes), then using multiplication to find out what would happen over many of that same period of time in a row (like 720 * 2 minutes = 1 day).

The problem is just that if the period you're extrapolating from—the amount of time you did your measurements over—is a lot smaller than what you're extrapolating to, you're risking it getting pretty inaccurate, pretty quickly. That's because when you do multiplication, you're assuming that every two minutes is going to be (on average) the same as the two minutes you started with.

So if it just so happened that those were a particularly busy two minutes (or they were two minutes where he got many fewer messages than what is really average!), then your whole extrapolation gets thrown off, because he doesn't have super busy or slow two-minutes all the time.

Another way to think of it is that you counted how many messages he got at about 22:30 in my timezone. That's about 9PM in a good portion of America, and past midnight in Europe! It's possible that he gets more messages in the daytime or evening than late at night (depending on what parts of the world his messages come from and how those people use Scratch). But if you only measure two minutes, or even two hours, your extrapolation will work as though that time is accurate for the whole 24 hour day.

Extrapolation is a good approach to this question, and it's super awesome that you figured out how to get it to work on your own! But you can make it a lot more accurate by starting with a bigger time period, and it's cool to have more data anyway, because then you can stick it all on a graph and realize more detailed patterns if you want, too

—

Proposal for this thread: let's write down how many messages griffpatch currently has every few hours (so we don't spam the discussion forums or anyone's messages…right…sure…). Then we can fit those values to some simple-ish graph, or just compute the overall average, and from there, extrapolate how many messages he will have tomorrow

And then guess and see how accurate we were. #math.

griffpatch's current messages: 3066082

Last edited by _nix (Nov. 1, 2022 11:09:43)

awesome-llama

We've got historical data too.

1,769,682
2021-09-30
https://scratch.mit.edu/discuss/post/5618465/ (of course there's already a topic about griffpatch's message count, there were estimates of 1-2k)

2,484,107
2022-04-02
@Robby_Blue
[site that isn't scratch]/channels/558340604627648539/558341218262843402/959471134292721684

2,657,446
2022-06-12
https://scratch.mit.edu/studios/6446649/comments/#comments-187631635

2,569,899
2022-10-01
https://scratch.mit.edu/discuss/post/6637537/

Something's wrong with the last two I listed, I doubt griffpatch checked his messages and had it grow that quickly to have a message count a little smaller than what it was a few months earlier. One is probably not correct.

Edit: It seems there are a few more in the wayback machine for https://api.scratch.mit.edu/users/griffpatch/messages/count
I'm going to grab them too.

Edit 2:

3,016,731
2022-10-14

2,942,778
2022-09-19

2,817,335
2022-08-20

2,414,958
2022-03-06

2,419,026
2022-02-18

1,352,585
2021-06-11

1,226,986
2020-11-08

Last edited by awesome-llama (Nov. 2, 2022 11:13:04)

_nix

Nice! I'll edit this post with more current counts (if I remember lol).

Current post time is ~10:50 my timezone (Atlantic), sync edits to that.

10:52 — 3,066,879

ZZC12345

_nix wrote:
Nice! I'll edit this post with more current counts (if I remember lol).

Current post time is ~10:50 my timezone (Atlantic), sync edits to that.

10:52 — 3,066,879

We can see the post time already :p

Anyways, 3,069,511.

Based on that, it's been 75,487,000ms = 1258.11667 minutes = 20.9686112 hours = ~21 hours. Therefore, based on this, Griffpatch gets 2632 per 21 hours works out to about 125 messages per hour = about 2 messages per minute = one message every 30 seconds. That seems a bit low, but oh well.

Last edited by ZZC12345 (Nov. 2, 2022 10:55:09)

Sonic_No_FNF

awesome-llama wrote:
We've got historical data too.

1,769,682
2021-09-30
https://scratch.mit.edu/discuss/post/5618465/ (of course there's already a topic about griffpatch's message count, there were estimates of 1-2k)

2,484,107
2022-04-02
@Robby_Blue
[site that isn't scratch]/channels/558340604627648539/558341218262843402/959471134292721684

2,657,446
2022-06-12
https://scratch.mit.edu/studios/6446649/comments/#comments-187631635

2,569,899
2022-10-01
https://scratch.mit.edu/discuss/post/6637537/

Something's wrong with the last two I listed, I doubt griffpatch checked his messages and had it grow that quickly to have a message count a little smaller than what it was a few months earlier. One is probably not correct.

Edit: It seems there are a few more in the wayback machine for https://api.scratch.mit.edu/users/griffpatch/messages/count
I'm going to grab them too.

Edit 2:

3,016,731
2022-10-14

2,942,778
2022-09-19

2,817,335
2022-08-20

2,414,958
2022-03-06

2,419,026
2022-02-18

1,352,585
2021-06-11

1,226,986
2020-11-08

What are you talking about?

SpeedBee411

_nix wrote:
I'm kind of annoyed the so-far final answer for this post was defined based on a sample size of only two minutes, rather than a whole day let alone multiple days

—
Edit because I should have explained instead of just being negative! What I mean is, you're doing pretty much exactly what I'm suggesting—a thing called “extrapolating”. That means finding out what happened in a small period of time (like getting 9 messages in 2 minutes), then using multiplication to find out what would happen over many of that same period of time in a row (like 720 * 2 minutes = 1 day) . . .

At 4:27 PM EDT, he had 3,072,133 messages.

At 7:27 PM EDT, he had 3,072,979 messages.

3,072,979 - 3,072,133 = 846.
846 messages in 3 hours.
24 / 3 = 8.
846 * 8 = 6,768.

MORE ACCURATE ANSWER:
6,768 messages per day.

Last edited by SpeedBee411 (Nov. 2, 2022 23:37:37)

Sonic_No_FNF

SpeedBee411 wrote:
_nix wrote:
I'm kind of annoyed the so-far final answer for this post was defined based on a sample size of only two minutes, rather than a whole day let alone multiple days

—
Edit because I should have explained instead of just being negative! What I mean is, you're doing pretty much exactly what I'm suggesting—a thing called “extrapolating”. That means finding out what happened in a small period of time (like getting 9 messages in 2 minutes), then using multiplication to find out what would happen over many of that same period of time in a row (like 720 * 2 minutes = 1 day)…

At 4:27 PM EDT, he had 3,072,133 messages.

(will update in 3 hours)

Me personally would not take that.

the total is 27563.

ISTILLMAKESTUFF

Sonic_No_FNF wrote:
SpeedBee411 wrote:
_nix wrote:
I'm kind of annoyed the so-far final answer for this post was defined based on a sample size of only two minutes, rather than a whole day let alone multiple days

—
Edit because I should have explained instead of just being negative! What I mean is, you're doing pretty much exactly what I'm suggesting—a thing called “extrapolating”. That means finding out what happened in a small period of time (like getting 9 messages in 2 minutes), then using multiplication to find out what would happen over many of that same period of time in a row (like 720 * 2 minutes = 1 day)…

At 4:27 PM EDT, he had 3,072,133 messages.

(will update in 3 hours)
Me personally would not take that.

?

Sonic_No_FNF

ISTILLMAKESTUFF wrote:
Sonic_No_FNF wrote:
SpeedBee411 wrote:
_nix wrote:
I'm kind of annoyed the so-far final answer for this post was defined based on a sample size of only two minutes, rather than a whole day let alone multiple days

—
Edit because I should have explained instead of just being negative! What I mean is, you're doing pretty much exactly what I'm suggesting—a thing called “extrapolating”. That means finding out what happened in a small period of time (like getting 9 messages in 2 minutes), then using multiplication to find out what would happen over many of that same period of time in a row (like 720 * 2 minutes = 1 day)…

At 4:27 PM EDT, he had 3,072,133 messages.

(will update in 3 hours)
Me personally would not take that.
?

Why are you doing that? I don't know why you put just “?”.

ISTILLMAKESTUFF

Sonic_No_FNF wrote:
ISTILLMAKESTUFF wrote:
Sonic_No_FNF wrote:
SpeedBee411 wrote:
_nix wrote:
I'm kind of annoyed the so-far final answer for this post was defined based on a sample size of only two minutes, rather than a whole day let alone multiple days

—
Edit because I should have explained instead of just being negative! What I mean is, you're doing pretty much exactly what I'm suggesting—a thing called “extrapolating”. That means finding out what happened in a small period of time (like getting 9 messages in 2 minutes), then using multiplication to find out what would happen over many of that same period of time in a row (like 720 * 2 minutes = 1 day)…

At 4:27 PM EDT, he had 3,072,133 messages.

(will update in 3 hours)
Me personally would not take that.
?
Why are you doing that? I don't know why you put just “?”.

What do you mean by “me personally would not take that”? Do you not want to take the information that @SpeedBee posted?

Sonic_No_FNF

ISTILLMAKESTUFF wrote:
Sonic_No_FNF wrote:
ISTILLMAKESTUFF wrote:
Sonic_No_FNF wrote:
SpeedBee411 wrote:
_nix wrote:
I'm kind of annoyed the so-far final answer for this post was defined based on a sample size of only two minutes, rather than a whole day let alone multiple days

—
Edit because I should have explained instead of just being negative! What I mean is, you're doing pretty much exactly what I'm suggesting—a thing called “extrapolating”. That means finding out what happened in a small period of time (like getting 9 messages in 2 minutes), then using multiplication to find out what would happen over many of that same period of time in a row (like 720 * 2 minutes = 1 day)…

At 4:27 PM EDT, he had 3,072,133 messages.

(will update in 3 hours)
Me personally would not take that.
?
Why are you doing that? I don't know why you put just “?”.
What do you mean by “me personally would not take that”? Do you not want to take the information that @SpeedBee posted?

NO!!? Plus, I don't know how many messages @griffpatch has now dummy! ):I

ISTILLMAKESTUFF

Sonic_No_FNF wrote:
ISTILLMAKESTUFF wrote:
Sonic_No_FNF wrote:
ISTILLMAKESTUFF wrote:
Sonic_No_FNF wrote:
SpeedBee411 wrote:
snip-
Me personally would not take that.
?
Why are you doing that? I don't know why you put just “?”.
What do you mean by “me personally would not take that”? Do you not want to take the information that @SpeedBee posted?
NO!!? Plus, I don't know how many messages @griffpatch has now dummy! ):I

(#35)

Sonic_No_FNF wrote:
SpeedBee411 wrote:
_nix wrote:
I'm kind of annoyed the so-far final answer for this post was defined based on a sample size of only two minutes, rather than a whole day let alone multiple days

—
Edit because I should have explained instead of just being negative! What I mean is, you're doing pretty much exactly what I'm suggesting—a thing called “extrapolating”. That means finding out what happened in a small period of time (like getting 9 messages in 2 minutes), then using multiplication to find out what would happen over many of that same period of time in a row (like 720 * 2 minutes = 1 day)…

At 4:27 PM EDT, he had 3,072,133 messages.

(will update in 3 hours)
Me personally would not take that. the total is 27563.

Last edited by ISTILLMAKESTUFF (Nov. 2, 2022 22:05:48)

Discuss Scratch