Home Creators Posts Import Register Favorites Logout

Content

I've been trying to figure out domino velocities. Look... I know it sounds trivial but good grief it can get difficult. Slippage on the bottom, variation on if the domino hits near the top or the bottom, twisting during impact, etc. My first test was inconclusive because it seemed like the dominoes accelerated in the first 20 or so. Obviously the only reasonable answer to this is brute force empirical scientific data collection. I've recorded 19 high speed videos. Recording each individual domino impact down to the nearest 100 microseconds is a tedious task. If you have a few extra minutes would you be willing to pitch in and count some camera frames?

Files

Crowd Sourced Domino Data Analysis: Part 3

Main page Wecome to the weirdest crowd sourced data project I' ve particiapted in to date. We' re going to figure out dominoes together. There are several high speed videos recorded of long lines of dominoes. Each video has a tab/ sheet below. Please pick a tab that hasn' t been completed yet,...

Comments

Jeff Dzado

Oooo... can't tell you how much I would love to do this (software engineer doing flight controls)....but work is busy and we've got a newborn at home!

Sharan

I'm unable to edit the spreadsheet

Patrick Ford

I filled in tab 2, but its' locked for saving.

Austin Burnham

For clarification: when the last red domino hits the first orange domino, we put this entry in the last red row of the spreadsheet, right? So in the column of the domino doing the hitting, not being hit.

Mark

I think so :) that's how Destin did it for Video 1

Luke Israelson

That's not what I expected to do tonight. Reminded me of data capture and analysis from physics class, oh the memories.

Everfalling

help. i'm editing #19 and it stopped letting me edit cells. i can edit cells on other tabs if i want but not on #19. i donno why,

Everfalling

yeah this is what i figure

Everfalling

also twist seems to go on the domino actively twisting before striking. so between red and orange if the red is twisting as a result of the red before it such that the twisting hits the orange then that last red gets a twist label.

Andrew Young

Well i at least got one done. Now it's time for bed.

Daniel Johnson

Sheet #9 needs the colors corrected, they don't quite match video #9. Also I'm noticing that a good number of dominoes made double impacts, the one in video 9 at frame 1839 for instance basically has two dominoes near-simultaneously pushing the next one.

Everfalling

also: i've been counting a strike as the frame before the next domino moves. i figure if i see any movement then the dominos have already touched and one is moving the other. so the frame right before should be the closest the previous domino is to the next one.

Everfalling

yeah i see what you mean. they're rattling back and forth. 1 will hit 2, 2 will hit 3, 2 will bounce back and hit 1 then bounce forward and hit 3 again right as 3 is hitting 4.

Daniel Johnson

Where it's clear I'm trying to note the frame where the movement of the preceding domino stopped, but they apparently refused to always line up on 0.1ms boundaries.

Patrick Ford

I did the same. As soon as there was perceptible movement, I stepped back one frame. So at least our observations will be consistent with each other. :)

Daniel Johnson

I didn't plan to spend nearly two hours looking at less than 1 second of footage, but it was fun. Let's do this again soon!

Steve

/r/Crowdsourcing/ there's a subreddit for this kind of project/request Destin. Could cross-post there & on your /r/SmarterEveryDay/ - would see this frame counting task done lickity-split (never underestimate the number of bored redditors). At this point I suggest posting/linking to a 2nd copy of the spreadsheet out there so you can compare the Patreon posted version with the Reddit posted one, and do secondary validation on any differences. That way you reduce the human error factor. The numbers need to be right for the statistical analysis to be right.

Steve

Also, good questions in the comments here around the "how" of the data gathering & recording. Be sure to note the particulars if you do post to Reddit. Just my two cents bud! :)

Mark

Looking back I think I misused the twist label a bit :p but I guess the frame number is more important anyway

Nathan Samchek

Not what I expected to do either but it was kinda fun. :)

Quester

I fully expected there to be many videos left to analyse at this point. This group of patreons is awesome!