Seems like quite a level headed approach to it. I’m always a little leery of these types of systems but they explain it well and I think their reasoning makes sense.
Thank God. I’m so sick of seeing low effort reviews with one or two actual reviews on the front page. From copy paste ASCII art to listing every single disease in the world when reviewing a bad game. And don’t even get me started on the “if this gets [blank] I’ll do [blank]” reviews
I will kill the next ascii cat I need to rate helpful to pet
When I like or dislike a game a lot and I’m able to articulate why, I usually write a long-winded 2-3 paragraph review on what I liked and didn’t like about it. These reviews tend to get 1-2 reactions at most
One time, I decided to write a positive review for a game I really liked but didn’t feel like saying “the shooting good” six different ways, so I wrote my one and only joke review
It then became my most popular review, with now 37 reactions
So yeah, I see why people do it and yes, I heavily agree with what steam’s doing
Sounds like AI will be used to detect the shitty meme reviews. Honestly a great use of ML
There’s a “not helpful” button under every review.
It has never worked though, as far more people would vote helpful if something made them laugh (because people just think “more upvote”)
I dunno, that’s why they added the “Funny” button and I see people use that all the time. Even the nearly-useless “Was this review helpful?” section on Amazon has some use to a customer making a purchasing decision.
If ML can be used to further help the issue, what’s the problem? At least “AI” is being used for something that’s actually trying to solve a practical issue in an attempt to improve the platform and not as an immediate way to extract maximum profit with minimum effort.
You could argue that Valve loves to automate its customer service to save money, and that would be valid and true, but I think improving the platform experience by trying to reduce (if not eliminate) unhelpful reviews is good.
Seems like not enough people click those. We are taking about gamers here.
Probably true.
Bionicjoey is right though, too. They’re using AI to scale the not helpfulls and other review responses. Personally I’m all for the change.
and some machine learning algorithms to help scale the human judgement calls
I’m just glad they called it machine learning instead of “AI”
You don’t even need advanced AI.
I never said advanced. They say they are using ML as part of the process in the post
I hate those copy-pasted reviews of an old person having lots of fun playing the game with their kid. It’s told in a way that’s emotional and I thought it was true until I started seeing the exact same review copy-pasted on every other game.
Wtf that review is fake???
I assume it’s fake because they don’t mention the game itself in the review and I see it copy-pasted in a lot of games. I think it’s a way to farm steam point rewards.
I hope this system works well, right now I only really read like 5 positive actual reviews and then read mostly negative ones, as that’s where you find a real grasp of what a game is or isn’t.
Good, the Steam reviews were in dire need of an overhaul. The amount of absolute garbage in the reviews and how it’s practically only used to bomb games for whatever reason is getting out of hand and it was often very difficult to get to the good informative reviews about the actual game.
The next problem will probably be that the people that review bomb games are now likely going to target helpful reviews instead. Because nothing is more obnoxious than haters taking down games and people who actually enjoy them, even if provided with helpful and informative descriptions.
A good example right now would be Star Wars Outlaws, most people who got to try the game are mostly optimistic about the game, while still being realistic about flaws and issues. But on the other side there are a load of haters using various excuses elevating the toxicity to a new level. Some of these reviewers are scared to even post their opinions because they get targeted by toxic trolls trying to take down their content because the reviewer’s opinion doesn’t fit their narrative. Some reviewers even ended up having to remove their content because it was hurting their image they’ve been working on for so long.
I think that Steam would really need are truly neutral curators judging whether reviews on Steam are legit helpful and informative or not. But with the amount of games on Steam that might be an impossible job.
A problem I’ve personally run into is a lot of reviews focus entirely on MTX/monetization and this can turn into a fun game getting thrown through the floor because the corporate side decided to put in a bunch of optional purchases.
I’m not saying monetization doesn’t matter … but sometimes I really don’t care … like Lego 2K Drive has a bunch of MTX bricks I’m never going to use and the option to grind them out with a lot of play time (another thing I’m not going to bother with). Those reviews almost definitely really hurt the sales of the game (which I did end up getting and it’s actually quite solid in terms of PC kart racing) and probably killed any chance of it ever living up to its potential.
Artifact Classic (the card game by Valve) also got review bombed to hell about monetization (and that one I get a little bit more because you had to buy card packs) … but if you actually play the game (and you can for free now with all cards unlocked)… I found it to be a really fun card game. I and all my friend stopped playing when Valve announced they were just going to redo the whole thing… I suspect a lot of people did the same thing which caused the chain reaction of “nobody’s playing our game… this looks hopeless…” and the eventual abandonment of both the original game and the rework Artifact Foundry (which I … didn’t particularly care for). The people I know that knew about the game that didn’t get the game said the negative reviews basically immediately disqualified the game from consideration for them.
I want to be able to rate games as ‘meh’, a lot of games i dont feel strongly enough to say its outright good or bad.
Example: there is an extremely highly rated game from a few years ago that i got around to playing, but to me it was just okay and i’ll probably never get around to finishing it. Based on my achievement progress i find it highly unlikely that 95% of the reviewers have finished the game either. They probably reviewed it after playing for an hour.
I dont want to review it as good or bad so instead i just wont.
But if 0 is bad and 1 is good then tallying 0.5 as meh shouldnt be that hard.
