Good to know. I'm sad that you spoiled my joke though.
Posted via the Android Central App
Turbo Producing Low Benchmark Scores?
- Thread starter Murph5150
- Start date
Good to know. I'm sad that you spoiled my joke though.
Posted via the Android Central App
- Jul 31, 2010
- 1,163
- 31
- 0
- Jul 31, 2010
- 1,163
- 31
- 0
That's great. You've provided your opinion and ducked away from fact.
Some people actually like a visualization of phone speed just like a Corvette owner may dyno his car.
I'm degreed in both mechanical engineering and theoretical physics. I don't think I could do a 30 minute manipulation to produce falsified scores. If benchmarking was an irrelevant component, why do Geek Bench, AnTuTu, Vellamo and others have over a million downloads? I guess there are several million Android idiots because you seem to be the ultimate authority on Android etiquette. Shame on us all....
Posted via the Android Central App
Actually is not my opinion, it is fact (well maybe they would need a little more than 15 minutes to do it, but it's easily done )
Is great that you have your degrees. Could you perform brain surgery right now? Not successfully because it's a different skillset, so your degrees are completely irrelevant to this conversation (although they seem to make you feel good about yourself which is great )
The fact is its been proven over and over and over and over and over and over and over that benchmarks are extremely easy to manipulate. Samsung was even busted for doing this very thing to make their phones look faster. Just because you can't do it with your degrees doesn't mean it can't be done. That's actually a pretty ridiculous notion to think if you can't do it then it's not possible.
The reason those sites get so many downloads is that people like to brag and people don't know any better. I mean they even got someone with a theoretical degree in physics to post their useless results.
And while i appreciate the compliment, i assure you i am not the ultimate authority on anything. But what i do know is that on Android phones, benchmarks are totally, completely and in all ways useless.
And i never said anyone was an *****. Nobody knows everything, and people don't know benchmarks are useless until someone tells them.
And thanks for bringing up the car analogy to help prove my point. Is it not possible to rig a speedometer or radar gun to show false speeds? To someone with the right skills it would be very easy to do. Same thing with benchmarks.
Posted from the Avengers: Age of Droid Ultra
I guess everybody forgot about the anandtech.com report that Samsung and HTC are known benchmark cheaters: AnandTech | They're (Almost) All Dirty: The State of Cheating in Android Benchmarks
The Note 3 cheated on Geekbench, so it's possible that the S5 did, too. The HTC One M7 didn't cheat on Geekbench, but perhaps they did on the M8, and they certainly cheated on other benchmarks. Motorola and the Nexus devices were the only devices that didn't cheat on anything.
I wouldn't get too concerned about benchmarks. It's also possible, as hinted earlier, that the Turbo software was still being tweaked, running debug code, etc., at the time they benchmarks were run. I'd see how they benchmark after the device is actually released.
The Note 3 cheated on Geekbench, so it's possible that the S5 did, too. The HTC One M7 didn't cheat on Geekbench, but perhaps they did on the M8, and they certainly cheated on other benchmarks. Motorola and the Nexus devices were the only devices that didn't cheat on anything.
I wouldn't get too concerned about benchmarks. It's also possible, as hinted earlier, that the Turbo software was still being tweaked, running debug code, etc., at the time they benchmarks were run. I'd see how they benchmark after the device is actually released.
NoYankees44
Well-known member
- Dec 20, 2011
- 1,258
- 22
- 0
Samsung only truly cheated on their own processors by raising the clock speeds to levels only available in the recognized benchmark apps.
All the snapdragon powered devices were just put under idealized conditions(forced max clock). This is not cheating per say, but it is potentially running them at a different set of conditions than your competition. If everyone ran the benchmarks with these conditions, it would in no way be cheating.
Cheating implies completely faked numbers. That is definitely possible, but not what Samsung and others were doing.
I agree that benchmarks should be taken with a grain of salt, but they are not worthless metrics. You just have to clarify the conditions that the tests were run in. Just like any other quantitative gauge.
Sent from my XT1096
- Jul 31, 2010
- 1,163
- 31
- 0
I'm glad you knew I was being facetious because 1) cellphone review companies almost always post a benchmark. I can't count the number of benchmarks the DROID Turbo had floating on the Internet. So I suppose we should discredit them for endorsing the concept of running a benchmark.
AnTuTu has over ten million downloads and is rated at 4.5 stars. There's another ten million people we can add to the ongoing list of cellular retards.
You have clearly stated your position regarding benchmarks. I, however, provided quantitative evidence that proves benchmarking a phone is extraordinarily popular. Speaking of theoretical physics - Do you notice the similarity between "theory" and "theoretical?" Using involved, scientific evidence, typically pure mathematics, I turn a theory into fact. I could postulate a cunning argument but the bottom line is my argument is purely opinionated until evidence to such is produced.
So make no mistake. You have not been factual nor have you produced evidence to substantiate your claim. I respect your opinion, but I caution you to remember it is only an opinion and not something I'd go around preaching as Gospel. My attachment identifies over ten million AnTuTu downloads, and that's just one of many benchmark programs. If I did a summation of every benchmarking application and counted the downloads, I can confidently say there are more than 50 million benchmark application downloads.
If you have evidence to refute my evidence, I'll give it a thorough look. Until then, I'm gonna stick to my guns because I have the evidence and in the scientific community, a supposition, conjecture, postulation, or theory mean nothing if you lack the information and evidence to support your claim.
Posted via the Android Central App
Attachments
For a smart guy with such impressive sounding degrees you sure don't sound smart. At all. You throw out all these scientific sounding terms and explanation, but the root of your argument, or your "evidence" a you're trying to call it, is that benchmarks are popular so they must be right. I can't think of a more unscientific approach. You know what had far more downloads then your benchmark sites? Facebook. Does that mean any references on Facebook are right? Look a few posts back for your evidence. Someone posted a link to an article showing how Samsung manipulated benchmark scores. There are plenty more articles and videos of regular people doing the same. Sure, they're not throwing up mathematical formulas on a chalkboard, but they're still cheating benchmark scores. But i suppose your scientific mind can't get past the idea that if something is popular it must be right.
And what's with your insistence on calling people idiots and retards? That's really not cool, man. And it makes you sound even more uneducated.
I get it. You have degrees so you have to be right. But no one is right all the time. Some day you'll learn that. But until then, you're more than welcome to continue with your "its popular so it's right" belief.
Posted from the Avengers: Age of Droid Ultra
Nah, just a guy who understands that benchmarks are useless having a discussion with a guy with a couple of degrees.
Other than calling people idiots and retards there's nothing unnecessary going on here. But that part really was unnecessary wasn't it?
Posted from the Avengers: Age of Droid Ultra
So to recap what I have learned from the last 2 pages. Benchmarks may be one of the most important aspects of a device or not, depends who is reading or reporting them. Samsung and HTC may be no good cheaters on benchmarks or not. There are some people on here with very impressive degrees. I leaned more about the scientific theory and a little vocabulary like theory vs theoretical. Most important I learned that Facebook is more popular than AnTuTu so it must be totally factual. Then again everything you read on the internet is fact.
What I have really learned from this and a few other Turbo threads is people are starting to get antsy waiting for this phone to be released. Tempers are rising and patience is drawing thin. Motorola, Verizon, for God sakes release this phone already. Then we may finally achieve World Peace.
What I have really learned from this and a few other Turbo threads is people are starting to get antsy waiting for this phone to be released. Tempers are rising and patience is drawing thin. Motorola, Verizon, for God sakes release this phone already. Then we may finally achieve World Peace.
Last edited:
- Jul 31, 2010
- 1,163
- 31
- 0
1) You have some sort of attachment disorder with my degrees, and 2) you still have yet to show me evidence that supports your argument. I'll ask again - if benchmarks are insignificant, false, and misleading, why have they earned north of 50 million downloads? You can't answer that question because the popularity with benchmarks are overwhelming and your emphatic argument still rests with no validation.
I ask, "prove it." And you can't. You have a fascination with my education and are bent on getting to "just simply believe you" because that's convenient, right? Good luck finding an opposing, supportive argument when 50 million Android users felt the curiosity to see how their phone scores; for whatever the reason. I'll save you the effort. There isn't anything you can dig up, true or not, that will take presidence over the overwhelming majority. 1/6 of the American population has downloaded a benchmark app, and we have haven't grazed iOS yet.
Just maturely exit this dialogue and accept that my numbers are too overwhelming to beat and maybe, just maybe, you are fallible. And for the record, I worked hard for my degrees. The dedication as a student and bring awarded a Ph.D in Theoretical Physics is an honor, and I make a good living without the burden of having a boss on my case. I am my own boss. And that's something you can never take away from me.
Your turn, pal.
Posted via the Android Central App
Revolutionary
Well-known member
- May 12, 2010
- 876
- 21
- 0
- Jul 31, 2010
- 1,163
- 31
- 0
No argument. I'm 100% Irish and my colors fly at parties.
Posted via the Android Central App
vzwuser76
Well-known member
- Jan 28, 2011
- 1,669
- 26
- 0
Actually yes that is cheating. In a competition, doing something to enhance performance and change the outcome is cheating. When using any test as a metric to compare one or more things, they must be tested identically otherwise the results are worthless. Especially if you're testing one under ideal conditions and the other under real world conditions. Or giving one device a boost that it will never see under normal operation and running the other bone stock.
To me it's no different than in sports using steroids or other performance enhancing drugs. And they call that cheating. If there was a way to test all these devices and make sure these enhancements weren't happening, then they would carry more weight, but since no one can be sure if these are happening or not, the results can't be trusted.
Ugghh. Ok, buddy. Since you asked so nicely.
1) You're the one who brought up your degrees in a way that implied that if someone with your degrees couldn't manipulate benchmark scores, it couldn't be done. I keep referring to them to point out how preposterous it is to bring that up. While its quite an impressive feat, and something you should be proud of, any degrees you have has zero bearing on if someone can manipulate benchmarks.
2)I hate to tell you, but you haven't brought any evidence or proven anything at all. (I'm not even going to point out the enormously huge error in your argument thinking that the number is in America only when those numbers are worldwide. Its hard to win an argument when you don't even know what the data you are using represents. And even if it was America only, 1/6th isn't close to a majority, let alone an overwhelming majority. That's basic math, I'm seriously questioning how you can possibly have degrees in mechanical engineering and theoretical physics when you seemingly can't even do basic math. wow. And I know I said I wasn't gonna mention it, but I just did, so...whatever)You say its been downloaded north of 50 million times, but there is no data to support that. This particular app has been downloaded 10-50 million times. it could be as little as 11 million which isn't even close to north of 50 million, but there are other benchmark apps too, so even though you've given absolutely no evidence that it is "north of 50 million", I'll give you that number anyways. But guess what? There are over 1 billion android phones in the world. That means less than 5% of people are downloading benchmark apps. If you think that having 5% of users using something qualifies it as more useful than not, then I can't really help you. But even if it was 50% and not only 5%, what kind of science relies on the "It's popular so it has to be useful" theory? That's a serious question. I'd really like to know how you can claim to use a scientific method and then use that as your evidence.
Furthermore cheating at benchmarks is incredibly easy and there have been ways to do it forever:
How to cheat at benchmarks (and why we should downplay their importance) | Android Central
How to Cheat benchmarks in Quadrant using the HTC Hero � Hacks, Mods & Circuitry
Not to mention the post earlier in this thread showing that Samsung and others were cheating on benchmark tests.
There's my evidence, but lets not stop there. Lets say that all users everywhere used benchmarks and it was impossible to cheat. How are benchmarks useful? Here's a scenario. Tell me how a benchmark would answer my questions.
My wife and I are both looking to get new phones. She uses her phone to do Facebook, I use mine for gaming. If she gets the Z3 with its score of 2860, and I get the Note 4 with its 2925, will either of those serve our needs based on those scores? Which one of us will see better performance based on those scores? If I lose my phone and want to do some gaming, how will the game perform on her phone?
Or just on a more basic level, what do those numbers mean? If I buy a car(to use your analogy) if it says it gets 35 mpg, i know that with a gallon of gas I can go about 35 miles. If its max speed is 125 mph, i know that if i get on a straightaway with no traffic, i can go 125 miles in one hour. What does 2925 mean on the Note? Will it explode after 2925 texts? Will it stop working after 2925 minutes? Is it 2925 frames per second on video playback? Or is it just comparing something between phones? If so, what?
So yeah, I've just provided evidence that benchmarks are barely used, easily manipulated and the results don't really tell you anything.
Your turn.
Edit: And for the record, I never said I was infallible. I've been wrong plenty of times about plenty of things...just not this
what is this "parties" that you speak of? Is it the thing with "girls"
ooohhh, I wanna go
mobrules
Well-known member
- Jan 21, 2011
- 1,170
- 0
- 0
delrey1900
Well-known member
- Mar 22, 2010
- 751
- 0
- 0
Let me put it this way. I highly doubt I'll notice any speed difference between my '14 X and the Turbo based on how I use my phone. Probably the only difference I'll notice is how much longer the battery will last and I will probably feel the phone in my pocket. It's the same way you won't feel the difference between doing a quarter mile in 13 seconds to 13.2, but the 13 second car is faster. Woopy do!
Similar threads
Trending Posts
-
The Real-World Consequences of a Malware Attack on Your Business
- Started by bydasarinai
- Replies: 0
-
-
-
-
Facebook
Twitter
Instagram