Linus Gabriel Sebastian & Linus Media Group / Linus Tech Tips - Narcissistic corporate shill YouTuber driving his media empire into the ground. KILL COUNT: 2

[mandatorylurk]

Awkward upgrade video for the guy in the Labs who does all the PSU reviews. Awkward not only because he seems like a tall quiet gay guy (just guessing by his rug he bought) who doesn't pretend to be stupid for the camera, but because you can feel like Linus has to acknowledge the old pet he said he was totally going to take care of forever and ever until the new shiny thing took away his attention. The way some of the other people have gotten upgrades before they left, almost as a conciliatory thing for being laid off, wonder if the same thing is in play here.

 

[HootersMcBoobies]

Awkward upgrade video for the guy in the Labs who does all the PSU reviews. Awkward not only because he seems like a tall quiet gay guy (just guessing by his rug he bought) who doesn't pretend to be stupid for the camera, but because you can feel like Linus has to acknowledge the old pet he said he was totally going to take care of forever and ever until the new shiny thing took away his attention. The way some of the other people have gotten upgrades before they left, almost as a conciliatory thing for being laid off, wonder if the same thing is in play here.

What the actual fuck is this math? Is Retard Tech Man 4'11" or is this dude 7 and the editor is just retarded? Because somehow an inch turned into a foot.

Also: Canadian and no centimeters.

 

[Homoshrexual2]

What the actual fuck is this math? Is Retard Tech Man 4'11" or is this dude 7 and the editor is just retarded? Because somehow an inch turned into a foot.

Also: Canadian and no centimeters.

It's called a joke you fucking retard.

 

[Hellwalker]

It may be too early to speculate but it seems that LMG are using thumbnails that are significantly less autistic than their previous ones and are, only slightly, reflective of their older thumbnails from the early 2010's. Maybe they realized that people are fucking sick of the Mr. Beast styled thumbnails?

 

[Susanna]

It may be too early to speculate but it seems that LMG are using thumbnails that are significantly less autistic than their previous ones and are, only slightly, reflective of their older thumbnails from the early 2010's. Maybe they realized that people are fucking sick of the Mr. Beast styled thumbnails?
View attachment 6743035View attachment 6743039

Algorithm changed. Everyone thought the exaggerated thumbnails looked retarded the whole time, but the algorithm liked them so if you wanted your videos to get promoted to people you had to follow suit.

 

[Boiling Oil]

JayZ2C had a surprise interview with the NXZT founder.

These tech youtuber x journalist crossover episodes are getting wildly popular. GN Steve cracked the code by mixing anti-corporation hatefuel with spergy PC hardware enthusiasts, an unexpected match that is generating a lot of drama and views.

The JayZ2C interview with NZXT CEO was pretty weird to say the least. CEO claimed that Steve misled his audience in a video in a few different ways, first by claiming that he didn't receive the price increase email, which the CEO assured he did as he got an email reply and has the receipts to prove it (timestamp 15:25). Secondly, Steve claimed the URL to cancel the subscription was broken, but again CEO had receipts that they sent a working link and that somehow, an asterisk was added to the address bar after the URL causing the link to not work (timestamp 27:33).

Will Steve expose how this owner/CEO is lying and provide more proof? In the original GamersNexus video, I distinctly remember them not going into the cancellation fees.

This seems suspicious as fuck, if Steve and GN are straight up fabricating problems with the service to push their argument, especially "I never got the email" or breaking the link on purpose, its a disaster waiting to happen.

Steve uploaded a video tearing down all the lies(and bringing up that NZXT casually dropped that he used an AMEX card. He is now saying in both this video and in community posts that he had a new lead about more shady NZXT stuff that will take time to investigate

I didn't see anything in Steve's response video addressing the price increase or cancellation emails, it seems that only one clip from the interview was shown and it was the "leak" that he uses amex. Jayz interview isn't even linked in the description, but given the timing and new lead i'm expecting to see more in the next video.

As much as I think rental PCs are an insane and scummy practice, Steve & Co. Vidya Hardware Journalism Inc. are maxing out both their hubris and viewcounts on these topics, they might end up a little too close to the sun.

 

[The Great Milenko]

As much as I think rental PCs are an insane and scummy practice, Steve & Co. Vidya Hardware Journalism Inc. are maxing out both their hubris and viewcounts on these topics, they might end up a little too close to the sun.

Steve does suffer from OCD autism. He has mentioned several times about temperature measurements related to specifications and % improvements. He has this innate desire to use Kelvin instead of Celsius and no one, but him, actually cares. But if you do use Celsius in your specifications then YOU ARE WRONG dick nigger! - Steve, probably. It's a retarded hill to die on.

 

[Spiral Architect]

Steve does suffer from OCD autism. He has mentioned several times about temperature measurements related to specifications and % improvements. He has this innate desire to use Kelvin instead of Celsius and no one, but him, actually cares. But if you do use Celsius in your specifications then YOU ARE WRONG dick nigger! - Steve, probably. It's a retarded hill to die on.

That sounds retarded. Can you link any examples where he does that?

 

[Jeff Q. Anime]

Steve does suffer from OCD autism. He has mentioned several times about temperature measurements related to specifications and % improvements. He has this innate desire to use Kelvin instead of Celsius and no one, but him, actually cares. But if you do use Celsius in your specifications then YOU ARE WRONG dick nigger! - Steve, probably. It's a retarded hill to die on.

hey, buddy, I care. percentage improvements are meaningless without reference points for maximum and minimum and they never supply what those are in marketing slides. what is 20% improvement in temperature? tell me, bitch! tell me!

It may be too early to speculate but it seems that LMG are using thumbnails that are significantly less autistic than their previous ones and are, only slightly, reflective of their older thumbnails from the early 2010's. Maybe they realized that people are fucking sick of the Mr. Beast styled thumbnails?
View attachment 6743035View attachment 6743039

they've openly been talking about a/b/c testing thumbnails for a while now. they have at least one or two videos about it. nobody's 'realizing' anything, they're making money over here, they're using whatever works best, and they now have tools to actually show what that is.

 

[Susanna]

Steve does suffer from OCD autism. He has mentioned several times about temperature measurements related to specifications and % improvements. He has this innate desire to use Kelvin instead of Celsius and no one, but him, actually cares. But if you do use Celsius in your specifications then YOU ARE WRONG dick nigger! - Steve, probably. It's a retarded hill to die on.

That argument is even more moronic than what he’s railing against. Your computer isn’t venting heat into absolute zero, it’s venting it into room temperature. So if an AIO claims 5% improvement over something else based strictly on the difference in Celsius, the comparison will be as if you’re venting into an ice cold room, which isn’t accurate but is a hell of a lot closer than venting into an absolute zero room.

 

[Kuritan Deplorable]

That argument is even more moronic than what he’s railing against. Your computer isn’t venting heat into absolute zero, it’s venting it into room temperature. So if an AIO claims 5% improvement over something else based strictly on the difference in Celsius, the comparison will be as if you’re venting into an ice cold room, which isn’t accurate but is a hell of a lot closer than venting into an absolute zero room.

Part of the instigator of his autism is they use those percentages in a void - they don't really talk about from what to what, so they're basically useless marketing fluff that often don't represent real performance, and use celcius or farhenheit, and set the 'zero' point as best benefits them as opposed to being convenient. IE, a 5% difference between load temps from ambient is different from being 5% less degrees celcius from zero, which is different from 5% less degrees in farhenheit towards '0' vs ambient at load temps. Unfortunately, his solution is accurate but useless, because 99% of people do not know what anything is in Kelvin, so a 5% kelvin decrease from absolute is technically correct but practically useless.

Really the solution isn't to do any of that shit, and just say "It went from running at 65c at load to 61c under [benchmark tool] load" and drop the percentages entirely. Its still not a super useful metric but at least its a concrete number and not a divination, and can be reasonably tested/reproduced.

 

[The Great Milenko]

Part of the instigator of his autism is they use those percentages in a void - they don't really talk about from what to what, so they're basically useless marketing fluff that often don't represent real performance, and use celcius or farhenheit, and set the 'zero' point as best benefits them as opposed to being convenient. IE, a 5% difference between load temps from ambient is different from being 5% less degrees celcius from zero, which is different from 5% less degrees in farhenheit towards '0' vs ambient at load temps. Unfortunately, his solution is accurate but useless, because 99% of people do not know what anything is in Kelvin, so a 5% kelvin decrease from absolute is technically correct but practically useless.

This is really what it is. Conditions must be the same for an accurate comparison. A room with a set temperature and then run the test to measure the increase. The same test must be run the same way for different products for an accurate comparison.

Specifically, since 1 °C = 1 K there is no issue. Celsius is simply offset from absolute zero, but it maps 1:1 to the Kelvin scale. % calculations of delta measurements will be different depending on Kelvin or Celsius, but you are an autistic CIA nigger if you care. No one cares about Kelvin outside of a limited disciplines where you want/need to know how much kinetic energy is involved.

Electronics in cumsoomer, automotive, medical, military, etc? No. No one cares. Pull up any semiconductor component datasheet. Temperatures will be in Celsius.

That sounds retarded. Can you link any examples where he does that?

It was an older video when he was talking to ThermalTake.

https://www.youtube.com/watch?app=desktop&v=vhkYcO1VxOk&t=0s

There was a video this year where he made quick off hand comment in the intro which I don't remember which video. But to paraphrase Steve, it is a hill he will continue to die on.

EDIT:
My original comment was related to the possibility of Steve flying to close to the sun. He has done good investigative work, but IMO he is not as smart as I perceive him to believe himself to be. It's about hammering on his own misinterpretation of details to push a narrative - to Boiling Oil's comment.

 

[HootersMcBoobies]

My original comment was related to the possibility of Steve flying to close to the sun. He has done good investigative work, but IMO he is not as smart as I perceive him to believe himself to be. It's about hammering on his own misinterpretation of details to push a narrative - to Boiling Oil's comment.

I do wonder about his hubiris too. So far he's only attacked people when they hit a real breaking point on being wrong to the point of misleading or borderline scammy. (LTT, Newegg, NZXT). He really should request for comments before publishing videos. Most journalists do and get nothing or boilerplate back and mention it. GN makes it a point to not request comment. Both are probably ethical so long as you state that you did/didn't, I guess?

He did meet with the Newegg guys, and honestly I still buy most of my gear from Newegg or eBay (cause fuck Amazon). I never buy refurb gear anymore though.

He also makes a point to issue corrections when they're wrong at the start of the next video. His whole staff/studio/process is autistically tuned to being correct. Him going lolcow would require him being wrong about something and not backing down on it. Or doing something that alienates all his Patreon supports and advertising partners (which he's probably more dependent on than YouTube for revenue) and makes it impossible to pay down the debt I'm sure he has on all that testing equipment.

 

[quaawaa]

they've openly been talking about a/b/c testing thumbnails for a while now. they have at least one or two videos about it. nobody's 'realizing' anything, they're making money over here, they're using whatever works best, and they now have tools to actually show what that is.

Whenever a Youtuber starts talking about a/b tests, ignore them. 99.99% of them are incapable of running properly controlled experiments and as such their conclusions are meaningless (the 0.01% being people like MrBeast). YouTube didn't even have a proper thumbnail testing framework until last year yet YouTubers had been "optimizing" their thumbnails for many years prior (in reality, they were essentially picking thumbnails at random).

Hyper-optimizing a thumbnail to get people to click is actually a bad thing if the video sucks because it increases bounce rate which is one of the key factors the algorithm uses to determine if a video is "good" or not.

Algorithm changed. Everyone thought the exaggerated thumbnails looked retarded the whole time, but the algorithm liked them so if you wanted your videos to get promoted to people you had to follow suit.

No, that's just YouTubers being dumb trend chasers. The core algorithm is really simple: it's just watch time (as a percentage of total video time) divided by number of impressions. The problem that most YouTubers have is that they make poor quality videos that viewers stop watching halfway through. not that the algorithm hates them.

Unless you're a conservative commentator on the shadowban list, your video not being recommended is your fault, not YouTube's. MrBeast said it best that whenever you hear someone complaining about the "algorithm", replace the word with "audience" to find the real issue with their videos.

 

[Administrator Flotsam]

I do wonder about his hubiris too. So far he's only attacked people when they hit a real breaking point on being wrong to the point of misleading or borderline scammy. (LTT, Newegg, NZXT). He really should request for comments before publishing videos.

I don't think he has enough hubris to hurt him. If his numbers are correct, its not hubris, its just speaking the truth. He doesn't seem to have a unquenchable hatred for anything to warrant his standards slipping either.
As for requesting comment, I think he's always informed with due time in all situations where it would've been practical. He'll know the risks that come with publishing a hit-piece on a huge corporation so I doubt he won't have that base covered. In the case of NZXT, there was nothing to comment on anyway; a scam is a scam.

He's safe as long as he sticks to good research, good numbers and facts and I think he's one of the few creators who are mentally stable and financially unmotivated enough to keep to that path.

 

[Jeff Q. Anime]

YouTube didn't even have a proper thumbnail testing framework

which is what the video was about, it showed their dashboard and sheeez mayne

 

[Flaming Dumpster]

He really should request for comments before publishing videos.

If I recall correctly, he stopped reaching out to companies after one of them (can't remember which at this point) tried to get ahead of the story and downplay issues rather than addressing them. Might've been the NZXT fire thing or the Gigabyte exploding PSUs.

 

[Administrator Flotsam]

they've openly been talking about a/b/c testing thumbnails for a while now. they have at least one or two videos about it. nobody's 'realizing' anything, they're making money over here, they're using whatever works best, and they now have tools to actually show what that is.

Whenever a Youtuber starts talking about a/b tests, ignore them. 99.99% of them are incapable of running properly controlled experiments and as such their conclusions are meaningless (the 0.01% being people like MrBeast). YouTube didn't even have a proper thumbnail testing framework until last year yet YouTubers had been "optimizing" their thumbnails for many years prior (in reality, they were essentially picking thumbnails at random).

This has been picking my brain for a long time and I'd like everyone on here's opinions on it. Dubiously useful tools like tubebuddy and vidiq have had a/b testing features since before yt brought it out as a feature but even then, please explain how a/b testing isn't absolute bullshit when you consider that:

1. Videos get most of their views in the first few days of and after them being published
2. Time is a major factor because more or less people will be active to click on videos depending on when they get on and off of school / work
3. Most videos have time-based relevance because the video topic will change in popularity across the testing period
4. Many people actually click on the same video with a different title or thumbnail a second time (after having already watched it the first time) thinking that its a second video on the same topic
5. In addition to time-based relevance, other creators will saturate the topic of a video by the time a second or third thumbnail is given to it

Is it really possible to account for all of these factors when testing thumbnails and titles? Seems to me like its a gambler's superstition that channel growth gurus have somehow made into a thing.

 

[quaawaa]

Is it really possible to account for all of these factors when testing thumbnails and titles? Seems to me like its a gambler's superstition that channel growth gurus have somehow made into a thing.

Yes, with the newish YouTube feature that segments the audience and shows different thumbnails to different people at the same time and ensures that each person only sees one thumbnail per video. Prior to that, it was only possible to change the thumbnail for everyone, which has so many confounding factors it's impossible to determine which thumbnail is better.

Note that YouTube only offers this feature for thumbnails, so YouTubers who are "A/B testing" titles are just "gambling".

YouTube's help even mentions this:

When using tools not native to YouTube, you may receive conflicting thumbnail test results. YouTube’s thumbnail experiments are true, concurrent A/B/C thumbnail tests. This means all thumbnail variations are shown to viewers at the same time. Many third-party tools run thumbnail tests sequentially, and thus may generate different results. These tools often optimize solely for click-through rate, which may determine a different winning thumbnail than measuring by watch time share. This may be due to videos with high overall engagement (strong click-through rate, watch time share, and other factors) are shown more frequently. We believe that evaluating winning thumbnails by watch time share will best support creators’ growth.

Not that YouTubers ever read that.

 

[#lurker]

A/B testing is cancer and one of the big reasons UI and navigation are in the dumpster these days. It's not a coincidence that things were easier to use 20+ years ago before these parasitical practices ran amok.