Let's talk rating systems

Ska_baron wrote: I'd prefer a 4 point scale. Something's either:
1) terrible
2) bad side of mediocre
3) good side of mediocre
4) great

And yes, then you're writing fills in the details and nuances more than a decimal place would.

Another vote for a small number of even-number of ranking bins! The reason you see so many 7s is because it's the safe bet. You're not really saying anything with a 7 on a 1-10 scale. You're saying it's above average, but not great. Then you use it so much that 7 is the new average. You might as well be using a 5-10 scale at that point.

Using a small number of even number of ranking bins like Ska_baron put out there forces you to make a choice. There is no safe middle. You need to commit to it being better than average or below average, and you need to justify it. 1-star, 2-star, 3-star, 4-star, no fucking half-stars. You provide the half-stars and the reasons in your writing, where it belongs.

Like Nate I love ranking games on BGG too. But if I'm going to use a 1-10 scale than 5 is the average. If it isn't I'm mis-using the scale and diminishing its usefulness. Looking at my ratings on my profile page I've got a pretty good bell-curve thing going. I looked at Nate's ratings too out of curiosity and I see a fair-looking bell curve there too. And it pleases me.

If I were to setup a rating system it would be:
-Bad
-Meh
-Good
-11

Bad is bad, there aren't many games that would hit there, but there are games that I feel should have never been printed.

Meh games are the games that fall into that "7" region. Games that are designed well but no talons that latch on and draw me back. Ones that I didn't hate playing, but ones that I won't seek out again.

Good are the games that I'd consider obtaining a copy for myself because I'd like to have it around and at hand.

11 are the games that are the Top Shelf, Crème de la crème, Keepers, Perma-shelf. Those games that you'd go to and save if your house was on fire (assuming the kids were safely out).

So it'd be a three-star system with a potential absence of stars as a fourth option.

5 is the fewest I'll work with.

I use 5 for my movie reviews and I don't really like it. I know I would hate out of 10 and 100 even more tho.

I think a thumbs system is perfect. Something is either good or it isn't. If you're in the middle, then it isn't good. You then use qualitative notes to describe how much of a thumbs up it is. I would give Gravity and Hunger Games both a thumbs up but they're not exactly in the same league.

Then, as an aggregate, you could really get telling information by what the percentage of people (you trust) have given thumbs up. The only problem with this is that the divisive stuff will get really underrated (like something like GTAV which some people like to give thumbs down but when you aggregate actual scores, it comes close to 100 and is very much GOTY).

I'm on the same page as VonTush, if and when I bother to rank games, here's how I break them down on BGG:

9 - Fun/Interesting

5 - Okay, I'll play.

1 - Kindling

I've also given a few games a 10, but it truly has to be a game I never tire of playing.

Given the choice, I wouldn't rate games. It detracts from what's being said in the review, and if the reader's been paying attention and it's a good review, the reviewer's opinion should be clear.

That said, I think it can help to separate (paraphrasing Ska Baron) "good side of mediocre" and "bad side of mediocre". And that's not a small thing, because in honesty most games fall into one of those categories.

I'm against binary rating systems because they give little to people who don't agree with the reviewer's steer. If you have similar tastes to a critic you can trust their binary ratings, and you've got a clear indicator. But if not, a thumbs down tells you virtually nothing about whether you yourself might like the game.

Whenever I rate games myself, I fall back on a 5-star system in my head and then multiply it by whatever's required to fall into the publishers review scale. But the key thing is that three stars is good because games are meant to be good. A lot of people, especially those dishing out sevens to average titles on BGG, ought to remember that.

Computer Gaming World, back when they were still alive, had one of the best descriptions of their 5-star system:

5 - Outstanding. The rare game that gets it all right. A must-play experience.
4 - Very Good. Worthy of your time and money, but there are drawbacks.
3 - Average. Either an ambitious design with major flaws, or just vanilla.
2 - Seriously lacking in play value, poorly conceived, or just another clone.
1 - Abysmal. The rare game that gets it all wrong. Pathetic. Coaster material.

I also seem to recall someone else (maybe an older CGW - I subscribed to it for forever) had a good description of 3 Stars. It's worth your time if it's a playstyle or subject that you really like. For example, Ticket To Ride: Middle of Nowhere is good enough that a TtR fanatic will really like it, but a casual fan will think it's merely okay.

wadenels wrote: Like Nate I love ranking games on BGG too. But if I'm going to use a 1-10 scale than 5 is the average. If it isn't I'm mis-using the scale and diminishing its usefulness.

This would only be true if you played games indiscriminately, like a professional reviewer. Almost all of us have a selection bias that prevents us from playing games we don't think we'll like.

The problem isn't that everyone rates games a 7 (well, it partly is, but I'm making a point); the problem is that very few people bother to play and rate the 4s.

I've often told myself I was going to switch all of my ratings to a 5-point scale like Mark Johnson, but I just get intimidated by the task. Hell, I'm still way behind on rating and commenting my current games, let alone everything in the past.

Bull Nakano wrote: I mention Mark Johnson's (5 star) system because he describes it as avoiding 7's and encouraging 10's. I like the concept of committing a game in that way, thus avoiding half stars, and having to decide if a game is a 6 or an 8.

I use a 5 star ratings system. It works because I aggregate scores from the entire group over 3 plays.

Here's my ratings system:
superflycircus.blogspot.com/2011/12/this...ow-we-dooooo-it.html

Here's some personal commentary on how I, personally, rate games, not how the Circus rates games:
superflycircus.blogspot.com/2012/07/pete...us-game-ratings.html

I do want Sag does..I'm trying to move towards nothing having the same ratings, so it is a true ranking. In principle, everyone's collection (whether it is what they own or games they have played) could be reduced to a ranking by asking for every pair of games, "given the opportunity to play Game A or Game B under optimal conditions, which would you rather play?" If everybody did this, the system on BGG would be really useful. As it is now, it is borderline useless.

Gregarius wrote: The problem isn't that everyone rates games a 7 (well, it partly is, but I'm making a point); the problem is that very few people bother to play and rate the 4s.

This is a really good point. On a 10 pt scale, most games I'm interested in playing tend to fall into the 6-8 range due to taste. What I like about the 5 pt scale is it differentiates the 6's from the 8's and, for me, puts my most common rating at 6. But I make it a point to rate every game I play, I try to comment on every one, but I'm a bit behind.

I also like that the 5 pt scale makes a "perfect" score, for me, more common. I feel often people put a strong emphasis on games they give a 10/10. "Go is my only 10." and I think removing the sanctity of "perfect" allows more of a personality to come through in ratings.