CheesecakeRecipe
Stormy Grey
Sega throwing PC players a Persona 5 shaped bone
with a PSO2 collab~
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
STEAM | September 2016 - Good job doing previous stuff, let's do new stuff
- Thread starter MRORANGE
- Start date
- Status
Yes. Rip valve. Gaben took the money and run.
PetriP-TNT
Member
Turok has brilliant stage design by the way
It's a great game too, when is the 2 coming out?
Sweet
Anyone ever tell you that your taste in video games is terrible? Because it is, no offense!
Black Soma
Member
Capcom still hasn't fixed the no audio issue in Dead Rising. I rushed to beat Doom so I could start this at launch but I can't even play it, not like this. Bah.
d00d3n
Member
That sucks. I was waiting for the first free new game to show up in my account after getting the "Valve Friends and Family Complimentary" due to the steam controller OSX debacle. But that won't happen then? No free L4D3 for me ...
(unless it is free to play for everyone, I guess)SunDimension24
Member
Why does every PS4 and PC thread turn into garbage?
Also I love how going exclusively to PlayStation is a fine idea but going exclusively to PC is unimaginable and the weirdest thing ever.
Also I love how going exclusively to PlayStation is a fine idea but going exclusively to PC is unimaginable and the weirdest thing ever.
Baron_Calamity
Member
Kinda sorta not really.
Story wise, it is not. Its set 16 years after the first dead rising. Frank returns to Willamette, Colorado, Mysterious outbreak at a new mall, Willamette Memorial Megaplex Mall. But theme wise, its very much a reimaging of the first game.
Yes indeed, well kinda I guess, its also going to be a timed windows store and xbox1 exclusive iirc. I didn't know about ^ that information though.
1. Sony fanboys.
2. Sony fanboys.
3. They want to play everything but not offer everything, which generates a "superiority" in selection that perpetuates their Sony > everyone else mindset.
That sucks. I was waiting for the first free new game to show up in my account after getting the "Valve Friends and Family Complimentary" due to the steam controller OSX debacle. But that won't happen then? No free L4D3 for me ...
Valve Friends and Family Complimentary is still a valid sub. You'll get all future Valve titles.
d00d3n
Member
Lol, Steam has changed its mind about my ownership status. I used to get the error message "You must own this product to review it.", now I get "You have used this product through Steam Family Sharing, but you must own it in order to post a review.". I have never used family sharing of course ...
Thanks, good to know!
Wok
Member
Nice, now I can buy the game for the same price as Humble Store, and my Steam review can still count! YAY!
Edit: It is actually 1 cent mor expensive to get my Steam review to count! What to do? What to do?
This is the cheapest Mu Cartographer has ever been.
HEADS UP STEALTH AND TENCHU FANS
Aragami is releasing on October 4th on PS4 and PC
http://www.neogaf.com/forum/showthread.php?t=1278460
http://store.steampowered.com/app/280160/
Nice find, Derrick.
hardcastlemccormick
Banned
UcchedavÄda
Member
There is obviously an ethical responsibility associated with presenting data. However, I think there are ways that you can make yourself more or less responsible. For example, making normative claims about what you present clearly implicates you more ethically. Presenting data selectively clearly implicates you more ethically. Refusing to comment when people draw inferences you know to be incorrect implicates you more ethically. And indeed, Valve has already commented by directly accusing developers of review-buying campaigns, they simply didn't name the developers.
Allow me to make an inference. Suppose we set aside the number of reviews, arguing that due to bundles, third-party site purchases, and yes developers giving out review keys, there are any number of innocuous reasons why the bulk of all reviews would be external purchases.
Suppose we focus only on the differences in score. To do this analysis, what we'll do is assume that one type of reviews (say, external reviews) are the honest truth, and then ask how likely the internal reviews are, given that. So, like, you know that if you flip a coin 50 times, you're going to get around 25 heads, right? So if you get 25 or 27 or 24 or whatever, you're probably going to say "yeah, that's a fair coin". But if you flip a coin 50 times and you get 2 heads, you're going to say "that's very suspicious". It's not impossible, obviously you could get 0 heads, but it's very very very very very unlikely, right? Well, the way we would calculate this is treating each coin-flip as a "binomial random variable", with a p=0.5 chance of being a "success" (heads) in each trial. Then, it's possible to figure out exactly how likely an event like only getting 2 heads out of 50 is. Well we can do the same thing for events that have a different probability. Like, suppose you have a 6-sided dice, and you roll it 50 times, and you roll 45 "1s". The probability of a 1 isn't 50%, but we can still use the same math to calculate how likely that outcome is.
Okay, so let's assume that the external reviews are the truth about the game and simply ask "given that the external key reviews are completely honest, what is the probability that we would see internal reviews be this different?".
Take the top game on SteamSpy's list--the game with the widest score differential: Simple Ball Extended Edition. With 7 internal reviews, it has a 14 percent positive review rate (1 positive review). With 51 reviews including those 7, it has a 88 positive review rate (45 positive reviews). As a result, we know that 44/44 external reviews are positive, and 1/7 internal reviews are positive. This is a problem for our assumption, because the tools of probability actually break down in cases where you get a 100% probability of something happening and it doesn't.
Let's say we do it a different way. Let's assume that the internal reviews are real, so Simple Ball above has a 14% chance of being positive. What's the chance that 44/44 external reviews are positive? 0.000000000000000000000000000000000006 percent chance. Hmm. Hmm indeed. It is very unusual that the external reviews would be so dramatic.
Take the second game: Lands of Devastation. 2 positive internal reviews of 12, 143 positive reviews total, 141/154 external reviews positive. Okay, this time we can use my original design. Assuming the external reviews are true, what's the probability that chance alone would give us such bad internal reviews. The external reviews are binomially distributed with a 92% chance of success (a positive review) What is the probability that the internal reviews were generated from the same process? 0.00000005 percent. There is a 0.00000005 percent chance that it's a statistical fluke that explains the difference between the reviews. Or we could be generous and calculate the chance that the difference is at least that big--i.e. the chance that we'd get internal reviews at least as bad? 0.00000006 percent chance. This is a thing that makes you go "hmmm".
These developers should definitely play the lottery, because the odds of winning are way higher than the odds of these bizarre statistical flukes happening. They should definitely avoid getting brought in for murder charges, because the odds of a perfect DNA match being wrong are way higher than the odds of these bizarre statistical flukes happening. These calculations, by the way, are conservative in the developers favour, because they assume that each review is totally independent of other reviews. In reality, there is a bandwagon effect for a few reasons. So the independence assumption actually raises the odds that totally divergent reviews occur by inflating the variance of the binomial process.
I'm not saying they cheated. I'm not saying they bought reviews. Who would do such a thing and what incentive would there be to do such a thing? I could never imagine why that might happen. What I am saying is that I am very certain that the internal reviews and the external reviews were not generated by the same process. Maybe that means that they're totally different groups of players reviewing the game from internal and external purchases and each has a different set of values. Maybe the reviews took place at different times and the game got patched substantially in between. But something is going on in these games and we can safely rule out that random chance is the cause of their big change in status.
Since I believe in "I report, you decide", here's a link to Simple Ball:
http://store.steampowered.com/app/487530/
Lands of Devastation:
http://store.steampowered.com/app/467560/
For Simple Ball, here is a review from the key redemption section, which is positive:
Here is a review from the purchases section, which is negative:
You can feel free to draw your own opinion about those games, and whether or not it is likely that the developers engaged in (or benefited from unknowingly) review rigging.
(BTW, you can see an example of a developer who benefits from this change if you check Elves Adventure, a game mostly reviewed from bundle keys and widely considered to be terrible. Some of the very few internal key purchases gave the game positive troll reviews--one positive review talks about wanting to drink bleach--and because there are so few internal reviews, the score looks very positive now that all of the honest and negative bundle reviews are removed. This is not a case of key rigging. It's a case of an orchestrated troll campaign working in the game's favour. LOL.)
Thank you for the long post.
But I have mixed feelings reading it, because I was not asking for you to do a statistical analysis, nor for an explanation of how you might do so, rather I was arguing that people who release data like that which the SteamSpy owner released, should do so with appropriate analyses and after consider the implications of doing so.
Also, while I appreciate you taking the time to explain binomial distributions, your application seems a bit misguided. The way you are formulating your tests, you are asking what the probability of one set of binomial values (e.g. internal reviews) being drawn from a binomial distributed described by a different set of results (e.g. the external reviews). However, we must remember that both sets of these results at best approximate the true underlying distribution, even if both sets of results were both drawn from the same distribution. This should be obvious, when you are, for example, inferring the underlying probability of success from a mere 7 reviews. We therefore need a test that can (potentially) reject the null-hypothesis that these were drawn from the same distribution, rather than than testing if one was drawn from the other.
If I recall my statisticals correctly, one appropriate choice would be two-sided Fisher's exact test. So given the pairs of results you have described, this could be done something like the following in R (I have cropped the results to keep this short):
Code:
> fisher.test(matrix(c(6, 1, 0, 44), ncol=2))
p-value = 3.887e-07
> fisher.test(matrix(c(10, 2, 13, 141), ncol=2))
p-value = 1.941e-08Additionally, we need to correct for multiple tests. We are not looking at these games in isolation, but rather as part of the ~7830 games on Steam, so with a p-value of 0.05 we would expect something like 390 false-positives purely by chance. The simplest and most conservative way to do this correction is simply to multiple by the number of tests (7830), which yields p = 0.003043401 and p = 0.0001520088. Better methods exist, but IIRC they require that you have access to all the p-values, and I'm not going to bother scraping the Steam store to do this calculation.
However, these were just two of the most extreme examples. Let's try to take something less extreme, for example Fossil Echo that Wok mentioned. For internal reviews there are 18 positive and 7 negative. For external reviews, there are 17 positive and 0 negative. This gives us
Code:
> fisher.test(matrix(c(7, 18, 0, 17), ncol=2))
p-value = 0.03001Which is statistically significant on its face (and amounts to a 11% difference in score), but not once we correct for multiple testing. Which is why you have to do this kind of analysis before you start pointing fingers at all. And pointing fingers is what the SteamSpy owner did, even if that was not his intention.
EDIT: Did a quick and dirty estimate based on the table that SteamSpy published. 68 of 427 had statistically significant differences in internal and external reviews at p <= 0.05, using the same method as described above, and assuming that I didn't screw up anything now or before. Of these 68, 57 went down in score following the recent change. Though this is of course based on estimating the numbers of positive and negative reviews based on the rounded before / after percentages.
PepsimanVsJoe
Member
Not at all actually. I'm in talks to write for another site, and I'm also going to work with my curator group, while we figure out the next step. Even if less review-keys are received, I still buy a ton of games, so I'm not hurting for material to write about.
I might start prioritizing reviews for games that I've purchased through Steam though.
Fuck me
What is the address?
EDIT: Necropolis quick impressions: It can be quite short but its fun to play. The combat is very Dark Souls ish but more simplified and less weight to it.The ending is kinda sucky unless I missed something but still good fun there to play coop with friends.
Baron_Calamity
Member
Is there any way to tell want games you are allowed to review at a glance?
sixteen-bit
Member
Aren't they doing a Dragon's Dogma Online collab too?
Ah, nice. I was actually a wee bit annoyed when I activated my Kickstarter key and noticed it wasn't tied to the beta testing sub as I then assumed backers wouldn't have earlier access.
HEADS UP STEALTH AND TENCHU FANS
Aragami is releasing on October 4th on PS4 and PC
http://www.neogaf.com/forum/showthread.php?t=1278460
http://store.steampowered.com/app/280160/
What's corrosivefrost doing on that banner
hardcastlemccormick
Banned
Before I had owned an actually modern PC, I had just used cheap Dells with integrated graphics, Macbooks, or budget low profile gaming cards. And it was a nightmare. I hated it, I was always fighting the machine to try and play the simplest game. This was also down to Windows XP, where sky's the limit on how many things can go wrong. And how shitty PC gaming was before Steam/GOG/Humble/etc.
Now I have a decent PC that I built this summer, and the experience is completely different and hugely superior. Modern PC gaming is much more hands off, easy to use, and requires minimal effort to get started. I'm actually flabbergasted how different of a platform it feels.
But it's hard to shake initial impressions, and I wouldn't recommend NeoGAF as a forum where posters are known for their nuance.
derExperte
Member
Origin got a 50% voucher which also works with Watch_Doges 2 (Gold). http://www.neogaf.com/forum/showthread.php?t=1278613 And you can get ME:C for cheap.
They just lowered the base price. That's almost the same.
They just lowered the base price. That's almost the same.
Tizoc
Member
Wish it was on sale so I could get it for cheaper ;P
Witcher 3 just went up from 1080p30 to 1440p60 for you.
Rockin' Ranger
Member
I figured it was a Mass Effect: Collection that had all games and DLC and now I feel dumb.
I should have known better than to think that EA would make an easy way to get the DLC for Mass Effect. Gotta sell those Bioware Points.
People that had problems downloading DOS 2 EEA because they bought the collector's edition, JaseC has good news:
So you can start downloading it now
So you can start downloading it now
I love that kinda stuff too!
Still thinking about waiting for it to see if it comes out on Steam
LurkerPrime
Member
Does the PC version still have inauthentic music compared to the console version? Because that's a gamebreaker.
Tizoc
Member
Oh wait I can get Mirror's Edge 2 for EUR 15 hmm....
http://www.neogaf.com/forum/showthread.php?t=613186
http://www.neogaf.com/forum/showthread.php?t=613186
SilentMatthew
Member
D:OS2 early access is coming on the 15th of Sept, not sure if I should spend time on the early access. Game of the year is near and still haven't played that many games yet to make a proper list.
If you backed it you can download it now. Early Early Access!
And if you don't play now, how will you know what build you'll use when the game is released!?
I hate almost everything about Watch Dogs 2 so much that even though I can rarely pass up on a pre-order deal, I still can't convince myself to get WD2 Deluxe for 30. Incidentally I bought the original Watch Dogs Deluxe also on Origin, it wasn't terrible but to me it looks like WD2 is almost the same game, except even more annoying/irritating. Time spent on any Watch Dogs game is better spent on those better games stuck in backlog queue.
The other option for the voucher is Mirror's Edge Catalyst for 15 but again from what I've seen, they made Mirror's Edge but just slightly worse and more irritating. Blah.
hardcastlemccormick
Banned
On the plus side, Bioware points are also 50% off in the Origin app through Mass Effect 3's DLC pane, so you can get all story DLC for ME2 and 3 for a cool $40.
you had me pretty stoked until the end there
No, the original music was patched in shortly after the Steam release.
hardcastlemccormick
Banned
I had myself pretty stoked until I added up how much I had spent.
Baron_Calamity
Member
I HATE the new origin. Major misstep.
LurkerPrime
Member
What is the new Origin like? I haven't launched it since I bought a Sims 4 expansion a few weeks ago for... some reason. (It's a really bad habit.)
Edit: And as if on command, someone made a thread about neOrigin!
Awesome! For some reason I thought SE would never get around to fixing it.
Edit: And as if on command, someone made a thread about neOrigin!
Awesome! For some reason I thought SE would never get around to fixing it.
Well, I've now finished both of the Half-Life 2 episodes, and I leave on a better note than I thought I would from just the base game experience. I know people have commented on the wonderful story telling of Half-Life 2 and the humour, but whilst I appreciate seeing the world it conjures, without any real plot devices the game just felt like one big journey to me with an ultimate goal but little indication why I was doing what I was doing at any one moment. And whilst Gordon is intentionally silent, the writing which remarked on this fell flat for me ("Leave the talking to me Gordon"). With that In mind I didn't think the episodes would change my mind.
Episode 1 was pretty good, and actually made me feel more involved in actually participating in the plot. I liked the interaction with Alyx and the little set pieces were nice. But it still wasn't the awesomeness that I expected from all the positivity I'd read about. After finishing Part 1 I genuinely wondered why people were still sore about Valve's failure to provide another game....and then I played/finished Episode 2. In this single game everything felt redeemed for me and I finally got it. There were exciting set pieces (the final portion of the Antlion attack in the mine where the music slowly builds, the Vortigaunt's jump in to help and all hell breaks loose....even though I knew it was probably impossible to lose it still felt exciting! Racing around trying to repel the Striders was also great, although difficult to keep up). The writing also felt better to me...the relationships were well written and the humour worked better for me (for example on solving a crate puzzle in the mine - "Ah, a puzzling predicament of the sort which the Freeman is famed for swiftly solving.") And so the ending, even though I knew it would be a cliffhanger, still felt like a gut punch.
So, overall thanks to Episode 2 I'm a convert. Without that final part It'd have gone right over my head and the absence of any third part wouldn't matter to me, but now I'd actually like to see the story evolve...roll on Episode 3 and the exploration of the Borealis!
Episode 1 was pretty good, and actually made me feel more involved in actually participating in the plot. I liked the interaction with Alyx and the little set pieces were nice. But it still wasn't the awesomeness that I expected from all the positivity I'd read about. After finishing Part 1 I genuinely wondered why people were still sore about Valve's failure to provide another game....and then I played/finished Episode 2. In this single game everything felt redeemed for me and I finally got it. There were exciting set pieces (the final portion of the Antlion attack in the mine where the music slowly builds, the Vortigaunt's jump in to help and all hell breaks loose....even though I knew it was probably impossible to lose it still felt exciting! Racing around trying to repel the Striders was also great, although difficult to keep up). The writing also felt better to me...the relationships were well written and the humour worked better for me (for example on solving a crate puzzle in the mine - "Ah, a puzzling predicament of the sort which the Freeman is famed for swiftly solving.") And so the ending, even though I knew it would be a cliffhanger, still felt like a gut punch.
So, overall thanks to Episode 2 I'm a convert. Without that final part It'd have gone right over my head and the absence of any third part wouldn't matter to me, but now I'd actually like to see the story evolve...roll on Episode 3 and the exploration of the Borealis!
freshVeggie
Member
Someday Atlus' fabulous backlog will be on the PC.
- Status