Fake benchmark channels.

[Dobke]

May call this a conspiracy theory if you want. But I can't help but notice that there are a lot of benchmark channels out there, mostly smaller ones, that I'm pretty sure are faking gameplay benchmarks. Lots of people that are in the looks to buy new hardware wants to look for a specific combination that it's not widely available on the bigger and trustworthy channels out there, and if you keep digging you find some smaller channels that might have that combination of hardware you're looking for that suits what you plan on doing.

I'm not saying all small channels do this, but I'm pretty sure a lot of smaller channels does.

How would I be skeptical that they're faking benchmark gameplay? They hide some very vital information for no reason at all, when looking around to see what difference would make if I had a 13900K instead of a 12900K on my system, I found some questionable videos, such as in this video for example: https://www.youtube.com/watch?v=t2ilkAJ9aoI

In this video, the uploader decided to hide core count, power draw, clock speeds, important information that one would have no reason to omit, unless, they're not using the REAL CPUs, or it's the same CPU just slightly overclocked, in the case of this video, I assume an overclock gone wrong, as his claimed "13900K" under performs the "12900K "in almost every game. Also notable that both CPUs run pretty much same usage, same temperature, which is quite sketchy, considering the 13900K has a significantly higher TDP.

Anyway, just being paranoid here, I don't take any benchmark videos that hide important information seriously. This is not exclusive to CPUs, some GPU channels with do the same thing, hide clock speeds, power draw, etc.

Now why would one do this? Views, why not, and if they can monetize on it, that's an extra. Having all these expensive hardware is well, expensive, so why not fake it? Most people are going to buy it anyway.

[hawkeye]

That's definitely true. Make sure that you know how a pc works so that you can spot fakes.

[Dobke]

Origineel geplaatst door hawkeye:

That's definitely true. Make sure that you know how a pc works so that you can spot fakes.

Why don't you use your hawkeye and point out the truth then? If you can't point out where I'm faulting your comments will remain invalid.

[Illusion of Progress]

I would never be going specifically to YouTube to start a search for hardware performance. That's not to say you can't look at videos along with other things, but YouTube is merely a part of the web, so starting your search there is limiting it to there.

Yes, unfortunately many people aren't as thorough or aren't as willing to be educated on hardware, and instead just want the results (and honestly, even though I love learning about it at times, I don't blame anyone who doesn't want to and simply wants to know what's a better given option), and despite multiple sources existing and enough in depth analysis on specific hardware, nobody wants to educate themselves or spend time researching and wants an up front "A vs B" answer, and these videos arrive where that demand exists.

I might be "old fashioned" but I typically go to written review sites first. The old guard like Tom's Hardware, Anadtech, and so on (there's plenty of others so these aren't the only reliable ones but I'm just listing a couple of old ones), and then I'll also look at videos from sources like Gamer's Nexus (ditto, not the only one that's reliable) and others that lean more into the video format for their publishing, to get a broader picture.

If you just search some CPU plus GPU combination and get one of "those" head to head comparisons, well... you're going to maybe trade accuracy for that promptness.

In reality no single source will do enough testing to cover the entire software landscape. It's simply impossible. Tom's Hardware is one of the ones I know of that has one of the largest ones for CPUs and GPUs in games, but that's not to say they're the best either. Learning about hardware isn't easy and even a lot of regulars here and within other communities will regularly disagree (this can be healthy as it can show different views). You should always try and compile your total picture by looking at the total landscape.

But many people don't want to actually learn hardware, and that's fine. They just want to play games and find out what the better option is. And while "those" videos might not be "as accurate", they're often close enough. And if anyone actually just randomly takes any single first video they find and it happens to be a way off the mark one, and does no further research... well that's on them, too.

[Bad 💀 Motha]

You just have to use a little common sense and find the right channels. Small channels do not tend to have the ability to benchmark using high end hardware, while some of the bigger channels have said FU to people with lower end systems and simply not testing any of that stuff to show what the average low budget gamer might expect.

Instead of limiting yourself to YouTube videos there are plenty of sites that have had more time to test games on a wider range of hardware and give you break down comparisons with charts and other data. So look into those as well. I know people find it boring to READ through articles these days and just want things quickly explained and covered in a short video but if you really want the data, you need to look harder then juat a few YT videos

[Overseer]

Depends. Some people are also just incompetent or don't care. Major concern i always have with random channels, something that anybody should keep in mind, is the recording process itself. Some channels mention this others don't. But if you record the video on the machine you are currently benching then all you do is benching the recording. Only external capture gives the full picture.
Overall you get a good gamer experience on YT when you look for data. Just don't take it as scientific peer-reviewed fact.

[A&A]

It's very easy to get the information from other YouTubers and websites and try to recreate it, even if you don't have the right hardware it will look legit by making synthetic tests, but falsifying the results (like in this case) shows that the creator wants to make the fastest buck from the platform.