What Sucks About HardOCP Video Card Reviews

Personally, while I do read, enjoy and am entertained by the reviews here, I find them next to useless from a buyers/consumer standpoint.

If I'm in the market for a GPU, I want to see how it stacks up against the competition. I want hard numbers/data taken in a controlled environment (AKA, all settings used are the same across the board) because ultimately, that tells me exactly what I need to know. When settings are adjusted across tests to hit a target framerate, that really doesn't tell me anything. Thats what I, as an enthusiast, already do to to make a game perform to what I consider an acceptable level.

For an example, I just built a 4K gaming box. I wanted to see how various GPU's stood against each other rendering games at 2160p. [H]'s reviews wouldn't of really helped me with that unfortunately. I wanted the BEST performing GPU at the resolution (all else being equal), and I'd adjust settings from there to get the framerates I wanted.

Would it not stand that the best performing card at any resolution is the fastest card on the market? I mean if all you are looking for is the best card, regardless of price, would you not naturally just automatically look at the Titan X? What need is there for a review when there is no card currently on the market that beats it?
 
The main thing is that I would like more cards compared - many times a card is just compared against one other card, or 4 total if you include OC'd variants of the same card. I find reviews to be most useful when you can compare results against a dozen or more cards. I'm not just talking about new cards, but older cards also, going back a few generations if possible. People buying a new card are usually upgrading from an older card, and being able to compare the performance of the newer card against the older card that they have is VERY valuable.

The 2nd thing is games. I would like to see a bit of diversity when it comes to games, including older games. Not everyone only plays the very latest games, and there is this misconception that only the latest and greatest games can put a new videocard to use. Plenty of "older" games can push a newer videocard when settings are maxed. I'd like to see games like World of Warcraft on there also. It's still the most popular MMO, and even though it's an 11+ year old game now, dramatic graphical improvements to the game have come with every single expansion, and this year will see the 6th expansion. It can work a card pretty hard on max settings, and there are very competitive aspects to the game which directly benefit from good computer performance.

Overall I'm also just not a huge fan of the apples to oranges review style. I do find it valuable to be told the particularities of how a card performs in relation to other cards, and when the apples to apples numbers should be taken in context for various reasons. At the same time, I don't necessarily feel that it should be the focus of the review unless those particularities are so extreme as to actually warrant that.
 
Last edited:
I play games at 1200p.

I would love to see reviews include stats for what max resolution can still attain playable settings with detail turned all the way up.

Using the max resolution playable as the baseline and then turning up features until unplayable doesn't really do it for me as the higher res tests are always way over what I play at.
 
  • Like
Reactions: raz-0
like this
I really like the reviews, the only thing I wish I saw was a comparison from red to green and vs versa. For example, in the MSI Golden review, throwing in a Fury X and Fury X OC would be interesting. Also some 4k reviews would be cool. Not saying you need a Fury X and 4k on every single game, more like "hey this is the FOTM game right now, this is what it would look like at 4k with these cards, this is what it would look like vs a Fury X". Otherwise, I really don't have much complaints at all. Keep up the good work.
 
On thing I'd really like to see is the Apples to Apples comparisons benchmark more GPUs, often ones from a previous generation. Lets say when the new GPUs hit some time in the next months I'd like to be able to see how much of an improvement a new card will be over my GTX 970.

Gotta agree with this. Hard to see value in upgrading my 970 if I can't see the difference between them.
 
Not so much a review, but have some place to compare cards. Have a game that scales well and keep it in the rotation (Q3 was one that lasted a while, despite getting 300+ FPS eventually) just for that comparison. Keep a 'overclockers comparison page' that you can compare different cards. Say the current card in your machine to the newest card reviewed. You get the basic stats and the FPS (and other details) of the cards in a nice format. So, you can do a direct comparison.

I have an AMD 7950. If I go back to the reviews on that, and then a new review, there will be little to compare it to. But, I KNOW a new card would be much faster. I just want to fully educate myself on how much of an increase I'll get. If I'm going to spend $300-500 or more on a card, I want all the information I can get.

Of course, if I have the old card, I can just test with a game in the review and then look at your numbers. But, you have a more standard way, and a standard machine set up, etc.. At the very least (and it may have been done previously, I cannot recall), put out an article on how you do the review, details on the OS (custom install? Install + updates + drivers? Fresh install?), game benchmark details (built in? custom benchmark?), things like that. So I could have a similar machine set up and be able to reproduce your numbers within a small margin of error.

I really enjoy the reviews now, and this site is one of the very few that I actually read and put a lot of faith into. I won't buy a new card without reading the reviews here. I've had my eye set on a card, read the [H] review, and ended up with a different card (same with other hardware). I'm sure there could be some improvement, but the reviews are far from bad. Very trusted. Brent really knows his shit.
 
I would like to know what clock speeds the cards are running at. With gpu boost 2.0 not all cards of the same model at least nvidia side run at the same clock speed. While the differences of frame rate may be small others may be a big difference. A 980ti at reference may be at 1086mhz while a lightning might be at 1400+ does make a difference as you know. Knowing the clock speed of the cards used in the test would help give a bit more reference as to what is being tested.
 
I hate the popout pictures. That's one thing that has bugged me for years. I wish the pictures would expand dynamically. It's a PITA when viewing reviews on a mobile site as this leads me to another page which causes me to close it out and go back to the main review.


Beyond that I agree with others stating a more streamlined approach of options that changed vs. ones that were the same across all cards. Less is more.
 
Hello there :)
Just signed up to answer the call.
As far as I'm concerned GPU reviews are great.
I'd just like to see with the introduction of DX12 a comparison with different CPUs.
I mean wasn't that one of the selling points of the API? (better CPU usage).
Keep going with your review, you're doing great guys!
Many thanks
 
I would like image quality comparisons. It keeps the companies more honest. Even one or two screenshots comparisons I would find useful. I am always worried when I see highest playable settings or benchmarks (rare cases when you show them) that one of the companies may be cheating by not honoring the settings in the game to give better performance or one company having a poorer implementation of the setting. You don't have to spend a lot of time on it, just find one busy scene and take a screen shot. It would also tell if I care if the highest playable settings on each card matter to me. For the DX11 vs DX12 I am wondering if it looks any different. For different resolutions I am wondering if it looks better at higher resolutions; it may not if the models and textures don't scale that high.
 
My only request would be to have some comparison to older, high end cards. Some of us skip generations between upgrades and it would be helpful to have more relevant data when we do make a purchase. I know it's not feasible do do multiple extra run-throughs of every game tested, but some fairly constant measure would be useful
 
At the risk of being lame and comparing to another site, I always found Anandtech's "Bench" to be useful in comparing. Not so much video reviews, but maybe a HardOCP version of that with your data? Something that has a few more bells and whistle, so that we can see how a card performs in different settings?

Realize that's asking a lot, but it would be a nice feature.
 
  • Like
Reactions: DF-1
like this
I hate the popout pictures. That's one thing that has bugged me for years. I wish the pictures would expand dynamically. It's a PITA when viewing reviews on a mobile site as this leads me to another page which causes me to close it out and go back to the main review.


Beyond that I agree with others stating a more streamlined approach of options that changed vs. ones that were the same across all cards. Less is more.
What browser are you using? chrome has zero issues with pop out pictures. I do agree that Internet explorer seems to do that last i checked
 
now this i agree with....matter of fact i mentioned i always skip the apples to oranges part myself. I wonder how many of us feel this way?

But what good is the apples to apples if the FPS is in the 20's?

My bullet points:

-Better formatting
-Standard test system (maybe a yearly update to it?)
-Some frame data on Crossfire/SLI
 
I don't think there is necessarily anything WRONG with the way the [H] does video card reviews. They definitely don't suck! In fact, I think the subjective "rubber hits the road" type playable settings approach is a huge asset to the H reviews that - to my knowledge - no one else is doing. Sometimes there is more to game performance than just numbers. It does come with the risk of bias - however - but I feel the team have been pretty good at avoiding that over the years

(You can tell by how fan boys from both Nvidia and AMD complain of bias over the years. If you are pissing the koolaid drinkers on both sides off, then you are doing a pretty damned good job.)

I also trust reviews here more than I do most other places, as there have been way too many reports of "pay to play" type reviewing out there from other sites.. I have a feeling Kyle and the gang have enough pride in what they do that they would never stoop to that level.

The H is always my first stop for reviews, and usually only if I don't find what I am looking for here, do I check out another site.

So, there are times I go elsewhere. When do I do this?

It's usually when I am looking for lots of data. I'm an engineer by profession, and an all around geeky type who has always been very into objective numbers. Sometimes the data here is a little bit on the light side. As an example, I don't necessarily think that frame time/stutter type measurements are necessarily warranted in every review. That would just wind up adding extra work, and adding very little benefit. That being said, if - in the subjective opinion section of the review - the reviewer notices something, like the game not feeling as smooth as it should at the reported framerate, that - to me - would warrant triggering further measurements and investigation to see what is going on, and present the data that explains it.

Other times I go elsewhere is when I am looking for apples to apples comparisons for more hardware than the 2-3 GPU's in any given review. I may be interested in trying to interpolate figures to get an estimate for potential performance might be in a build planned in my head. The "playable settings" approach is great if you are comparing things 1:1. If I am interested in the exact GPU and game combination being reviewed, this is perfect. Usually - however - for me, since I am an old fart who doesn't care about the currently most popular titles, I'm usually looking to find out how much better this new GPU under review is going to perform in some obscure game that I play, but most people don't, compared to my current GPU that is NOT under review.

It is unreasonable to expect every combination of game and GPU to be tested, so then I know I have to do some sort of educated guessing. Scenarios like the above, are when I go to - for instance - Anandtechs massive benchmark database, find my exact current GPU, and the GPU I am interested in and look at comparative performance in the most similar title there (maybe it happens to use the same game engine, or at least is a similar style (large map outdoor fps, for instance). This way I can say "Well, you know, I am currently getting 40fps with my current GPU in this similar title, so what I need in order to hit 60fps is something 50% faster. I may not be looking at benchmarks from the same title, but this is fairly close, and it is only 30% faster than my current hardware, so it probably won't be enough". That type of stuff.

Anyway, these are just some thoughts. The H is still my favorite site for reviews, and I'll keep coming bak here no matter what.
 
Last edited:
I know you guys have a good relation with ASUS, the truth is they make good cards but their RMA is severly lacking. I have seen more horror stories than good ones with them and before anyone says otherwise, do your research to look at reviews of spcifically their RMA, at BBB, Reddit, hell even this forum. I know that a lot of people who have issues will be more vocal than the folks who don't but look at the consistancy and you will see ASUS rises above them all in complaints.

I will honestly never buy an ASUS product so most of your reviews I just ignore the brand part iself (so temps, OC ability, extra features all get thrown out to me which is a shame), it can be the best thing out there but the fact if anything goes wrong I will be in some serious troube makes me cringe. If I start seeing more positive things with their RMA I will change my mind, but right now it is what it is.

I know you guys have butted heads with EVGA which is a shame, I wish I could see more of their reviews.
 
I always wonder why these threads turn into a laundry list of each respective poster's own personal preferences, versus offering insights into how they can make the reviews they're already doing more consumable by their readership. Most y'all want them to octuple, if not more, their workload.

That said, I'm in the group that, if nothing else, would like to see the apples-to-apples framerate trace (pending technical difficulties at the moment) return, as that's the most generalizable (IMO) to extrapolation to other configurations (lower/higher resolutions, etc). The science nerd in me (I know, I know, y'all said that wasn't a direction you wanted to go) would prefer that you pin to a fixed resolution throughout a series of tests, like you did in the Tomb Raider DX11 DX12 eval, rather than letting that one be a floating variable.
 
I think you guys should answer the questions we really want to know.
Like if i drill holes through my brand spanking new gpu in order to mount a bracket for my aio liquid cooler, would it still work?
 
I always skipped the "highest playable" settings parts of H reviews. They are meaningless to me. I'll decide what are my highest playable settings. All I'm interested in reviews is apples to apples comparisons. And if DX12 can't provide that, well then DX12 sucks ass, as far as testing goes. Hopefully there will be tools available for this right? Not that it matters much to me yet, as I'm not planning on going to W10 for the foreseeable future.

Maybe you could make the "highest playable" settings comparisons more meaningful and insighful, and actually helpful. By quantifying it.

What I mean is assigning score to each graphics setting in a game.
And showing the score of the cards in charts as you would show the FPS charts.
 
I always wonder why these threads turn into a laundry list of each respective poster's own personal preferences, versus offering insights into how they can make the reviews they're already doing more consumable by their readership. Most y'all want them to octuple, if not more, their workload.

That said, I'm in the group that, if nothing else, would like to see the apples-to-apples framerate trace (pending technical difficulties at the moment) return, as that's the most generalizable (IMO) to extrapolation to other configurations (lower/higher resolutions, etc). The science nerd in me (I know, I know, y'all said that wasn't a direction you wanted to go) would prefer that you pin to a fixed resolution throughout a series of tests, like you did in the Tomb Raider DX11 DX12 eval, rather than letting that one be a floating variable.

Unfortunately that's what people want to know, how their card will run at specific settings and resolutions. Yes reviews are a lot of work but if the reason they do it the way is to save time then they are doing it wrong if you don't get the information you want. I am in no way saying that's how [H] operates, just answering the general question. I don't care if you can play with high settings at 1440P, I care about playing Ultra at 1080P. Things don't always scale linearly so you just can't assume they will. I want to see the actual numbers and frame times if possible. That is the reason I normally skip the numbers and go right to the [H] summary. Then I go to different reviews and get actual numbers. More info the better.
 
Since upgrading to gsync, my minimum playable fps is around 75. If I have to cut some settings, so be it.

It'd be nice to see highest playable settings for different preferences. 45, 60, 75, etc.

The problem with this is it's subjective. There's no way to change that except if YOU are performing the review :D

That being said, when [H] does a review, they hold all the cards up to the same standards. That means you still get a comparative review between multiple cards, because the difference should scale pretty closely (at the same resolution).

So if you like 30-40% higher frame rate than the reviewers from [H], take their reviews and just turn down the texture level AND shader level by one.

That's an easy way to tell if you can live with the card at the same resolution as the review, but higher frame rates.
 
Too be honest I usually skip to the last page to see if the card got a pass or not. Then I go back to see what the big deal was about. I make a list of the cards and browse other sites to get a general feel. If the price is fair from my point of view I will go for it.
 
A place for getting what you can from older stuff is also necessary. might be more suited for a game review. something like "you can max everything else out if you turn of soft shadows" stuff. there seems to be less and less content on specific settings. in a card review, mention what the card is great at and also what slider will have to be the first to be turned down for the next set of games. take my r9 285. sure no max res textures, but can keep up fine with particles and lighting.
 
TukfKg9.gif


Lets see. It is pointless to show data that doesn't change, because it gives us no insight and makes reading the rest of the data much more difficult. So, what I would do is only write about the data that doesn't change (IE "we found that X, Y and Z settings can be kept at W level throughout the cards and systems we tested") and graphically show the things that do change. This makes understanding the differences quite easier, as your eyes get to see straight what is what. In this particular instance the tables would be less than half the size. So not only you reduce clutter, you simplify the understanding part of the review.

In this context, I don't agree.

In a job situation where streamlining the user experience and presenting relevant data is valuable - Yes - you are correct.

In a review where I want to easily see all of the details, take my time to absorb and not have to guess what setting might be - No - you are not correct.
 
On thing I'd really like to see is the Apples to Apples comparisons benchmark more GPUs, often ones from a previous generation. Lets say when the new GPUs hit some time in the next months I'd like to be able to see how much of an improvement a new card will be over my GTX 970.
This would be my biggest request. Yes it is nice to see how the new Nvidia and AMD cards compare, but how much are they improved from the card I already have?
 
I trust [H] GPU reviews more than any other site.

If the framerates cannot be monitored in the old way for now, then the subjective playable experience that [H] is already doing becomes even more important.

Some are complaining about the data that didn't change between runs, not being important. But I disagree. I think people come to [H] to know what settings to set to make a game playable on their system. To do that, we need to know all the settings. In some games with a lot of settings it might get messy, like in the Tomb Raider review, but I believe the information should be provided. Some tweaks to how it is presented wouldn't hurt, but don't just leave it out.

What sucks is no apples to apples, which allowed us to see different GPU's strengths and weaknesses on an even playing field. It was a good way to quantify it, and satisfied those who wanted benchmarks. And it could still be performed and interpreted in a subjective manner by [H], who could comment on how sluggish or smooth it was, etc.

Is there no hope for a future tool that can measure framerates and frametimes? I bet someone could mod an LCD in such a way as to be able to measure all those things, in the LCD itself. Or maybe in something that sits between the GPU and the LCD, monitoring the data going across.

Not too worried really about the format, or the current lack of the hard numbers. Someone will find a way to measure it...
 
Personally, while I do read, enjoy and am entertained by the reviews here, I find them next to useless from a buyers/consumer standpoint.

If I'm in the market for a GPU, I want to see how it stacks up against the competition. I want hard numbers/data taken in a controlled environment (AKA, all settings used are the same across the board) because ultimately, that tells me exactly what I need to know. When settings are adjusted across tests to hit a target framerate, that really doesn't tell me anything. Thats what I, as an enthusiast, already do to to make a game perform to what I consider an acceptable level.

For an example, I just built a 4K gaming box. I wanted to see how various GPU's stood against each other rendering games at 2160p. [H]'s reviews wouldn't of really helped me with that unfortunately. I wanted the BEST performing GPU at the resolution (all else being equal), and I'd adjust settings from there to get the framerates I wanted.

I liked this post but one can argue that this information is already present in the apples to apples chart. What other information would you want?

-Edit-

Well other than the issue that DX12 makes this impossible (for now) to create.
 
I would like to see an "Ashes of the Singularity" DX12 (or similar game that supports it) multi GPU section added. I am intrigued with the idea of increasing graphic quality in my games by adding a GPU from a different vendor. Maybe compare to dedicated Crossfire and SLI setups?

Not sure if this should be a standalone review or have a couple of dedicated cards from different vendors to throw into the mix. So if reviewing the latest Nvidia GPU, throw the AMD card in to run some test, reverse it if reviewing the latest AMD GPU...
 
IMO, the only problem I've had with the reviews from the years that I've been reading HardOCP was the fact that I don't give a flying F about the playable settings really. I care about the Apples to Apples only... I want to know how much better card A is better than B and weather I should upgrade. Using old benchmarks to compare new cards to old is helpful as well as I usually only upgrade when I get double the performance.

For instance, I had a Radeon 5800 and moved to 780GTX and looking to move possibly again.... It just has to be worth it
 
What I like see in stead of using High End CPU use Low CPU i3 or FX older even old CPU with High GPU would this give us a better present chart ?
 
What sucks? = Nothing, What could be added? =

I think a 60FPS minimum fps achievable with the following settings on our test setup section can be added. In the rare event the fps were way higher than 60fps shoot for 120fps or 144fps
A max OC vs max OC on the Apples to Apples test would be cool if you can squeeze it in the article and get to publication on time.
More 1080p results even if it were only the apples to apples section and not the max playable settings. This would mean the same type of apples to apples section you normally do and add a 1080p apples to apples alongside it.
 
I have mixed feelings about this new format. On one hand I don't tend to take notice on FPS anymore and rarely even look at the carts. So I read the articles and take note on the findings an conclusions.

But now that I don't have them well I kind of miss them :p
 
There are a couple of things I miss in most if not all hardware reviews, not only on this website, and it is probably not possible or worth doing but still irks me, for instance I have a GTX780, I will almost never see it compared vs any card due to it's age or because it falls between the 770 and the 780ti so I have to kinda guess how it performs compared to the other cards.

Also reviewers almost always tend to test all GPU's against 1 usually the most powerful CPU they have access to which always makes me wonder how big the differences are on lower end machines, but this is a smaller issue to me.

Maybe the first problem can somewhat be solved by having a list available with (semi)regularly updated benchmark numbers for a couple of the most recent generation of cards with the latest drivers so you don't have to look at 2 outdated series of nr's to have a rough idea of the difference between the new card and your old one.
 
Given how many different monitor configurations there are these days, I wonder if it would be possible to derive a Meapixels Per Second metric for a given card in a given game. That might allow us to theoretically predict performance on our 1080p, 1440p, 4k, various VR headsets, etc. Maybe the relationship is not linear enough to interpolate. Maybe that is something worth finding out. It's just a thought for something new to try.
 
I'm pretty happy with how you guys do all your reviews.
I really like how you include the system specs used with each hardware test. I use that information as a base guideline when I am putting together a system.
 
I appreciate the apples-apples comparison as much as the highest playable settings comparison. A to A helps gets a direct comparison which sometimes helps to decipher the highest playable comparison. Other times the highest playable helps find other things the A to A doesn't. They are mutually beneficial comparisons.

Things I would like to see:
-More roundup reviews comparing several video cards at once. Things to focus on here are build quality, heatsink/fan temps and acoustics.
-More niche video card reviews, e.g. SFF cards, watercooling blocks including the AIO watercooling units, etc.
 
Long time reader (bought stock in 3dfx back in the day), first post.

Personally, I am going to buy a graphics card for a reason, not because it just came out on the market (real or imagined launch).

Reasons like....
  • I want to play games at 4k at 30, 60, 90, 120 fps on my tv/monitor. What graphics card(s) do I need?
  • I want to play game X at 1080, 1440, 2160 resolution. What graphics card(s) do I need?
  • I want to get a VR headset, which graphics card is going to be best for Vive, Oculus, pornhubs Inyoureye?
  • I want to build a new system to play game X @ X resolution. Do I pull the trigger now or wait for Polaris/pascal...
  • My 52cnd AMD XFX graphics card melted, which graphics card melts less often?
I would love to see question like these and others answered specifically.

On a side note, while I appreciate HardOCP being upfront and open when having problems with a specific vendor (the AMD debacle last year), there are only two GPU manufacturers of substance. From a business perspective, you might be better off letting the stinking pile lay, you still have to work with them. From the community perspective (which is what HardOCP is in many aspects) I made some popcorn, sat back, and enjoyed the shit show :) I am sure it was a tuff call, I am not sure the right one was made... I will give y'all this though, you have been consistent. Damn near the same thing happened in 2005 with ATI, and I remember some beatdown on Nvidia for interesting image quality issues.

For what its worth, keep up the good work, the good news, and stay hard (awkward).
 
In this context, I don't agree.

In a job situation where streamlining the user experience and presenting relevant data is valuable - Yes - you are correct.

In a review where I want to easily see all of the details, take my time to absorb and not have to guess what setting might be - No - you are not correct.

You wouldn't have to guess anything. All the settings used would be presented... but those that were kept equal would not be on the table, they would be elsewhere.
 
There's not enough attractive, scantily clad (or even naked) women holding the cards in suggestive manners...

THIS OFFENDS ME! FIX IT!

;)
Full disclosure: I purchased my current cards (2x AMD Sapphire 7950) based entirely on the box art. My friend was trying to convince me to get an R9 or GTX 960 but I chose the rendered marketing ploy.

As far as the graphics reviews here, I've always wished there was a section focusing on optimizing high frame rates, as I'm more interested in hitting the human retina limit of ~83fps. I dislike how most gamers feel 60fps is acceptable for the benchmark of "playable" just because for so long they have been force-fed 60hz LCD's and were sold the lie that frames above this level are indistinguishable. I can usually estimate the settings needed to reach a playable frame rate based on the data given in your reviews but there is a lot of unwanted data like 4k resolution and the sections where you talk about how the game play smoothness feels. 4k resolution will only matter when monitors supporting >80hz refresh are available. To me it is just a waste of space to scroll over in the reviews.

You're hardware reviews are still and have always been the highest quality on the web, and you don't really need to change anything.
I do have to say the most recent Tomb Raider review of DX12 was abysmal, and it left a sour taste in my mouth.
 
A Dollar/FPS chart. The majority of us don't have unlimited funds to spend on playing video games. A good portion of our paychecks go towards our mortgage, utilities, groceries, etc. Getting the most bang-for-your-buck is the primary goal for many people when shopping for a new video card. A Dollar/FPS chart for video cards (similar to your Dollar/Performance Ratio charts for your CPU cooler reviews) would be welcomed.
 
Back
Top