WEBVTT

00:00:00.000 --> 00:00:07.760
[Intro Music]

00:00:07.997 --> 00:00:11.402
Hey! I'm Cam, and you
are watching Reality Check.

00:00:11.402 --> 00:00:18.272
That's right, 64% of people can't tell the
difference between PS4, XBox One, and PC.

00:00:18.272 --> 00:00:21.975
At least, that's according to a study
done by me.

00:00:21.975 --> 00:00:23.770
and, uh, some of you in fact.

00:00:23.770 --> 00:00:28.555
Here's the skinny: Over 2,700 people took
part in an online test

00:00:28.555 --> 00:00:32.830
which tasked participants to watch
3 different gameplay clips back-to-back,

00:00:32.830 --> 00:00:39.154
and then attempt to identify which was
PS4, which was Xbox One, and which was PC.

00:00:39.154 --> 00:00:42.539
The results were then collected
over the course of a few weeks.

00:00:42.546 --> 00:00:44.853
Okay, so a few disclaimers:

00:00:44.853 --> 00:00:48.995
Was the test perfect? No.
Was it scientific? Kind of.

00:00:49.039 --> 00:00:52.022
Should it be taken as gospel? Probably not

00:00:52.022 --> 00:00:56.628
But does it make an interesting point
about how we value graphics? Definitely.

00:00:56.628 --> 00:01:00.963
But, before we dive into the details of
the test, reveal the correct answer,

00:01:00.963 --> 00:01:04.847
and talk about what went well and what
I would maybe do differently next time,

00:01:04.847 --> 00:01:08.990
let's take a moment to recap and rewind.

00:01:08.990 --> 00:01:11.715
On Monday, the 16th of February,

00:01:11.715 --> 00:01:14.125
I ran an experiment in our GameSpot studio
in London.

00:01:14.125 --> 00:01:18.473
I got 7 volunteers to play 3 different 
games on 3 identical setups.

00:01:18.473 --> 00:01:24.289
Their task: to tell me which they thought
was PS4, PC, and Xbox One.

00:01:24.289 --> 00:01:27.886
All other ways of identifying each machine
were controlled for.

00:01:27.886 --> 00:01:31.903
All they had to base their decision on was
how the game looked: the graphics.

00:01:31.903 --> 00:01:34.830
To get the full lowdown of the test
and how people got on,

00:01:34.830 --> 00:01:36.879
watch that episode of Reality Check.

00:01:36.879 --> 00:01:40.281
The problem, though, was that the
test pool size of 7 people

00:01:40.281 --> 00:01:44.232
was WAY too small to draw any meaningful 
conclusions.

00:01:44.232 --> 00:01:49.346
So, I devised a more simple one that would
allow anyone, anywhere to take part.

00:01:49.346 --> 00:01:54.509
So, how did I do it? Well, I added 3 video
clips at the end of that previous episode,

00:01:54.509 --> 00:01:57.275
all consisting of Far Cry 4 gameplay.

00:01:57.275 --> 00:02:02.153
One clip was captured on PS4, one on Xbox
One, and one on our gaming PC.

00:02:02.153 --> 00:02:06.205
I asked those watching to vote on which
they thought was which in an online poll

00:02:06.205 --> 00:02:07.554
Simple stuff.

00:02:07.557 --> 00:02:10.260
Now obviously, web compression was
always gonna be a big problem.

00:02:10.260 --> 00:02:14.104
As a standard for the show I uploaded
the video to GameSpot and Youtube

00:02:14.124 --> 00:02:17.644
at 1080p resolution and 60 fps,

00:02:17.652 --> 00:02:20.561
and both were compressed using the
H.264 codec.

00:02:20.566 --> 00:02:24.266
Now this is fine for general viewing, but
there is always a degree of

00:02:24.365 --> 00:02:29.149
quality loss and artifacting, so it was
far from ideal for this test.

00:02:29.149 --> 00:02:32.876
So, in an attempt to combat this, 
I also uploaded all three clips

00:02:32.876 --> 00:02:38.049
at 1080p-60, packaged up as one movie,
uncompressed, to Google Drive,

00:02:38.049 --> 00:02:41.079
for anyone to download and
watch for themselves.

00:02:42.059 --> 00:02:47.238
Okay, time to reveal the correct answer:
Clip #1 was PS4.

00:02:47.238 --> 00:02:51.959
#2 was PC, and the final clip was XBox One

00:02:51.959 --> 00:02:57.298
This was the single most popular answer,
with 982 votes, or 36%.

00:02:57.298 --> 00:03:03.115
The other 64% of voters, though, did get
it wrong, either getting 1 or 0 correct.

00:03:03.115 --> 00:03:07.097
Interestingly though, if you shift the
focus of the test to see how people

00:03:07.097 --> 00:03:08.611
performed at picking out the PC

00:03:08.611 --> 00:03:12.008
from the other 2 consoles, then the
results actually get a lot better,

00:03:12.008 --> 00:03:16.250
with 58% of people correctly identifying
the non-console.

00:03:16.280 --> 00:03:20.421
That was made up of the 36% who got it
correct, plus 22% of people

00:03:20.490 --> 00:03:22.928
who just got the consoles
the wrong way around.

00:03:22.928 --> 00:03:24.788
As for the rest of the results,
here they are:

00:03:24.788 --> 00:03:29.456
391 people, or 14%, voted PS4/Xbox One/PC,

00:03:29.456 --> 00:03:33.734
311, or 11%, chose PC/Xbox One/PS4,

00:03:33.734 --> 00:03:39.003
291, or 11% again, thought it went
Xbox One/PS4/PC

00:03:39.003 --> 00:03:46.033
And finally, 104, or 5% of voters,
chose PC/PS4/and then another PC.

00:03:46.033 --> 00:03:49.410
The eagle eye amongst you will notice 
that that last one was wrong.

00:03:49.410 --> 00:03:53.905
The final option should have been
Xbox One, not a second PC.

00:03:53.905 --> 00:03:55.967
So how did that happen? Well,

00:03:55.967 --> 00:03:59.624
that's what we call a derp. (Otherwise 
known as human error.)

00:03:59.624 --> 00:04:03.480
It was very late and I was very sleepy
when I made that poll, simple as that.

00:04:03.480 --> 00:04:05.640
Whoops!
But it doesn't really matter, though,

00:04:05.640 --> 00:04:08.753
because it was an incorrect answer
either way.

00:04:08.753 --> 00:04:11.038
So, that was the experiment and 
the results,

00:04:11.038 --> 00:04:12.867
but like I said, it wasn't perfect.

00:04:12.867 --> 00:04:15.721
So it's time to look at some of the
comments you guys had

00:04:15.721 --> 00:04:18.061
but also think about how I might do this 
differently

00:04:18.061 --> 00:04:19.771
if I was gonna do the test again.

00:04:19.771 --> 00:04:24.239
One big issue a lot of people commented on
was my choice of clips in the experiment.

00:04:24.239 --> 00:04:28.803
Now I used 3 clips from the same early
stage in Far Cry 4.

00:04:28.803 --> 00:04:33.148
The areas, though, weren't identical,
but this was a deliberate choice.

00:04:33.148 --> 00:04:37.038
You see, I wasn't attempting to emulate
the artificial setup

00:04:37.038 --> 00:04:38.967
of a graphics comparison, rather,

00:04:39.006 --> 00:04:43.295
I was trying to create real-world gaming
conditions, if that makes sense.

00:04:43.295 --> 00:04:46.645
Sure, identical clips would have made it
slightly easier,

00:04:46.645 --> 00:04:49.375
as you could have compared identical 
textures like for like.

00:04:49.375 --> 00:04:53.804
But the way I see it is that texture
quality, dynamic lighting, and all other

00:04:53.804 --> 00:04:57.812
graphical settings are all consistent in a
whole game area.

00:04:57.812 --> 00:05:01.828
And, more to the point, if graphical
differences between platforms really

00:05:01.828 --> 00:05:05.367
weren't the fuss that a lot of people tend
to make over them,

00:05:05.367 --> 00:05:09.128
then we should be able to tell
under the conditions I set.

00:05:09.128 --> 00:05:12.721
That was my logic, anyway, but I'm willing
to be talked around on that one

00:05:12.721 --> 00:05:16.030
if I were to repeat this test.
So let me know what you think.

00:05:16.030 --> 00:05:18.377
The main issue for me, though, with these
results is that

00:05:18.377 --> 00:05:22.967
I have no way of knowing how many of those
who voted did so after watching the

00:05:22.967 --> 00:05:26.148
online version and how many actually
downloaded the clips.

00:05:26.148 --> 00:05:28.623
Downloaded votes are far more powerful

00:05:28.623 --> 00:05:30.843
as you can eliminate all that nasty
compression

00:05:30.843 --> 00:05:33.512
and give people a much fairer test.

00:05:33.512 --> 00:05:36.750
Next time I'd find a way so only
after actually downloading the file

00:05:36.750 --> 00:05:38.291
could you cast a vote.

00:05:38.291 --> 00:05:40.929
Then, there's the case of the gaming PC.

00:05:40.929 --> 00:05:44.496
The specs of the one I used are on screen
now, and I think it's fair to say that

00:05:44.496 --> 00:05:48.866
it makes for a reasonably powerful gaming
PC, probably to the higher end of average

00:05:48.866 --> 00:05:51.053
that most PC gamers have.

00:05:51.053 --> 00:05:54.360
Plus, Far Cry 4 was running at the preset
ultra settings.

00:05:54.360 --> 00:05:57.157
I'll admit, though, that it would've
perhaps been fairer

00:05:57.157 --> 00:05:59.918
to have the most powerful PC I could get 
my hands on.

00:05:59.918 --> 00:06:04.254
So next time I'd jam in a Titan X in there
alongside a better CPU.

00:06:04.254 --> 00:06:07.143
But despite these changes I
would make in the glory of hindsight,

00:06:07.143 --> 00:06:10.362
I still think these results are really
interesting.

00:06:10.362 --> 00:06:14.535
For me, this test simply underlines how
graphics are not the most important factor

00:06:14.535 --> 00:06:18.164
to consider if you're trying to decide
which setup is best for you.

00:06:18.164 --> 00:06:24.161
Ultimately, PS4, Xbox One, and a gaming PC
are fairly close in visual quality.

00:06:24.161 --> 00:06:28.433
Sure, you can line up labeled textures and
spot the highest resolution one

00:06:28.433 --> 00:06:29.398
in a graphics comparison,

00:06:29.398 --> 00:06:33.228
but when it actually comes down to sitting
on your ass and having a great time gaming

00:06:33.228 --> 00:06:37.423
your ability to tell, and therefore the
importance of those extra pixels

00:06:37.423 --> 00:06:39.255
or quality of lighting, or whatever—

00:06:39.255 --> 00:06:41.063
start to diminish.

00:06:41.063 --> 00:06:44.363
Other things like online services, the
ability to host servers,

00:06:44.363 --> 00:06:48.278
digital distribution platforms,
controllers, what your friends use—

00:06:48.278 --> 00:06:53.103
these are far more important when deciding
which is best for you.

00:06:53.103 --> 00:06:55.991
It's all about our own
individual preference.

00:06:55.991 --> 00:06:57.859
But that's just what I think.

00:06:57.859 --> 00:07:00.109
Do you guys agree 
or do you disagree?

00:07:00.109 --> 00:07:03.228
Or do you have any ideas about how I maybe
could have done the test differently?

00:07:03.228 --> 00:07:07.002
Let me know in the comments down below
or by hitting me up on Twitter @camfrazrob

00:07:07.002 --> 00:07:09.955
Thanks for watching.

00:07:09.955 --> 00:07:27.597
[Outro music]

00:07:28.760 --> 00:07:31.032
So, it's time to think about some of the
comments you guys had,

00:07:31.032 --> 00:07:34.921
but also what I might do differently—
that was squeaky!

00:07:34.921 --> 00:07:37.747
—comments you had, but also think about 
how I might do this differently,

00:07:37.747 --> 00:07:40.519
Oh there's my chair!

00:07:40.519 --> 00:07:43.274
but also consider what I might do
differently, if I was going

00:07:43.274 --> 00:07:44.494
to do this test again.

00:07:46.850 --> 00:07:52.280
[laughs from Cam and the cameraman]
How did that happen?

00:07:52.339 --> 00:07:56.289
If you guys agree or disagree, or if you 
have any other ideas about how I

00:07:57.507 --> 00:08:05.191
[electronic alarm sound]