Wszystkie dyskusje > Fora Steam > Steam Community > Szczegóły wątku
Snake4eva 11 stycznia 2021 o 7:39
Numeric Rating System (1-10 / 0 -5)
Hey community, i'm wondering what are your thoughts on Steam's binary rating system (Recommend or Not Recommend) and do you think a numeric rating system would be more useful? Also which rating would you prefer? 0 - 5 or 1 - 10 and why?
Początkowo opublikowane przez Kargor:
I don't necessarily want numbers -- words have a clear meaning. However, I would like at least 3 options like yes/maybe/no, and I'd prefer 5 like definitely, yes but, oh well... maybe, no unless, hell no.

I often feel the gap between one game and another can be rather large -- especially when it was a clear choice on one, and contemplating whether it's yes or no for the other...
< >
Wyświetlanie 16-30 z 153 komentarzy
Snake4eva 11 stycznia 2021 o 14:47 
Początkowo opublikowane przez DiceDsx:
Początkowo opublikowane przez Snapjak:
There's nothing wrong with the recommendation system as is. It's a very simple yes/no thing and people can explain their reasoning quite well in the review portion.

Having #/10 doesn't help when most people will do 1/10 or 10/10 just because.
I get the feeling that some people can't comprehend a review system that doesn't have a score at the end.

The main focus of the review should be the text, the reason why one's recommending/not recommending a game, not which direction the thumb is facing.
But some people do neither. Some people give meaningless text like "worst game ever, don't buy" with a not recommended. Such reviews don't help. A 0 out 5 is equally meaningless but at least we'd be able to ascertain that no thought was given to the score. A 0 rating like a max rating should be a rarity. Also giving games 1/5 or 1/10 also requires a reasonably meaningful review
Snake4eva 11 stycznia 2021 o 14:52 
Początkowo opublikowane przez Meatbug:
How do you put "fun" in a numeric rating, especially when it's also split up in different categories like graphic, story etc. that each need to be rated seperately?

And how do you rate a game like, let's say Rimworld?
It's definitely not a 5/5 in graphics, is it? Probably also not in the sound department. And story? I mean, what story? So a 0 out of 5 then?
Yet it still is, in it's category, an absolute marvel to play and i would recommend that game to everyone asking.

I really don't see the issue with "I recommend (or don't recommend) a game and here's why" (maybe an additional "maybe" would be fine as well).

And most users here aren't professional reviewers, so why pretend they are?
Graphics and sound as in your example can be contextualized. A pixel graphics or 8-bit sound game won't be judged by the standards of a blockbuster because it's genre is contextualized. I do agree that most persons aren't professional reviewers but we can agree that some people at least try to give decent reviews while others don't even bother.
Snake4eva 11 stycznia 2021 o 14:54 
Początkowo opublikowane przez unca.alby:
Alls I can say is asking for a granularity of more than 5 (hate, dislike, meh, like, love) is asking for more than most subjective minds are going to be able to provide.

Maybe on Planet Vulcan, Mr. Spock and his cohorts can give something a score of 7.835 out of 10.000 -- but not here on Planet Earth.
Maybe you're right that only Vulcans could use such a granular system to full effect. Well Vulcans, Romulans and the Q Continuum among others.
Meatbug 11 stycznia 2021 o 14:55 
Początkowo opublikowane przez Snake4eva:
Graphics and sound as in your example can be contextualized. A pixel graphics or 8-bit sound game won't be judged by the standards of a blockbuster because it's genre is contextualized. I do agree that most persons aren't professional reviewers but we can agree that some people at least try to give decent reviews while others don't even bother.

Exactly, the ones that do give proper reviews and put at least some effort into it, are doing that right now and would do so with your suggestion.
The ones that don't put in any effort, won't do so any way, regardless of how the rating system looks like.

If someone writes "worst game ever" as in your example, they would probably give a 0/10 with your suggestion. How is that different? How is that better?
Both things have basically zero meaning and give zero insights.
Snake4eva 11 stycznia 2021 o 14:59 
To recommend something indirectly you are rating it. Read the reviews and people create whatever categories they deem relevant and then they "rate" the game accordingly. The Recommend or Not Recommend is as a result of some inherent rating system. I'm not saying Steam should change their system just wanted to hear the communities thoughts on a numeric system with more categories than the current binary system. Also Steam's system is a numeric system just binary, Recommend = 1, Not Recommend = 0. We at least can agree that a third Maybe would be appreciated
cinedine 11 stycznia 2021 o 15:07 
Początkowo opublikowane przez Snake4eva:
To recommend something indirectly you are rating it. Read the reviews and people create whatever categories they deem relevant and then they "rate" the game accordingly. The Recommend or Not Recommend is as a result of some inherent rating system. I'm not saying Steam should change their system just wanted to hear the communities thoughts on a numeric system with more categories than the current binary system. Also Steam's system is a numeric system just binary, Recommend = 1, Not Recommend = 0. We at least can agree that a third Maybe would be appreciated

It's the other way around.
By recommending something - you recommend something.
By giving a rating, you also recommend something. Giving it a 7/10 means its not worth it for most of your readers. Giving it a 9/10 means you recommend everyone to take a look at.

A recommendation is a recommendation and dependend on criteria. A provides the trap of having objectivity. How many threads do you see where people are asking what's the better game? A racing game or a shooter? Well, it obviously depends whether you want to play a racing game or a shooter. And then you have the inevitable guy recommending a role playing game instead. The later might be of higher rating on the usual suspect sites, but it doesn't make it a better fit in this situation.
Snake4eva 11 stycznia 2021 o 15:10 
Początkowo opublikowane przez DiceDsx:
Początkowo opublikowane przez Snapjak:
There's nothing wrong with the recommendation system as is. It's a very simple yes/no thing and people can explain their reasoning quite well in the review portion.

Having #/10 doesn't help when most people will do 1/10 or 10/10 just because.
I get the feeling that some people can't comprehend a review system that doesn't have a score at the end.

The main focus of the review should be the text, the reason why one's recommending/not recommending a game, not which direction the thumb is facing.


Początkowo opublikowane przez Meatbug:
Początkowo opublikowane przez Snake4eva:
Graphics and sound as in your example can be contextualized. A pixel graphics or 8-bit sound game won't be judged by the standards of a blockbuster because it's genre is contextualized. I do agree that most persons aren't professional reviewers but we can agree that some people at least try to give decent reviews while others don't even bother.

Exactly, the ones that do give proper reviews and put at least some effort into it, are doing that right now and would do so with your suggestion.
The ones that don't put in any effort, won't do so any way, regardless of how the rating system looks like.

If someone writes "worst game ever" as in your example, they would probably give a 0/10 with your suggestion. How is that different? How is that better?
Both things have basically zero meaning and give zero insights.
The difference is you can not recommend a game for a number of reasons, you could even like a game and not recommend it. If you give a game a 0 then its justification better be meaningful or else it is obvious you are trolling. Not recommended and "worst game ever" is not the same as 0 and "worse game ever" especially if there are categories involved. To objectively give a game 0 would require justification why all of its elements, graphics, sound etc were given 0. Likewise its highly unlikely that two individuals who give the same numeric score gives all categories the same score or even judge based on the same categories. Not recommended and "worst game ever" is ambiguously meaningless while a 0 and "worst game ever" comes off as lazy and no effort placed in the review. People who give a lowest or highest score should provide the most justification. In a binary system a review is not even necessary since all scores are automatically the highest (recommended) or lowest (not recommended) score
Snake4eva 11 stycznia 2021 o 15:16 
People who give a lowest or highest score should provide the most justification. In a binary system a review is not even necessary since all scores are automatically the highest (recommended) or lowest (not recommended) score. Steam's system is a test taking system where you either pass or fail. Most practical systems have boundaries for pass or fail and degrees of excellency or failure. Person who give lengthy reviews with recommend or not recommend are essentially dissecting the review into categories which are better suited for a numeric review. The most helpful reviews are the detailed ones and if persons are going to lengths to write extensive reviews then a numeric system could save the user the time to just glance a review score to gauge the quality of the review.
Darren 11 stycznia 2021 o 16:00 
So you want people to provide numeric ratings on multiple criteria, but you don't want those ratings to actually mean anything. Taking RimWorld as an example it is an amazing game, but graphics and sound are definitely not it's strong suit if you contextualise those to allow for them to be anything more than 1 or 2 you are making those numbers absolutely meaningless for the purpose of using them for any assessments.

If someone is looking for games with really good graphics they aren't going to be happy to see a game like RimWorld at the top of the list with lots of 4s and 5s simply because the game itself is so awesome that nobody cares about the graphics. Conversely you'll find it difficult for people to objectively give RimWorld penalties for it's graphics because they don't care and feel it's an awesome game so you will get people ignoring the scale and giving it the maximum because it's that awesome a game (and they don't want it to be below what are to them clearly worse games).

Recommendations get around this problem entirely because it's just a do you or do you not recommend this game to your friends? And then you can explain why, and that explanation will never even touch on graphics if they aren't important to you, or might briefly mention "Yeah the graphics aren't great but you'll never notice because of ..." followed by a list of all that is amazing about the game.

To be clear ratings and recommendations are very different things. If you were to rate RimWorld using any objective scale it would never score highly, but it is an extremely fun game that most of the people playing it get hundreds of hours out of hence why it gets so many recommendations.
K O N J I G 11 stycznia 2021 o 16:59 
nice
Start_Running 11 stycznia 2021 o 18:11 
Początkowo opublikowane przez TKOdarkshadow:
Początkowo opublikowane przez Start_Running:
This gem again.
This system would be rather redundant to answer the question you are asked. If you want to rate a game ona 1-10 scale. You can do so on metacritic. If you want to use a X/Y score you can do so in the text of your review.
Metacritic doesn't require you to own the product, so it's effected by review bombing with ease.
Also, rating it in your review doesn't affect the actual score, so it's not really helpful. :lunar2019deadpanpig:

Having a 1-5 star system or 1-10 rating scale would only result in more accurate ratings, for slightly more time.


And what makes you think the current system is inaccurate?
Can you show an example?
because as far as I can see it is quite accurate in representing people's answers to the posed question. Remember- The Question is *can/do you recommend this game?' not 'How much do you recommend this game?"

TKOdarkshadow 11 stycznia 2021 o 18:21 
Początkowo opublikowane przez 76561198043285599:
Początkowo opublikowane przez TKOdarkshadow:
Metacritic doesn't require you to own the product, so it's effected by review bombing with ease.
Also, rating it in your review doesn't affect the actual score, so it's not really helpful. :lunar2019deadpanpig:

Having a 1-5 star system or 1-10 rating scale would only result in more accurate ratings, for slightly more time.


And what makes you think the current system is inaccurate?
Can you show an example?
because as far as I can see it is quite accurate in representing people's answers to the posed question. Remember- The Question is *can/do you recommend this game?' not 'How much do you recommend this game?"
It's effectively rating every game as 0/10 or 10/10, of course it's not accurate. There's no reason not to make it more accurate, since people that won't care can still resort to 0/10 or 10/10 ratings. :gk1_skull:
Darren 11 stycznia 2021 o 18:44 
Początkowo opublikowane przez TKOdarkshadow:
Początkowo opublikowane przez 76561198043285599:


And what makes you think the current system is inaccurate?
Can you show an example?
because as far as I can see it is quite accurate in representing people's answers to the posed question. Remember- The Question is *can/do you recommend this game?' not 'How much do you recommend this game?"
It's effectively rating every game as 0/10 or 10/10, of course it's not accurate. There's no reason not to make it more accurate, since people that won't care can still resort to 0/10 or 10/10 ratings. :gk1_skull:

No it's asking a very different question. Do you recommend the game doesn't mean you'd give it a 10 (or any other specific number) / 10 on a rating scale. You can recommend a game that on your objective rating scale would be a 2, because of all the things it does well while acknowledging it's flaws.

Trying to imply a rating from the answer to a very different question is obviously going to lead to confusion. The question being asked gives (in my opinion) a much more useful result which is I know how many people actually liked the game enough to recommend it (in spite of or even because of it's flaws) and therefore an idea of the odds that I'll like the game (if I like those sort of games).
DiceDsx 11 stycznia 2021 o 18:44 
Początkowo opublikowane przez TKOdarkshadow:
Początkowo opublikowane przez 76561198043285599:


And what makes you think the current system is inaccurate?
Can you show an example?
because as far as I can see it is quite accurate in representing people's answers to the posed question. Remember- The Question is *can/do you recommend this game?' not 'How much do you recommend this game?"
It's effectively rating every game as 0/10 or 10/10, of course it's not accurate. There's no reason not to make it more accurate, since people that won't care can still resort to 0/10 or 10/10 ratings. :gk1_skull:
Except it's not: "recommend" doesn't mean 10/10, it simply means "you should buy this because..."

It's like a friend telling you "You should go to this restaurant because this dish is great, the atmosphere is good and the people are nice". It doesn't mean the place is a 5 Star restaurant, right?
TKOdarkshadow 11 stycznia 2021 o 18:51 
Początkowo opublikowane przez Darren:
Początkowo opublikowane przez TKOdarkshadow:
It's effectively rating every game as 0/10 or 10/10, of course it's not accurate. There's no reason not to make it more accurate, since people that won't care can still resort to 0/10 or 10/10 ratings. :gk1_skull:

No it's asking a very different question. Do you recommend the game doesn't mean you'd give it a 10 (or any other specific number) / 10 on a rating scale. You can recommend a game that on your objective rating scale would be a 2, because of all the things it does well while acknowledging it's flaws.

Trying to imply a rating from the answer to a very different question is obviously going to lead to confusion. The question being asked gives (in my opinion) a much more useful result which is I know how many people actually liked the game enough to recommend it (in spite of or even because of it's flaws) and therefore an idea of the odds that I'll like the game (if I like those sort of games).
They should just change the system anyway, the vast majority of reviews are from people you don't know. Especially with more obscure games, you might not even have any friends who've put out a review.

Also even then, you can easily slap on "how much do you recommend a game?" Onto it, not much really changes. :calm_creep:
< >
Wyświetlanie 16-30 z 153 komentarzy
Na stronę: 1530 50

Wszystkie dyskusje > Fora Steam > Steam Community > Szczegóły wątku
Data napisania: 11 stycznia 2021 o 7:39
Posty: 153