LOL! It has crossed my mind a few times...
but then a bunch of us would really be flaunting how great we are at wasting time..
🎪🎯🎻🥁💥🦆🦓🐘🐖🐑🍄🥨🤪
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Top 50 Posters Extrapolated, January 2020
- Thread starter chown33
- Start date
- Sort by reaction score
LOL! It has crossed my mind a few times...
but then a bunch of us would really be flaunting how great we are at wasting time..
🎪🎯🎻🥁💥🦆🦓🐘🐖🐑🍄🥨🤪
That would be like awarding people for drunk driving.
Personally, I don't even think that other forum should exist.
I added it to my ignore list. It's really not necessary. But I digress.
Well everyone ELSE is wrong and in need of eternally being corrected
That's a value per post you can't buy.
Hmmm... selling "likes" could be a new revenue source!
Same, was one of the main reasons I stopped posting on these forums as much as I use to. Love talking tech but good grief when political stuff gets posted on here it just goes right down the tubes. Way to many rude comments and when they don't stop publicly they take it to private.. then I have to block 20 or 30 accounts from the same person who is still on this forum.. blows my mind how they continue to allow this person to keep posting but I don't own the forum so best to do is stop posting and just read the tech sections. Definitely glad they don't reward those in the other place.
I'd looked into doing this once before, but the numbers were all over the place. This is because Likes are counted in some forums where Posts aren't. So any ratio isn't a real reflection of the other stated numbers for the user. For example, someone could have 5 counted posts and 1000 likes, but that won't be their real Likes/Posts ratio, because they have 3000 posts in an uncounted forum (such as PRSI or Political News).
On the up side, it's probably going to be more amusing than the Followers count, which almost never changes, and when it does, it's often by just a few. So if Followers is out and L/P Ratio is in, then we'll have a more amusing measure of how much time we're wasting. What's life without whimsy?
Then there won't be popcorn till the Heat Death of the Universe.
In the current period, Jaffa Cake gained 1 Anomalous post, while mad jew gained 5 Anomalous posts. If they both continue at that fictional anomalous rate, then Jaffa Cake will never surpass mad jew.
It will take until April 8265 for Jaffa Cake to surpass mad jew's current post count. But by then, mad jew will have added 5 times as many posts. This assumes both users are receiving posts from the same subspace anomaly (or sofa cushions).
For obvious reasons, the calculation for surpassing mad jew didn't take into account that mad jew's post count would change. It's just a constant in the script that actually produces the dates. I guess this proves the old programming adage: "Constants aren't. Variables won't."
I'll look into this later today, and post what I find.
I’m waiting with baited breath and a tub of popcorn for the good doctor. Although I though @Doctor Q was more of a pie man!
Supplemental Projections
In post #22 @Doctor Q asked:
An approximate answer to this question is provided below. Following that is a detailed explanation of how the answer was obtained.
Executive Summary: An Approximate Answer
First, @Apple fanboy will have to surpass @Go Cubs Go. This should happen about 13 weeks from now.
Second, @Apple fanboy is expected to surpass @Doctor Q about 30 weeks from now.
Detailed Explanation, with Photos and Annotations
Part 1: Surpassing Go Cubs Go
Rather than use numerical methods to answer the question, I decided to go Old School and do this using mechanical graphical methods.
That is, given a graph of the posts for the 3 users of interest, with an X axis representing time, use a straightedge and other mechanical drawing equipment to approximate the X-position where @Apple fanboy's extrapolated line crosses @Go Cubs Go's line, and then @Doctor Q's.
I figured as long as I was wasting time, I should be extravagant about it: Go big or go home.
Making the graph was straightforward. I simply tweaked one of the Detail graphs shown earlier in the thread, so it contained only the 3 users of interest, and a sufficient range of time to allow decent linear extrapolation. Here is that graph as a GIF:
This graph originated as a PDF, which has much better accuracy. I've attached it to this post, so interested readers can download it and repeat the experiment.
In the graph, notice that Go Cubs Go is nearly a horizontal line. This means that a very simple linear extrapolation will suffice to estimate this first crossover.
To do this extrapolation, I used dividers set to the vertical distance between Apple Fanboy (AFB) and Go Cubs Go (GCG), for the last date on the graph: 2020-01-05.
I then "flipped" the dividers so one point was below AFB's final graphed datum. I put a pencil mark here.
Next, I placed a drafting triangle so one edge was on the mark, and the perpendicular edge was parallel to the last vertical date line. I then struck a horizontal pencil line across until it intersected AFB's graphed line.
The number of weeks between then and now is how long it took AFB to increase his post-count by the nmber of posts now remaining between him and GCG.
Here's a photo of the printed graph, showing the dividers flipped down, and the upper edge of the triangle at that distance, perpendicular to the last vertical gridline in the graph.
Here's a detail showing the horizontal line and one of the divider points.
The detail photo above shows the crossover is around 13 weeks ago. So it should take AFB about 13 weeks to surpass GCG.
Part 2: Surpassing Doctor Q
Extrapolating for Doctor Q won't be as simple in the mechanics, but it might be easier to understand because it's easier to visualize.
Go Cubs Go was mechanically simple because her post rate is nearly zero, so her post count barely changes. That means Apple Fanboy is effectively approaching a stationary target.
Doctor Q is a moving target, not a stationary one. This means that by the time AFB has reached where DrQ's post count is now, DrQ will have increased his own post count, and thus be that much farther away.
Oh sure, I could probably write a program to iterate both counts using their respective post rates, or perhaps develop a simultaneous equation, but that's too conventional. Instead, I used the same graphical methods as above, making an actual linear extrapolation.
First I had to place an additional piece of paper so I could draw the extrapolated lines. I used the back of a previously printed graph, so in the following photos, you'll see some faint lines showing through that should be ignored.
Here's a photo showing all the constructed lines, along with some letter labels for significant points:
The first extrapolated line extends AFB's line graph into the future. It crosses GCG's extrapolated line at the point I've labeled C.
The next extrapolated line extends DrQ's line graph into the future. I bungled the first line, so the upper line is the one to use. It crosses AFB's extrapolated line at the point I've labeled B.
I've also expanded the dividers so they match the distance between A and B. This distance represents how long from now it will take AFB to surpass DrQ's post count, taking both their current post rates into account.
The next photo shows the dividers moved onto the graph so they extend over a countable number of weeks. Then it's simply a matter of counting the vertical gridlines to reach the estimate of when AFB will surpass DrQ.
That count is about 30. This places it after the next Statistics Day, which will occur in about 26 weeks.
I also checked whether the horizontal distance between A and C matched up with the 13-week approximation made above. It did.
The extra marks and lines on the drawing were some additional points of reference.
The large + mark with the half-hidden "37333" is an extrapolation of the top Y-axis mark from the printed graph.
The "25 wks" line was an earlier estimate for a different projection: how long it will take for AFB to reach DrQ's current post count of 37333.
Going Bigger
For thecurious intrepid, the numerical data to populate a spreadsheet and perform a linear interpolation can be extracted from the PDF file. It can be opened in a text editor, and the uncompressed PDF examined.
The polylines for each line graph are in Object 5, which starts with this:
Each polyline begins with 2 coordinates (X Y) and a PDF 'move' command (m). That's followed by a series of XY coordinates and 'lineto' commands (l). The end is an 'S' (stroke) command that draws the polyline assembled at that point.
The numbers themselves are PDF page coordinates, measured in points (72nds of an inch). The fractional digit is rounded from a double, starting from the raw data (integer counts).
Strip off the 'm' and 'l' commands, and the numbers should go into a spreadsheet just fine. Graph them and it should be clear which one belongs to whom.
In post #22 @Doctor Q asked:
How long until Apple fanboy is expected to pass my post count? When the time comes, I'd like to get a big tub of popcorn and watch it happen.
An approximate answer to this question is provided below. Following that is a detailed explanation of how the answer was obtained.
Executive Summary: An Approximate Answer
First, @Apple fanboy will have to surpass @Go Cubs Go. This should happen about 13 weeks from now.
Second, @Apple fanboy is expected to surpass @Doctor Q about 30 weeks from now.
Detailed Explanation, with Photos and Annotations
Part 1: Surpassing Go Cubs Go
Rather than use numerical methods to answer the question, I decided to go Old School and do this using mechanical graphical methods.
That is, given a graph of the posts for the 3 users of interest, with an X axis representing time, use a straightedge and other mechanical drawing equipment to approximate the X-position where @Apple fanboy's extrapolated line crosses @Go Cubs Go's line, and then @Doctor Q's.
I figured as long as I was wasting time, I should be extravagant about it: Go big or go home.
Making the graph was straightforward. I simply tweaked one of the Detail graphs shown earlier in the thread, so it contained only the 3 users of interest, and a sufficient range of time to allow decent linear extrapolation. Here is that graph as a GIF:
This graph originated as a PDF, which has much better accuracy. I've attached it to this post, so interested readers can download it and repeat the experiment.
In the graph, notice that Go Cubs Go is nearly a horizontal line. This means that a very simple linear extrapolation will suffice to estimate this first crossover.
To do this extrapolation, I used dividers set to the vertical distance between Apple Fanboy (AFB) and Go Cubs Go (GCG), for the last date on the graph: 2020-01-05.
I then "flipped" the dividers so one point was below AFB's final graphed datum. I put a pencil mark here.
Next, I placed a drafting triangle so one edge was on the mark, and the perpendicular edge was parallel to the last vertical date line. I then struck a horizontal pencil line across until it intersected AFB's graphed line.
The number of weeks between then and now is how long it took AFB to increase his post-count by the nmber of posts now remaining between him and GCG.
Here's a photo of the printed graph, showing the dividers flipped down, and the upper edge of the triangle at that distance, perpendicular to the last vertical gridline in the graph.
Here's a detail showing the horizontal line and one of the divider points.
The vertical distance downward of that intersection is identical to the vertical distance between AFB's current data point and GCG's current data point. That is, it represents the number of posts that AFB needs in order to match GCG's current post count. Call that number of posts P.
Going downwards from AFB's final graphed point then represents the same number of posts, P. We can then look backwards (leftwards) to see how long ago AFB had that post count. If we measure how many weeks ago that was (W), then it should take AFB the same number of weeks (W) to increase his post count by P, assuming a linear relationship.
The detail photo above shows the crossover is around 13 weeks ago. So it should take AFB about 13 weeks to surpass GCG.
Part 2: Surpassing Doctor Q
Extrapolating for Doctor Q won't be as simple in the mechanics, but it might be easier to understand because it's easier to visualize.
Go Cubs Go was mechanically simple because her post rate is nearly zero, so her post count barely changes. That means Apple Fanboy is effectively approaching a stationary target.
Doctor Q is a moving target, not a stationary one. This means that by the time AFB has reached where DrQ's post count is now, DrQ will have increased his own post count, and thus be that much farther away.
Oh sure, I could probably write a program to iterate both counts using their respective post rates, or perhaps develop a simultaneous equation, but that's too conventional. Instead, I used the same graphical methods as above, making an actual linear extrapolation.
First I had to place an additional piece of paper so I could draw the extrapolated lines. I used the back of a previously printed graph, so in the following photos, you'll see some faint lines showing through that should be ignored.
Here's a photo showing all the constructed lines, along with some letter labels for significant points:
The first extrapolated line extends AFB's line graph into the future. It crosses GCG's extrapolated line at the point I've labeled C.
The next extrapolated line extends DrQ's line graph into the future. I bungled the first line, so the upper line is the one to use. It crosses AFB's extrapolated line at the point I've labeled B.
I've also expanded the dividers so they match the distance between A and B. This distance represents how long from now it will take AFB to surpass DrQ's post count, taking both their current post rates into account.
The next photo shows the dividers moved onto the graph so they extend over a countable number of weeks. Then it's simply a matter of counting the vertical gridlines to reach the estimate of when AFB will surpass DrQ.
That count is about 30. This places it after the next Statistics Day, which will occur in about 26 weeks.
I also checked whether the horizontal distance between A and C matched up with the 13-week approximation made above. It did.
The extra marks and lines on the drawing were some additional points of reference.
The large + mark with the half-hidden "37333" is an extrapolation of the top Y-axis mark from the printed graph.
The "25 wks" line was an earlier estimate for a different projection: how long it will take for AFB to reach DrQ's current post count of 37333.
Going Bigger
For the
The polylines for each line graph are in Object 5, which starts with this:
Code:
5 0 obj
<< /Length 1809 >>
streamEach polyline begins with 2 coordinates (X Y) and a PDF 'move' command (m). That's followed by a series of XY coordinates and 'lineto' commands (l). The end is an 'S' (stroke) command that draws the polyline assembled at that point.
The numbers themselves are PDF page coordinates, measured in points (72nds of an inch). The fractional digit is rounded from a double, starting from the raw data (integer counts).
Strip off the 'm' and 'l' commands, and the numbers should go into a spreadsheet just fine. Graph them and it should be clear which one belongs to whom.
Bravo for doing this computation mechanically! I can tell you had fun.
I have a compass in my desk drawer, and I wish I had more occasion to use it for something, as you did here. I remember trying to see if I could divide an angle in thirds with a compass and straightedge, even though it's known to be impossible. The last time I used my compass was to draw a design of concentric circles around a point for a decoration I was making.
It's simplest to assume that a user will continue posting at their latest post rate, which isn't always the case. You were fortunate that Apple Fanboy and I have almost linear post growth. This mechanical computation would have been a lot harder (or simply a lot less accurate) if either of us had an accelerating or decelerating post rate.
On the negative side, 30 weeks seems a bit too soon. Using the algebraic solution, you'd compute
WIth the data from Table 1 you'd get
which is 335.21 / 7 = 47.89 weeks. Of course the data is 3 days old, so it's obsolete anyway!
I have a compass in my desk drawer, and I wish I had more occasion to use it for something, as you did here. I remember trying to see if I could divide an angle in thirds with a compass and straightedge, even though it's known to be impossible. The last time I used my compass was to draw a design of concentric circles around a point for a decoration I was making.
It's simplest to assume that a user will continue posting at their latest post rate, which isn't always the case. You were fortunate that Apple Fanboy and I have almost linear post growth. This mechanical computation would have been a lot harder (or simply a lot less accurate) if either of us had an accelerating or decelerating post rate.
On the negative side, 30 weeks seems a bit too soon. Using the algebraic solution, you'd compute
numberDays = ( user2posts - user1posts ) / ( user1rate - user2rate )
WIth the data from Table 1 you'd get
( 37330 - 34122 ) / ( 17.86 - 8.29 ) = 335.21 days
which is 335.21 / 7 = 47.89 weeks. Of course the data is 3 days old, so it's obsolete anyway!
Re two previous posts, strictly my opinion: you guys have gone off edge of coffee table. (not an insult)
It's fascinating even if your day job employers might call for butterfly nets and straitjackets.
It's fascinating even if your day job employers might call for butterfly nets and straitjackets.
Absolutely fascinating. Outstanding detail!
That’s a tremendous amount of effort and I can see you enjoyed doing it. Thank you.
[automerge]1578379238[/automerge]
Better put a note in your diary for 30 weeks time: buy popcorn!
Last edited:
Agreed. I really appreciate the work that goes into providing this information.
I may have almost signed up to go down the road of that video a few times when someone has asked me to convert some quilt block pattern to a custom measurement that doesn't suit the pattern's base unit layout. Like some 8x8 layout they want to end up with a finished side of 7", 9", 11"... rather than the more usual 16, 12 or 8" finish for such a pattern.
Block pattern conversion is not a serious problem unless a block has 30º/60º angles in it and they want the thing shrunk down to where 1/4" seam allowances become impractical (setting in the seams for diamond shaped pieces can be challenging enough with big pieces!), but then my notebook page would have ended up looking like the doodle-decorated pages in that video. Still, I'd have just been trying to come up with reasonable alternate seam allowances, not changing shapes of the unit components. A good example of a block that becomes difficult to mess with is the basic Carpenter's Wheel block.
But: there are patterns for variations on the Carpenter's Wheel that do use arrangements of squares and right triangles instead of squares and diamonds.
So yeah, I "might could have" a few pages of doodles that approximate stuff in that video!
I always decline to provide a pattern conversion for a miniature if I realize I'm not going to want to make a block from the converted pattern just to check my work, not least because I'd have to include notes on irregular seam allowances too. I sometimes end up suggesting they experiment for once with a glued collage instead of trying to sew the thing. Then I don't mind making them new cutting instructions.
Pro tip #1: yeah it's true, never use fabric shears to cut paper; the minerals or other impurities in paper can quickly dull or even nick the blades, even though most nicks are by way of the tool having been dropped. Paper scissors are worthless for cutting fabric as their blade angle is usually 20º and the fabric will curl up. Fabric shears typically will have a 25º and a 45º blade (which is why there are shears made for lefties). Quilters and other surface artists can end up with as many different types of scissors in their studios as chefs have knives in their kitchens... one of my favorites is something any of you who may be a trout fisherman might be familiar with: a fly-tying scissor with curved blades and one of them serrated. Perfect in lieu of embroidery scissors for working with certain custom-content threads.
But: there are patterns for variations on the Carpenter's Wheel that do use arrangements of squares and right triangles instead of squares and diamonds.
And similarly there are workarounds for people who don't even like to sew triangles on the bias (stretchy angles of the weave of a fabric). For instance you can lay two differently patterned or colored squares of fabric right sides together, draw a diagonal line across the one facing you, sew the pieces together 1/4" away from both sides of that line, then cut on the center line, open up the resulting pieces et voila, a couple squares each having 2 right triangles in it. To cut the starting squares to the needed size, you have to know what size you want your finished piece to measure, but only in terms of the square block and not having to consider measurement of the hypotenuse... and not having to sew the stretchy edges of the hypotenuse together directly, as one would have to do if cutting and sewing the triangular components.
So yeah, I "might could have" a few pages of doodles that approximate stuff in that video!
I always decline to provide a pattern conversion for a miniature if I realize I'm not going to want to make a block from the converted pattern just to check my work, not least because I'd have to include notes on irregular seam allowances too. I sometimes end up suggesting they experiment for once with a glued collage instead of trying to sew the thing. Then I don't mind making them new cutting instructions.
Pro tip #1: yeah it's true, never use fabric shears to cut paper; the minerals or other impurities in paper can quickly dull or even nick the blades, even though most nicks are by way of the tool having been dropped. Paper scissors are worthless for cutting fabric as their blade angle is usually 20º and the fabric will curl up. Fabric shears typically will have a 25º and a 45º blade (which is why there are shears made for lefties). Quilters and other surface artists can end up with as many different types of scissors in their studios as chefs have knives in their kitchens... one of my favorites is something any of you who may be a trout fisherman might be familiar with: a fly-tying scissor with curved blades and one of them serrated. Perfect in lieu of embroidery scissors for working with certain custom-content threads.
Last edited:
wow, I quit math when they started adding letters to it, **** find X*** I did, she was with another guy....