>> Rating calculation method: ARL

[snifner]

Last thought:

If everyone on this site entered all 3's 90 % of the time, in 6 months the lowest score would be 2.5 and the highest score well be 3.5 and thats it.

[snifner]

Ok problem solved.

Just make it a 4 star system. Then there is no star for being average.

1 Star = Never took a single turn or missed too many turns & was rude ect...
2 Stars = Missed a few turns or deadbeated the end or something.
3 Stars = Above average. Didn't miss any turns and played fair.
4 Stars = Played a perfect game. Took fast turns. Played smart. Enjoyed playing with the person.
If you thought the person was average in some area leave it blank.

This is a system I can live with.

[PLAYER57832]

What seems clearer and clearer is that this rating system will just get muddier and muddier because so many people have their own ideas.

I like the "check off" system.

I like knowing how many negs someone has left.

I like automating the attendance rating.

BUT, I think just going back to the old feedback, unmoderated, other than specific language (these can even be programmed to become #@#) would be simplest and best overall.

[Hrvat]

To make the current system work there should be 3 choices.

Leave a 5 if you feel played well, ect.
Leave a 1 if they played like a fool or missed a lot of turns.
And leave nothing if you think they are average.

(then their number well be an average of the good and bad..)

It just seams stupid to give everyone a 3.

[Lexitonia5]

Looks very complicated but the end result should be great. I think its a good idea. I am guilty of leaving fives all the time. In my next life in CC I am going to name myself "happyhappyjoyjoy1"

[BBoz]

Seems to me that this just serves to over-complicate things. If you just indicated the average rating left in each user's profile, someone who really had to get down into this level could. This seems way too confusing to joe-average player.

I also think you lose balance on this when you have someone who makes a point of only trying to play with good players. Their ratings are going to naturally skew high.

[kbkamka]

1-Poor
2-Bad
3-Average
4-Good
5-I'd kiss you

Make it an average rating like initially suggested. You're right, you'll end up with a tighter variance which may not mean much in the end.

To solve this problem, I'd then figure out a way to make the more experienced ratings count for more and the opinions of noobs like myself count for less. Also make the frequency of ratings count and if you wanted to, even the win percentage of the player leaving the rank.

So you'd end up with 3/3.5 for my rating of an average player, lets just say. The initial '3' is my rating for Player X. The 3.5 would be my average rating. So I'd give this player a -0.5 versus the majority of the players I've played in my past. Take the -0.5 and multiply it by how experienced of a player I am. (# of games, premium, and winning percentage should all count)

Then let the ratings system fly. You might end up with some outliers in the beginning, but after a while, the more experienced players ratings will dominate the system which will hopefully give an approximately fair representation of the player being rated's level.

Just a thought.

[ueli]

I did not go thru all the responses, so please forgive me if I repeat what has already been said.

BUT, If I understood the proposition right, then to leave an average rating of 3 will augment the impact of my 1 and 5, right? And with that you hope to bring players to leave more 3 for average play? Well, it will be just as easy to leave anough 1 and 5 so that on average you get........... a 3.

Is that the goal then? That for every 5 I leave i'll leave also a 1?

I usually only rarely leave neg ratings, 90% of my ratings left are for people I had a good game with. Now I would be forced to leave also 1.... right now those that deserve a 1 get a 1, but usually for stupid players I just put them on ignore...

[edgarhons]

I give the rating system a 1 on any scale.

Quit trying to make 3 the average score, it was a poor decision in the first place and now it's being treated as a sacred principle for no good reason. If someone doesn't screw up a game in any significant way, how is that not 'Excellent'? If everyone in a game shows up for every turn, it's both average and excellent, is it not?

I think it's also fair to say that you're a lot more impressed by teammates than opponents. If one of my teammates plans an awesome move and wins the game for us, I'd see the results and be very excited, thus more likely to leave the highest rating. But if the other team does the same and beats me, A. I probably don't notice, and B. If I do, I feel dejected and usually cry a little. What exactly constitutes an 'Excellent' beatdown from an opponent anyways? If they played fair, they don't deserve any less recognition than my partners, so I think it's better that I use 5 as a good-game standard than 3.

The system's been around for only a few weeks and I'm averaging only about one 5.0 per game. The rest are scattered around the 4-5 range, and the system is working well enough that I can see sub-4.0 players in games awaiting players and can avoid them.

What about adding a super-star that you can only give to 1 person per game?

[lackattack]

I'm getting to this part of the Ratings Reloaded project and I'm having second thoughts. It won't fully solve the problem - some people will still typically leave 5's and others will still typically leave 3's and get complaints about it. Also the solution is quite hard to explain and understand.

I'm looking at other suggestions brought up here:

* 4 stars (1-Very Bad, 2-Bad, 3-Good, 4-Very Good) with no "average", so people could typically leave 3 with less controversy.
* 3 stars (1-Bad, 2-Good, 3-Very Good) so that "average" would be more positive sounding as "Good".
* 4 stars with a cap on how often you can rate 1's or 4's. For example, only one "super" rating per game to force ratings to be more moderate and therefore (hopefully) accurate.

What do you guys think? I need some more feedback!

[trapyoung]

the biggest issue is that a rating appeal system is needed, it's not just when people leave 1's for attendance when you don't miss a turn that's an issue, how about a 1 for fair play or 1 for attitude where the game is a 1 on 1, there are no issues or confrontations and the only words left in the game chat are "hey, good luck" and "gg" - i have four such ratings and i know most people on the site do as well. a rating moderator needs to be introduced or a system needs to be instituted, similar to new recruits and team games, where you aren't allowed to use the rating system until you've been adequately introduced to the site because it is mainly new recruits, privates, cadets, and cranky low ranked vets who leave the inaccurate feedback.

[ParadiceCity9]

So this is suggesting that if we give mostly fives, it's going to hurt our ratings? What if we actually think people deserve the 5...

[Ffraid]

When this 5 star system first went into effect, I thought 5 stars was overkill. I've always hated those surveys, "Do you: 1.Strongly Agree, 2.Agree, 3.Neither Agree nor Disagree, blah, blah, blah?" So, I thought a 3 star system would be better, but then this thread was started and I thought, "Well, that might work." It wouldn't work on a 3 star system, however, so I tossed that idea out the window. But, if you're considering other ideas, I definitely think it should be lowered from 5.

As to 3 or 4 stars, I just assumed 3. I mean, seriously, if you don't like something, do you ever not like it just a little bit? Has anyone received a 2 rating? I don't believe I've seen one, but I haven't been perusing the ratings all that much (Actually, I haven't been playing much lately, but for personal/time reasons, not because I'm boycotting the site or anything ridiculous like that. Still love the site, even though I don't like the new rating system).

On the other hand, I have seen 4s. I think the problem with a 3 star system is that you're going to see mostly 1s and 3s, just like most feedback was either positive or negative. I think people would regard a 2 as neutral, which would be fine with me, but it doesn't sound like that's what you're looking for. It sounds like you would like to see mostly 2s and I just don't think that that's what would happen.

Of course, the problem with a 4 star system is that I don't think you'll see any 2s. It goes back to not liking something "a little bit." If someone has a bad attitude (or is perceived by the rater to have had one, anyway) is anyone going to stop and think, "Well, it could have been worse," or are they going to go straight for the 1 radio button?

Personally, I guess I'd prefer a 3 star system. Fewer shades of gray = better, imo. Actually, I'd prefer jiminski's idea of community moderated feedback, but that discussion doesn't seem to be on the table.

[BBoz]

I am of the opinion that you should simplify all this. The real reason people get so upset seems to be their numbers slowly going down and the need to compare your score to your buddies scrore. (I am guilty of it) The purpose of ratings is to find the good people to play with vs. the twits. Rather than stars and number specifics, I suggest using something like a smiley face / color rating scale. (although ultimately choosing different icons would be great). Stars are so subjective, but happy faces are pretty easy to read. And, if you replace the "number" rating, it would remove a lot of the competitive comparing of "ratings scores" that go on in people's heads.

So I suggest 4 like so... When a person gives a rating, have the choices something like this...

VERY BAD BAD GOOD VERY GOOD

People could select the ratings and you use that to calculate a score which then translates into an icon that sits next to the person's name, like the star does. But instead of a star, it would be one for Gameplay (I like the for this) one for Attitude (smiley faces) and one for attendance (stopwatch?).

Scale them based on color rather than put numbers next to them that everyone compares. Great! (Green): Good (Yellow): Not so Good (orange): Awful (Red): . (I would make Awful solid Red).

This makes them easy to tell, but not so specific that everyone has their rulers out comparing size. Another added advantage to this, you can keep the formula for the calculations hidden and "tweak it" if you need to based on how the community does the ratings.


I would like to see the Attendance piece automated. Have the system track how many turns you miss averaged across your games. Then use this to calculate your attendance rating. Takes all the subjectiveness out of it.

For added "depth" on the ratings you could do a glowing icon for one that is in danger of changing to the next lower level.

Simple, less "comparing" to your buddy on specific numbers, and flexible (formula can be changed as needed behind the scenes).

[trapyoung]

the problem with the new rating system is that the old one you could describe attitude & caliber of player, the new one you are supposed to only rate attitude, the caliber of player is shown by rank, but too many kids are sore from losses and rate all 1's and similar things. if you rate someone for a poor attitude, attendance, or low on fair play (ie, a 2 or lower) you should be able to back it up should you be required to be to a sorely needed rating mod.

[KidWhisky]

I dont think that there should be a limit on how many 5/4/whatever the best rating is. I just finished a game with a few guys that was a great game. It took us months to play out because we all played it really well. I would hate to finish a game like that and not be able to leave the other players the ratings I feel they deserve.

[leolou2]

i saw the new rating tablet the only thing missing is player got a bad flop then it would be complete . thank you for your time

[e_i_pi]

Just a quick note to the people who think this over-complicates things... here's the formula to calculate the rating left after ARL is taken into account:

=Max(Min(Max(Min((3-ARL),2),-2)+RL,5),1)

...where...
ARL is Average Rating Left (by the rating player)
RL is Rating Left (for the rated player)

[ZeroDJoe]

great ideas in this thread!
lack make it happen!
and i dont think some people will keep leaving 5* to all, cause theyll realise that it will only make them not be able to actually giva a good rating to someone... maybe it will take some time but the 3* average rating will be what averyone will want to leave as an average
not sure if a rating reset isnt need to make this change