Hm...YT is basically using a computer to identify other potential computers or “bots”. So, let’s think like a computer to solve this.

Computers do not make emotional decisions. To start an algorithm, you need to have a common denominator that is not based on emotion +

A thread. https://twitter.com/btbtbtsssssss_/status/1277793307603750912

that common factor is the IP address. Every connected device has one. It all starts from that point.

The first consideration is how FAST are views coming from one ip.

The next consideration is how MANY views are coming from one ip. +

The only thing we don’t and won’t know is what is the trigger point for deletion, or simply put, how fast is too fast and how many is too many?

But we can make a VERY close estimate.

Let’s analyze it..how fast is too fast?
+

We know YT counts a view after 30 seconds, we also know the average length of a MV is approx. 3 min. A person would not normally watch a video back to back.

So, if you are looking for “abnormal behavior” any video repeats within a 6 to 10 min window would be suspicious +

and TOO fast and probably trigger the view deletion. The time frame is doubled to account for the problematic b2b window.

That means for a 3.5 min video you could safely watch it about 6x an hour per ip.
+

Let’s look at history...we had about 1.3M people at ON preview... and got 10M views in about a 1.5 hrs. Why?

Accounting for the people who joined during the first hour that watch rate falls within the “safe” window.
+

So the takeaway on how FAST? I would guess, 1 view every 2x the length of the video but a safe estimate is probably every 10 minutes to be safe.

Now, we advise ARMY of this EVERY cb. We tell people to watch other videos...the problem is back to the ip address. +

If you’re using two devices on one ip, you think it’s ok to rotate views.

Watch on one device and then watch on the other...but it’s NOT OK because each view is on the same ip and the computer is reading it as b2b views.
+

This problem compounds itself as the day goes on...

we get tired, we don’t search for the video as often, we get frustrated with views freezing and not rising...

so we watch the video even MORE on our various devices. And guess what, deletion.

I have done it myself.
+

Solution: let’s get specific...1 watch every 10 minutes per ip address and account. Most of us have access to a cell phone AND WiFi. That’s two ip addresses.

If we maintain that we could average 5M an hour, easy.

The second algorithmic concern is how MANY views from an ip. +

The issue is really the same as the how FAST, the ip address.

Realistically, one ip address should not be generating more than 5 or 6 views per hour. Using the same criteria listed above.

More than that and you’re in the deletion zone.
+

Remember, computers are emotionless and So is this algorithm... it is based on data (numbers) not emotion.

It doesn’t care how much we love BTS or that we’re a large fandom.

It cares about how many views are coming from your ip address..that’s it.
+

We have addressed some aspects of these factors but maybe not as explicitly as needed.

As simple as this sounds, all things being equal...we are just watching too much, too fast.

We break crap all the time and we’re breaking YT.

So let’s slow it down.

As a note:

Do I think there was manipulation and tricks going on with the other group/agency...yes.

But, right now, I’m concerned about how we can make this better for us because we are getting eaten alive by YT and we’re working our tails off for BTS. +

I don’t care about YT records, I think they have served their purpose for BTS.

But I care a lot about the lost streams that could help us chart better.

I also think we are the hardest working fandom out here, so this is not an insult to anyone.

My analysis may not be correct. But we gotta have these convos to figure this out.

We’re all disappointed that we’re not getting fair play. But we can’t be impulsive and emotional.

Ironically,they call us bots but we’re actually fighting a YT bot right now.

I think we can win.