[David Herrick]

Here’s another week of data for Tell Me Why:

For April 1st, V = 8173 and C = 14.8. For April 2nd, V = 6031 and C = 10.9. For April 3rd, V = 4227 and C = 7.63. For April 4th, V = 2940 and C = 5.31. For April 5th, V = 1696 and C = 3.06. For April 6th, V = 3820 and C = 6.90. And for April 7th, V = 3826 and C = 6.91.

[David Herrick]

That’s a fascinating little fellow, Jung! But I’m having trouble visualizing the eating and drinking part.

[David Herrick]

Ah, I get it now; thanks.

Funny you should mention Star Wars, because earlier today I suddenly remembered something about that film that I hadn’t thought of in many, many years. After seeing it in the theater and being introduced to C3PO and R2D2, I assumed the name of the old guy was OB1. I wonder if anyone else thought that!

[David Herrick]

I’ve been analyzing the evolution of the early viewing rates for various videos in an attempt to find a correlation with long-term viewing rates, with the goal of being able to estimate the eventual steady-state rate for any new video. That is, I’d like to be able to say that you can take the number of views per day after x days and divide it by some number n to determine what will be the approximate long-term rate. It turns out that it doesn’t seem to be that simple.

I looked at nine MLT videos that were posted on YouTube in 2022 for which I took daily data for each of the first nine days. All of them by this point have probably come close to levelling out, so I used the numbers for this March as their long-term viewing rates. I can say that on average they dropped to ten times their equilibrium rates after about eight days, but there’s a tremendous amount of scatter: after eight days, Here Comes the Sun was down to a little over four times its current rate, whereas Tired of Waiting was still cruising along at over 33 times its current rate.

There also doesn’t appear to be any significant correlation between decay rate and popularity: the most-viewed videos don’t consistently drop off either faster or slower than the less-viewed ones. This confounds my attempt to use the early numbers to predict the later ones. It would be so cool to be able to say after the first week that a given video will probably level out at a certain number of views per day, but I don’t think I can get there. I guess we’ll just have to enjoy the ride without worrying too much about the destination.

[David Herrick]

You’ve absolutely nailed it, Jung!

I’m basically a loner and definitely not a club joiner. But after discovering MLT I was so overwhelmed with intense feelings of joy that I had to commune with similarly afflicted people as a form of therapy, because honestly I was so constantly giddy that it was hard to focus on my day-to-day life.

As you stated, Mona and Lisa keep supplying the joy on a regular and frequent basis, and remarkably always make the time to communicate with us individually. It’s just mind-blowing to have such a special relationship with people that we may never meet in person (fingers crossed, though).

The club members are a huge part of the experience too. I’ve learned so much and shared so much over the last four years with a fun and diverse group of people that otherwise would not have been a part of my life, and now I can’t imagine not having all you folks to exchange ideas with.

I plan to stay on top of this wave and ride it all the way to the shore, which I hope remains far off on the horizon for a long time to come.

[David Herrick]

Here are the milestone predictions for the next month or so:

Here Comes the Sun: 500,000 views on April 14th

Vincent: 300,000 views on April 14th

Close to You: 300,000 views on April 28th

And that’s it! But Johnny B. Goode and I’ll Be Back should both hit a million views in mid-May.

[David Herrick]

Time for the March report! All these numbers cover the period from March 2nd through March 31st, except for Any Other Day, which is for just the last week. Tell Me Why is excluded because it’s still evolving quickly, but some recent data for it are given in the previous post in this thread.

.

simultaneous views:

1) Please Mr. Postman / Wipe Out: C = 12.11 (+2.05)

2) While My Guitar Gently Weeps: C = 3.08 (+0.18)

3) Here Comes the Sun: C = 2.90 (+0.74)

4) Drive My Car (2012): C = 2.26 (+0.21)

5) When I’m Sixty-Four: C = 1.85 (+0.05)

6) I Saw Her Standing There: C = 1.76 (+0.04)

7) You Can’t Do That: C = 1.76 (+0.06)

8) If I Fell: C = 1.66 (0.00)

9) Nowhere Man: C = 1.62 (+0.25)

10) If You Raise Your Head: C = 1.47 (-0.61)

11) Day Tripper: C = 1.46 (-0.03)

12) Wish You Were Here: C = 1.45 (-0.07)

13) Twist and Shout (w/ Mike Sweeney): C = 1.32 (-0.27)

14) This Boy: C = 1.24 (-0.06)

15) Reacting to Our First Music Videos: C = 1.20 (-0.23)

16) Africa: C = 1.19 (-0.05)

17) Vincent: C = 1.19 (+0.03)

18) Lola: C = 1.11 (+0.10)

19) Till There Was You: C = 1.11 (+0.14)

20) Any Other Day: C = 0.99 (-0.87)

.

views per day:

1) Please Mr. Postman / Wipe Out: V = 4272 (+724)

2) Here Comes the Sun: V = 1194 (+307)

3) Drive My Car (2012): V = 1167 (+109)

4) While My Guitar Gently Weeps: V = 970 (+54)

5) When I’m Sixty-Four: V = 916 (+26)

6) You Can’t Do That: V = 808 (+28)

7) If I Fell: V = 752 (+1)

8) I Saw Her Standing There: V = 749 (+19)

9) Nowhere Man: V = 721 (+113)

10) Day Tripper: V = 636 (-13)

11) Till There Was You: V = 583 (+72)

12) Twist and Shout (w/ Mike Sweeney): V = 543 (-112)

13) Please Please Me: V = 542 (+37)

14) This Boy: V = 527 (-28)

15) Any Other Day: V = 516 (-459)

16) If You Raise Your Head: V = 488 (-204)

17) I’m Looking Through You: V = 441 (+27)

18) Wish You Were Here: V = 428 (-21)

19) In My Life: V = 421 (+48)

20) Vincent: V = 402 (+9)

.

A lot more increases than decreases this month, which is great to see! And the latest MLT originals are still hanging in there. There was a huge uptick percentagewise for Here Comes the Sun, which perhaps makes sense at this time of year in the northern hemisphere.

[David Herrick]

Here are some daily data for Tell Me Why:

For March 27th, V = 22936 and C = 41.4. For March 28th, V = 13671 and C = 24.7. For March 29th, V = 7037 and C = 12.7. For March 30th, V = 6582 and C = 11.9. And for March 31st, V = 6524 and C = 11.8.

Meanwhile, for the past week, Any Other Day is at V = 516 and C = 0.99, and Vinyl Announcement! is at V = 140 and C = 0.47.

[David Herrick]

That’s an excellent suggestion, David. I’ve never paid attention to the number of likes. I’ll have to start doing that and see if they have any predictive power.

I noticed that the viewing rate for Tell Me Why has actually increased over the last couple of days. The last time we got a significant delayed increase was for their previous Beatles cover, Here Comes the Sun. Since Beatles songs are always their most popular, that makes me think that if the viewing rate remains high enough for long enough, the algorithm recommends the song to more people, creating a second wave of new views.

[David Herrick]

I agree, Tim: human behavior is much too complicated to be predicted with simple mathematical models. In fact, that’s precisely why I never considered taking a sociology class!

[David Herrick]

Wow, Jeffery, I thought I had a singular talent for acronyms, but you are definitely a Professional Employment Enhancement Renderer!

[David Herrick]

Thanks, JP. I think I’d prefer a pompous but vacuous title like Numeric Engagement Analytics Technician. (That would be NEAT!)

[David Herrick]

I can picture it now…

https://www.youtube.com/watch?v=O4zT7VhpDcQ

[David Herrick]

That’s right, Tim. For videos that have had time to “settle down” I use the number of views in the past month. For new ones I start with daily numbers, and after a week or two I go to weekly numbers, and finally after a month or two I figure it’s accurate enough to use monthly numbers.

Great question about the race to a million! It’s probably not very accurate to use the current viewing rates to extrapolate even a few months into the future, but it’s the best I’ve got. So here are some extremely tentative predictions for MLT originals that could reach a million views within the next ten years:

Questionable: February 2028

Any Other Day: April 2028 (probably way too optimistic, since it’s still settling down)

If You Raise Your Head: May 2028 (also probably too optimistic)

Still a Friend of Mine: September 2028

I Bought Myself a Politician: April 2029

This Boy Is Mine: February 2030

So yeah, we need to find some way of supercharging the viewing rates so that these milestones aren’t so distant!

[David Herrick]

Sorry, Tim, there’s sort of a history there that I should probably explain.

This thread started a couple of years ago when someone wondered how many people are watching any given MLT video at any given moment. I came up with a way to estimate that based on the total number of views, and I called it S for “simultaneous views”.

Later I realized it would be more meaningful to base that stat on just fairly recent views rather than the entire history, so I redid the calculations that way and named the new number C for “current simultaneous views”, so as not to confuse it with the old way of doing it.