LMG's obsession with misleading graphs

[NumLock21]

[Atmos]

Why is this a debate? 

LTT aren't attempting to show the exact full measurement of a given data point, that's not what they're using graphs for. they use them to visualize a difference in a quick, time efficient way. often the difference in bars on graphs with starting points at 0.0 in this particular context would be so small that the user wouldn't be able to determine a difference without pausing to study the graph, or without making the graph exceptionally long and narrow, only making it even harder to read from a normal sitting distance on a computer screen.

 

There is nothing wrong with their use of graphs, they are designed with a particular use in mind, and work in the given context. Incorrect scaling on the other hand is an entirely different point. 

[Guest]

5 hours ago, H0R53 said:

-snip-

lol are you running FRAPS while watching a YouTube video?

[mariushm]

Those product vs product value graphs with curves and annotations are super confusing to me and i consider myself a smart person

not to mention you really have to pause video on each slide if you wanna read text because they go by so fast... won't kill anyone if they last more time or make more sense.

[H0R53]

1 hour ago, Atmos said:

Why is this a debate? 

LTT aren't attempting to show the exact full measurement of a given data point, that's not what they're using graphs for. they use them to visualize a difference in a quick, time efficient way. often the difference in bars on graphs with starting points at 0.0 in this particular context would be so small that the user wouldn't be able to determine a difference without pausing to study the graph, or without making the graph exceptionally long and narrow, only making it even harder to read from a normal sitting distance on a computer screen.

 

There is nothing wrong with their use of graphs, they are designed with a particular use in mind, and work in the given context. Incorrect scaling on the other hand is an entirely different point. 

The graphs are poorly drawn and poorly organized, and require an actual look to understand. My IQ is above 135 (I have Asperger's so I'm a social idiot, but a genius in the workshop) and I still find the graphs annoying.

1 hour ago, NovaMan01 said:

lol are you running FRAPS while watching a YouTube video?

It's not on the video, so who cares? And I always have FRAPS active, it monitors any 3D application, that includes HTML5 video players. YouTube's player overlay is rendered in 3D, which is one reason Chrome uses so much memory.

57 minutes ago, mariushm said:

Those product vs product value graphs with curves and annotations are super confusing to me and i consider myself a smart person

not to mention you really have to pause video on each slide if you wanna read text because they go by so fast... won't kill anyone if they last more time or make more sense.

Point and case.

[Princess Luna]

1 hour ago, NovaMan01 said:

lol are you running FRAPS while watching a YouTube video?

Please when you're quoting a gigantic reply, can't you at least spoiler it so it doesn't pollute and fill the page all too quickly?

[GoodBytes]

1 minute ago, Princess Cadence said:

Please when you're quoting a gigantic reply, can't you at least spoiler it so it doesn't pollute and fill the page all too quickly?

I suggest to put "-snip-" (or something like that), like I have just edit @NovaMan01 post (sorry), like this the web browser dowesn't load once again the images for those on mobile devices, or slow connection, or slow device, because Spoilers still gets the images, just tells the web browser to hide the content.

[Imbellis]

6 hours ago, H0R53 said:

That's not the point. Even if the video's paused, the graphs are still misleading.

How are those bar-graphs misleading? None of them are truncated. I"m curious as to what is misleading.

As for those last four - Those (I'm guessing) are equations based on fps per dollar (Including the remainder of the computer's build). I do understand how those can be harder to understand.

Edited August 14, 2017 by Imbellis
Edited to specify bar-graphs instead of just graphs.

[Imbellis]

Now as for how they are organized - that could be improved. Personally, I'd like it more if they kept product names static instead of organizing them by highest to lowest. It works better with a lower number of comparisons but helps viewers read it better.

[EPENEX]

Because they're not smart enough to take a statistics class. I was taught statistics in high school, but it seems like most people aren't.

[vanished]

6 hours ago, H0R53 said:

Well truncated graphs are bad, mmkay?

No, they have a legitimate use-case.  When showing small changes is more important than showing comparative values, that's when you use it.

 

That said, I see where your'e coming from in the context of this thread; there's no place for them in a comparison between the performance of two products.

[AshleyAshes]

6 hours ago, H0R53 said:

 

I really can't help but look at the bottom graph and say 'Is this supposed to be evidence why UNtruncated graphs are bad?  cause that's what I feel when I look at this.'.

[mariushm]

3 minutes ago, Imbellis said:

How are those graphs misleading? None of them are truncated. I"m curious as to what is misleading.

As for those last four - Those (I'm guessing) are equations based on fps per dollar (Including the remainder of the computer's build). I do understand how those can be harder to understand.

 

They're stupid.

 

There's nothing to tell you how exactly those graphs are determined...

* did they factor in the cost of all the components in the system or just the card?

* did they price the electricity costs over one year assuming 6-8 hours of gaming a day or the costs are ignored .. 

* what if the 7700k requires some expensive board while the ryzen can do with a a320 chipset based motherboard to get same results ?

 

Look at this picture below...

 

What does the 100% even mean?

 

Why do you have to underline WOULD ... is it really some important word? Why is 56 written as ⁵⁶, just for the kicks? Why do you tell me that "crossing this line would mean blah blah" couldn't you just say "GTX 1070 is never a better value" and be done with it?

 

Do you really have to use THREE rows of text in a bubble when there's so much horizontal space there... just put an icon or something and put the text length wise, lots of space and it's easier to read and faster ... and way to go with the textured backgrounds which are a pain in the ass with video compression making text harder to read if watching the video at 480p or 720p ... just make the background of texts more opaque and more high contrast.

 

Who in the right mind would do a 3000$ computer with a single Vega56 or a gtx 1070 and what could a system with 1700x and a vega or gtx1070 could even have to affect this chart .. it's pointless long chart up to 3000$

 

where the f. are the specs of the systems or how are those curves produced?

This kind of graph doesn't help me make decisions about which card is better, it's pointless graph.

 

If you're afraid people will just stop watching the video if it's too long, just put the graphs with longer times AFTER the end ads and click to subscribe and all that crap.

 

 

[juri-han]

4 minutes ago, AshleyAshes said:

I really can't help but look at the bottom graph and say 'Is this supposed to be evidence why UNtruncated graphs are bad?  cause that's what I feel when I look at this.'.

i agree the first one is misleading true and unless you pause the video witch most people dont the quick cuts make it even more sketchy, but the second one is almost useless with out focuses in on numbers to the naked eye they look the same  

[AshleyAshes]

Just now, juri-han said:

i agree the first one is misleading true and unless you pause the video witch most people dont the quick cuts make it even more sketchy, but the second one is almost useless with out focuses in on numbers to the naked eye they look the same  

The first one if only misleading if you're too stupid to look at the actual numbers.

[juri-han]

10 minutes ago, AshleyAshes said:

The first one if only misleading if you're too stupid to look at the actual numbers.

like i said they show each graph for about 2 seconds and if you dont pause you literally cant read all the numbers, not that it matters every vega card is soldout even the bundles lol

 

my post was agree with you about the second example he made being useless but to call someone stupid for not trying to pause on quick cuts is being naive, if someone asked me for a good video of a review id link them one of the 100 other ones out there not one that does rapid fire quick cuts on misleading bar-graphs 

[Bcat00]

Lol graphs.... am i the only one that skips them in a video?

 

if i want to read a graph, i would go to a article. 

[Imbellis]

17 minutes ago, mariushm said:

 

They're stupid.

 

There's nothing to tell you how exactly those graphs are determined...

* did they factor in the cost of all the components in the system or just the card?

* did they price the electricity costs over one year assuming 6-8 hours of gaming a day or the costs are ignored .. 

* what if the 7700k requires some expensive board while the ryzen can do with a a320 chipset based motherboard to get same results ?

 

Look at this picture below...

 

What does the 100% even mean?

 

Why do you have to underline WOULD ... is it really some important word? Why is 56 written as ⁵⁶, just for the kicks? Why do you tell me that "crossing this line would mean blah blah" couldn't you just say "GTX 1070 is never a better value" and be done with it?

 

Do you really have to use THREE rows of text in a bubble when there's so much horizontal space there... just put an icon or something and put the text length wise, lots of space and it's easier to read and faster ... and way to go with the textured backgrounds which are a pain in the ass with video compression making text harder to read if watching the video at 480p or 720p ... just make the background of texts more opaque and more high contrast.

 

Who in the right mind would do a 3000$ computer with a single Vega56 or a gtx 1070 and what could a system with 1700x and a vega or gtx1070 could even have to affect this chart .. it's pointless long chart up to 3000$

 

where the f. are the specs of the systems or how are those curves produced?

This kind of graph doesn't help me make decisions about which card is better, it's pointless graph.

 

If you're afraid people will just stop watching the video if it's too long, just put the graphs with longer times AFTER the end ads and click to subscribe and all that crap.

This one is totally on me. I meant to specify bar graphs. I agree with you about the last four - Too many un-explained aspects(not accounted for) / not simple enough to flash by one every four seconds.

 

Note: I'm going to edit the previous post to specify it. Thanks.

[Imbellis]

19 minutes ago, juri-han said:

like i said they show each graph for about 2 seconds and if you dont pause you literally cant read all the numbers, not that it matters every vega card is soldout even the bundles lol

 

my post was agree with you about the second example he made being useless but to call someone stupid for not trying to pause on quick cuts is being naive, if someone asked me for a good video of a review id link them one of the 100 other ones out there not one that does rapid fire quick cuts on misleading bar-graphs 

Yet again - what aspect of the bar graphs are misleading?

[suits]

I guess whatever floats your boat. As a consumer you should know how to read graphs that are important to you, when I see a benchmark that I am actually interested in I pause and look at the numbers, everything else I want just a glance of what's best, which is where these graphs shine, if they aren't set up that way you can't tell the difference between anything. 

[juri-han]

3 hours ago, Imbellis said:

Yet again - what aspect of the bar graphs are misleading?

when they are done right nothing, when you keep them on screen and talk about the findings like say gamer nexus

[Godlygamer23]

It simply makes the differences more obvious, and if you keep within the context of the graph, you are left with a more accurate picture. For people just looking at the graphs without considering the data contained within, yeah it can be a problem, but people shouldn't be doing that, and should be ensuring they're getting all the proper information from a source before coming to a conclusion.

 

LMG has done nothing wrong here. I would argue that the second graph you showed is more misleading because it makes it seem like they're all the same when they're not. A truly accurate depiction of something won't try to hide subtle differences as your second graph tries to do.

[CoolMarquis97]

Isn't the simple answer just because the misleading graphs looks a lot more appealing and is able to show viewers the difference they want to see between products rather than having a more boring graph in which viewers are going to assume that there truly is no difference between the products. 

[LAwLz]

1) Nobody should watch LTT for information. They are constantly wrong and misleading.  I think it's sad that LTT still pretends to be a legitimate source of information, because I think they are borderline doing more harm than good these days. 

2) Yes, those graphs are misleading as fuck, especially those last few ones. 

3) There is no point in critizising Linus on this forum because everyone will just do world class mental gymnastics to justify it. 

[BingoFishy]

On 8/14/2017 at 3:18 PM, Atmos said:

Why is this a debate? 

LTT aren't attempting to show the exact full measurement of a given data point, that's not what they're using graphs for. they use them to visualize a difference in a quick, time efficient way. often the difference in bars on graphs with starting points at 0.0 in this particular context would be so small that the user wouldn't be able to determine a difference without pausing to study the graph, or without making the graph exceptionally long and narrow, only making it even harder to read from a normal sitting distance on a computer screen.

 

There is nothing wrong with their use of graphs, they are designed with a particular use in mind, and work in the given context. Incorrect scaling on the other hand is an entirely different point. 

That's OP's point. If the scores are so similar it's hard to tell, then it doesn't matter so much.