Userbenchmark does more shady stuff

[leadeater]

Just now, mr moose said:

It does matter, because in an effort to avoid my own biases,  I like to get information from sources that don't piss me off.

Data is data, accuracy is king here.

[GDRRiley]

1 minute ago, leadeater said:

If we are going to cherry pick then lets just drive right in to the deep end, go big or go home.

The 4900HS is over 3 times faster than the i7-9750H, that's right you heard it here first!

they did mention right below it that for whatever reason the intel chip would only run at 60FPS

driving it in deep would be looking at encryption.

[leadeater]

1 minute ago, GDRRiley said:

they did mention right below it that for whatever reason the intel chip would only run at 60FPS

driving it in deep would be looking at encryption.

I dunno, I just call it how I see it 

[mr moose]

Just now, leadeater said:

Data is data, accuracy is king here.

And humans are notorious for changing how much credit they attribute to that data depending on the person presenting it.  We are all guilty and that's why I want to see the data from someone else.

[mr moose]

2 minutes ago, leadeater said:

I dunno, I just call it how I see it 

I make a conscious effort not to do that.  

[Guest]

1 minute ago, leadeater said:

I dunno, I just call it how I see it 

Anandtech, Gah are we really going there?

That's almost like saying user bench is the tell all reliable source for PC benchmarking. 

 

Why don't you people just go to HWBot, participate and fill the damn database. Then, we won't need reviews (Not you directly, the public) **Not directly an advertisement** but that would be the best end user case reliable source.... if the submissions happen to be there.

[GDRRiley]

7 minutes ago, leadeater said:

I dunno, I just call it how I see it 

this battery life

3 minutes ago, ShrimpBrime said:

Anandtech, Gah are we really going there?

yes they do good work

[leadeater]

2 minutes ago, GDRRiley said:

this battery life

I think that is potentially more amazing than the performance, historically AMD based laptops have been just so utterly trash at battery run time.

[Guest]

3 minutes ago, GDRRiley said:

 

 

yes they do good work

 

We should gather information from at least 3 sources to be sure the numbers all add up. 

 

 

[GDRRiley]

Just now, leadeater said:

I think that is potentially more amazing than the performance, historically AMD based laptops have been just so utterly trash at battery run time.

yep. 4000 can get 12h+ with a 75W battery even in web browsing tests. I can get about 5-6 on my 3500U with a 45W battery but I haven't run a very scientific test.

3X00U ryzen with a maxed out 99W hour battery wasn't bad at 12+ hours but 4000 looks like it maybe 16+

[leadeater]

9 minutes ago, ShrimpBrime said:

Why don't you people just go to HWBot, participate and fill the damn database. Then, we won't need reviews (Not you directly, the public) **Not directly an advertisement** but that would be the best end user case reliable source.... if the submissions happen to be there.

I will never trust user submitted benchmarks over a known person doing known things that I can analyze their methods. We can all disagree with peoples analysis and conclusions taken from data and test results but as long as the data is sound we can make our own.

 

Things like HWBot are good for filling in the void if there are no specific reviews of the product.

 

There is a lot more to standardized testing than just the configuration of the software test pass.

[GDRRiley]

1 minute ago, ShrimpBrime said:

We should gather information from at least 3 sources to be sure the numbers all add up.

Spoiler

 

 

[Guest]

Just now, leadeater said:

I will never trust user submitted benchmarks over a known person doing known things that I can analyze their methods. We can all disagree with peoples analysis and conclusions take from data and test results but as long as the data is sound we can make our own.

 

Things like HWBot are good for filling in the void if there are no specific reviews of the product.

 

There is a lot more to standardized testing than just the configuration of the software test path.

It's a laptop. There's nothing that would really change the data between 2 people's systems because they will always be exactly identical while a desktop, just a motherboard swap could make a world of difference. So I cannot agree there.

[GDRRiley]

1 minute ago, ShrimpBrime said:

It's a laptop. There's nothing that would really change the data between 2 people's systems because they will always be exactly identical while a desktop, just a motherboard swap could make a world of difference. So I cannot agree there.

have you seen what software can do to a system?

[leadeater]

1 minute ago, ShrimpBrime said:

It's a laptop. There's nothing that would really change the data between 2 people's systems because they will always be exactly identical while a desktop, just a motherboard swap could make a world of difference. So I cannot agree there.

Room temperature is already a huge one for laptops. And then are you testing the laptop or the CPU? Because different laptops with the same CPU perform differently. Laptops are far worse to test than desktop are, way more firmware customization and power profile tweaking happens.

 

If I wanted to skew the results I'd just put the laptop in a walk-in freezer for a good 30 minutes before doing the test.

[GDRRiley]

Just now, leadeater said:

If I wanted to skew the results I'd just put the laptop in a walk-in freezer for a good 30 minutes before doing the test.

you could also cool it with DICE and remove the stock cooler.

[Guest]

1 minute ago, GDRRiley said:

have you seen what software can do to a system?

You mean by overclocking? 

It's still all the exact same hardware. 

ANd if Ryzen mobile follows the same OC path as desktop..... you're really not making much gains past stock Boosts except additional power consumption..... which let's face it, this forum is not an OC forum and likely the OC on these laptops would be in smaller percentages. It's designed to power on and game.

[Guest]

1 minute ago, leadeater said:

Room temperature is already a huge one for laptops. And then are you testing the laptop or the CPU? Because different laptops with the same CPU perform differently. Laptops are far worse to test than desktop are, way more firmware customization and power profile tweaking happens.

 

If I wanted to skew the results I'd just put the laptop in a walk-in freezer for a good 30 minutes before doing the test.

My thermostat has a 4 degree swing. I fully understand your angle. 

 

So I'm talking about real world numbers not lab coated. There's a difference. I'd take your personal experience over a review in a heartbeat.

[GDRRiley]

1 minute ago, ShrimpBrime said:

You mean by overclocking? 

It's still all the exact same hardware. 

ANd if Ryzen mobile follows the same OC path as desktop..... you're really not making much gains past stock Boosts except additional power consumption..... which let's face it, this forum is not an OC forum and likely the OC on these laptops would be in smaller percentages. It's designed to power on and game.

no I mean by doing stuff in the background crippling the processor.

[leadeater]

1 minute ago, ShrimpBrime said:

So I'm talking about real world numbers not lab coated. There's a difference. I'd take your personal experience over a review in a heartbeat.

Well I wouldn't because your real world might be the outback in Aussie or the desert in the middle east and I'm in NZ in a much colder climate so I might have a vastly different experience.

 

Law of averages will always win out but a properly controlled test will give you the information much more quickly.

[Guest]

2 minutes ago, GDRRiley said:

no I mean by doing stuff in the background crippling the processor.

Well sure. There's always that chance dude was proning after gaming DLed some bad stuff. 

But with larger data numbers, you have better data. 

 

That's why when I look for information, I go to over clock forums quite a bit. The user experiences are different from reviews 90% of the time.

 

 

[Guest]

1 minute ago, leadeater said:

Well I wouldn't because your real world might be the outback in Aussie or the desert in the middle east and I'm in NZ in a much colder climate so I might have a vastly different experience.

 

Law of averages will always win out but a properly controlled test will give you the information much more quickly.

I agree. But those living in the Desert likely have air conditioning. At least I hope so. My father in law lives in Arizona. His house a cool 75f. 

Step outside it's like stepping into an oven. You're not using a PC in that weather. Definitely not in the direct sunlight. But that's reaching a bit I think. 

 

 

[leadeater]

1 minute ago, ShrimpBrime said:

You're not using a PC in that weather.

Then why have a laptop 

 

Really does come down to user submitted data is not quality data and requires large amount of data points for the averaging to work out properly, sample size is a big issue. If I want to know something now there is no substitute for controlled testing.

[GDRRiley]

8 minutes ago, ShrimpBrime said:

 You're not using a PC in that weather. Definitely not in the direct sunlight. But that's reaching a bit I think.

I live in California by the shore with no AC we do have days here for about a week or 2 a year were it is near 100F. there are people inland who I wouldn't be surrprised if it was 110 and they had no AC and would be using a laptop. hell my room is normally above 75F right now.

[Guest]

2 minutes ago, leadeater said:

Then why have a laptop 

 

Really does come down to user submitted data is not quality data and requires large amount of data points for the averaging to work out properly, sample size is a big issue. If I want to know something now there is no substitute for controlled testing.

But see you leave out data!! A laptop is.... a personal computer (PC) lol. 

 

I understand what your saying. It's good that people can review the data is close enough. Obviously reviews lack those strange scenarios where weather could be a factor.

 

One issue I have though with reviews. If we look at a review for something not yet released, it's on the original starting bios for the system, be it laptop or desktop. That right there is never taken into account when 9 months later, we use the same review for demonstration in a forum such as this to discuss accuracy. And then why I look to end user data. Imagine bios updates that increase memory compatibility performance and frequency. That changes quite a bit of numbers there. 

 

So in the end..... there is no such thing as "controlled testing" because the thermostat on the wall just kicked in and lower my ambient temps by 4 degree. But is 4 degree a game changer? nah. But that bios update might be.