Reform the rating and ranking system of Steam &

[小糖人愛音]

Steam's rating and ranking system is rubbish and lacks reference value to a certain extent.For example, can you trust Steam's system to think that "HoloCure - Save the Fans!" and "Senren*Banka are the best games? Obviously there are some problems here. The problem lies in the fact that Steam's game rankings are ranked according to the percentage of positive reviews of each game to all reviews of the game. This causes games with 99% positive reviews to automatically reach the top. However, this is unreasonable and completely ignored. Issues with different number of reviews for each gameSo Valve, could you care more about it?

I want steam to be better, so I think this is a good change, of course, if it can be implemented

[nullable]

Publicado originalmente por 柊史曉·將臣昂晴:

Steam's rating and ranking system is rubbish and lacks reference value to a certain extent.So Valve, could you care more about it?

How so. Saying it's not good and to fix it, isn't really a suggestion. It's just empty criticism.

Publicado originalmente por 柊史曉·將臣昂晴:

We players all want steam to be better, so I think this is a good change, of course, if it can be implemented

You claim to be speaking for everyone. Lots of people do this, because they;r personal opinions have no value. Well exaggerated claims don't have value either because it's not so easy to just say "we" and deceive everyone that you represent any group of users.

Also since you didn't really define what you're talking about or how you think things should change, lying about who you represent is the least of your problems.

[小糖人愛音]

True, but this does not affect the fact that the rating and ranking system of Steam is poor

[Crazy Tiger]

Every rating and ranking system is poor, since they all ultimately only display scores without context.

[小糖人愛音]

For example, can you trust Steam's system to think that "HoloCure - Save the Fans!" and "Senren*Banka are the best games? Obviously there are some problems here. The problem lies in the fact that Steam's game rankings are ranked according to the percentage of positive reviews of each game to all reviews of the game. This causes games with 99% positive reviews to automatically reach the top. However, this is unreasonable and completely ignored. Issues with different number of reviews for each game

[小糖人愛音]

And thanks for the suggestion, I changed "we" back to "I" and added the specific problem.

[BJWyler]

Publicado originalmente por 柊史曉·將臣昂晴:

For example, can you trust Steam's system to think that "HoloCure - Save the Fans!" and "Senren*Banka are the best games? Obviously there are some problems here. The problem lies in the fact that Steam's game rankings are ranked according to the percentage of positive reviews of each game to all reviews of the game. This causes games with 99% positive reviews to automatically reach the top. However, this is unreasonable and completely ignored. Issues with different number of reviews for each game

"Best" is completely subjective for this form of entertainment. And I see no problem with a game with 99% positive rating getting a high ranking. That means people who played the game enjoyed it. That's exactly how it should work. That you don't like those games is completely immaterial.

[nullable]

Publicado originalmente por 柊史曉·將臣昂晴:

For example, can you trust Steam's system to think that "HoloCure - Save the Fans!" and "Senren*Banka are the best games? Obviously there are some problems here. The problem lies in the fact that Steam's game rankings are ranked according to the percentage of positive reviews of each game to all reviews of the game. This causes games with 99% positive reviews to automatically reach the top. However, this is unreasonable and completely ignored. Issues with different number of reviews for each game

Well my guess is you can't produce the math Valve is using in its current system. Methinks you just don't like the results sometimes, so the system must be wrong/bad/too simple in your opinion.

What sort of mathematical formula would you prefer Valve uses?

I'm gonna have to agree with Crazy Tiger, no rating and ranking system is good or without flaws if you listen to everyone's opinions. Some always prefers some other system or implementation.

Besides even if you're worried about some new game with suspiciously high reviews and low number of reviews, I don't think that's fooling anyone. Anyone who is mindlessly buying games because it's 99% positive with no other info ain't the sort of person who's going to be helped by or able to use a different system anyway. And it's not really your job to manage their purchases by manipulating the review system.

[Crazy Tiger]

Publicado originalmente por 柊史曉·將臣昂晴:

For example, can you trust Steam's system to think that "HoloCure - Save the Fans!" and "Senren*Banka are the best games? Obviously there are some problems here. The problem lies in the fact that Steam's game rankings are ranked according to the percentage of positive reviews of each game to all reviews of the game. This causes games with 99% positive reviews to automatically reach the top. However, this is unreasonable and completely ignored. Issues with different number of reviews for each game

Steams review system is for finding games that get recommended, not whether they are "good" or "best". Which is also why I often quite like games with Mixed ratings, plenty of those try new things or do things differently resulting people being divided on the mechanics.

For me personally the system works fine. I only look at negative reviews, since I want to know why people stopped playing, or even better refunded, the game.

But I'm looking for fun games, not "good" games.

[Mad Scientist]

The chances of hitting and maintaining 99% positive would likely be extremely difficult, so it's not like something that reaches that would be able to hold that for very long. It's also up to users to do enough research before buying to see if the total number of reviews at a percentage appears correct, and more importantly, if they will likely enjoy the game before buying it.

[aiusepsi]

Yeah, the way that Steam currently boils down the ratings to a single number is not very good.

That single number is supposed to be a good estimate of the actual quality of the game. And when there's a lot of ratings, it is. With low number of ratings, you have a simple statistical problem that because the sample size is low, it's not a good estimate.

What you need is a method which takes into account the uncertainty implied by small sample sizes. The rating method used by SteamDB[steamdb.info] for example, does this. With that said, "HoloCure - Save the Fans!" is still number 7 on SteamDB's top-ranked games list[steamdb.info].

[Start_Running]

Publicado originalmente por 柊史曉·將臣昂晴:

Steam's rating and ranking system is rubbish and lacks reference value to a certain extent.For example, can you trust Steam's system to think that "HoloCure - Save the Fans!" and "Senren*Banka are the best games? Obviously there are some problems here. The problem lies in the fact that Steam's game rankings are ranked according to the percentage of positive reviews of each game to all reviews of the game. This causes games with 99% positive reviews to automatically reach the top. However, this is unreasonable and completely ignored. Issues with different number of reviews for each gameSo Valve, could you care more about it?

I want steam to be better, so I think this is a good change, of course, if it can be implemented

This is already counted for in the system. The aggregates are gated by review number. Youb can't get more than Posiotive unti you pass a certain threshold. , Same for Ver, Mostly, and Overwhelmingly. Each one requires a certain number of reviews. and its the aggregate ranking that determines ranking.

If you're suggesting something more hands on, then why shiould a game that is designed to appeal to a small target audience be penalized over some genwric grey paste mas marketed trash?

Maybe you bneed to stop thinking of review scores as some form of absolute objective metric , which they never werem not on steam or any where. Instead see them for what they are.. a measure of how well the game was received by it's target audience.

[Tanoomba]

Holocure is freaking awesome.

[小糖人愛音]

Publicado originalmente por Tanoomba:

Holocure is freaking awesome.

I'm not saying is not a good（）

[vova.khayan.84]

As for me, the average value of "maybe" is missing in the assessment.As a result, reading some reviews, you cannot draw conclusions for yourself.After all, it happens that a player scolds the game, but it's worth "recommending" it.Yes, I've come across this myself a couple of times when writing a review.It seems that you like something about gre, but you can't recommend it, and on the contrary!