Recommend
453 
 Thumb up
 Hide
103 Posts
Prev «  1 , 2 , 3 , 4 , 5  Next »   | 

BGG» Forums » Gaming Related » General Gaming

Subject: Rating with confidence - going TOTALLY NUTS wth statistics! rss

Your Tags: Add tags
Popular Tags: stats [+] statistics [+] bgg [+] ranking [+] rating [+] [View All]
Re: Rating with confidence - going TOTALLY NUTS wth statisti
vernicus wrote:
Yeah, I get the histogram justification, but it doesn't seem like it would be that difficult to just show a bar for every distinct rating entry. But then again, I only use halves. If someone tried to create distinct rankings with their ratings (going back to your comparative statement analysis -- see how I tie it all together?) by going so far as rating 8.01, 8.02, etc. that could be a bit taxing as a histogram, eh?
I know of several people who frequently rate out ot the second decmal place.

Quote:
So do you know about my other question? Have the BGG ratings always been integral but allowing reals?
Ratings have been reals for at least as long as I've been on the 'geek (4 years now?).
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Mike Windsor
United States
Fort Worth
Texas
flag msg tools
badge
Avatar
Microbadge: Calvin and Hobbes fanMicrobadge: Calvin and Hobbes fanMicrobadge: Strong to the Finish, 'Cause I Eats Me Spinach!Microbadge: René DescartesMicrobadge: I like tacos
I'm just not smart enough to comment on much of this, but I have a few observations for what its worth:

While some people may skew the results with a 10, there are some people who never give out a "perfect 10."

People may only rate games that they feel strongly about. If I really like a game, I may want to take the time to give it a high rating; if I hated it, I may want to take the time to give a low rating. If I'm so-so about a game, I may not take the time to rate it. Therefore, some people's ratings may not contin many games in the middle.

The 1 - 10 rating system may be too precise. Most things are only rated on a 1 - 4 or 1 - 5 system. A game I rate as a 4 one day could be a 6 on another day. Maybe the ratings ought to be:

5 - One of my top choices.
4 - I'd rather play this than many other games.
3 - Doesn't do much for me, can play it or leave it.
2 - I'd rather not play it.
1 - I don't want to play this one again.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Philip Thomas
United Kingdom
London
London
flag msg tools
badge
Avatar
Microbadge: Medieval Theme fanMicrobadge: Isleworth BoardgamerMicrobadge: Golden MeepleMicrobadge: EurogamerMicrobadge: Roman Empire buff
I rate every game I play, if I can. I suspect a lot of geeks do that. Moreover, most people will have more ratings in the middle than at either end, so it will even out.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dominic Crapuchettes
United States
Bethesda
MD
flag msg tools
designer
publisher
North Star Games designs party games that don't suck! Play them with your non-gamer friends over the holidays.
badge
First there was Hearts, then there was Spades, and now we bring you Clubs. The suit of clubs finally gets some respect!
Avatar
Microbadge: 2015 Ultimate SupporterMicrobadge: 2016 Ultimate SupporterMicrobadge: 2017 Ultimate SupporterMicrobadge: 2018 Ultimate SupporterMicrobadge: 2019 Ultimate Supporter
Re: Rating with confidence - going TOTALLY NUTS wth statisti
Nice work Joe!

Joe Grundy wrote:
The rankings may be adversely affected by people "gaming" the ratings system by adding multiple user accounts.
Many of the users who only rate one game and give it a "10" are probably friends and family of the game designer and (wrongfully) thinking they are doing them a favor.

There are two people who have rated Wits & Wagers a "10" and who have also not rated any other games. This is because Satish and I encouraged people to rate Wits & Wagers on the BGG after we hosted our game show events at conventions. A few people from Origins and GenCon had never heard of the BGG. We encouraged them to check it out. It did not occur to us that they would rate our game a 10 and never come back. I would remove their ratings if I could.

Furthermore, a classmate of Satish and myself found the BGG on their own when searching for our game. He promptly rated Wits & Wagers a 10 and started trouble by trolling in one of the W&W forums. I tried several times to get him to tone it down. Predictably, a prominent member of the BGG accused me of creating his account. soblue

So I chalk up a lot of those accounts to misdirected friends and family who think they are being helpful.
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Regarding the "double spike" effect, with a pronounced number of "extra 4s":

I think this may be due to two factors: 1) BGG's wording of the description for each rating integer, and 2) BGG rounding any decimal point up to the next integer (e.g., 8.1 = 9 on the bar graph).

I have two spikes on my bell curve of ratings. A "natural" one at 7, and one that's a little out of place at 4 (although this latter spike is quite small).

Per the first point above, I find I'm less likely to feel truly neutral about a game -- thus rating it a 5 -- than to lean a little toward liking or disliking it (rating somewhere between 4 and 6, but not equal to 5). Also, my 4s are boosted as an effect of the second point since I'm more inclined to rate something a 3.5 than a 4.5. That is to say, I have relatively few game that I "dislike only a little" (rated a 4.5) compared to the number of games that I definitely dislike, but don't quite fit my view of a 3 or 4 (and are thus rated 3.5).
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joe Grundy
Australia
Sydney
NSW
flag msg tools
badge
Avatar
Microbadge: Pregnant - baby on boardMicrobadge: It's a Boy!!Microbadge: Pregnant - baby on boardMicrobadge: Microbadge: Offline from The Geek for a while
rootbeer wrote:
Regarding the "double spike" effect, with a pronounced number of "extra 4s":

I think this may be due to two factors: 1) BGG's wording of the description for each rating integer, and 2) BGG rounding any decimal point up to the next integer (e.g., 8.1 = 9 on the bar graph).
Jesse you're on a winner with point (1). Regarding "8.1=9" ... I'll just note that in all charts above I've rounded to nearest integer so 8.1 would be 8. However, 3.5 would still be 4.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Brian Diggs
United States
Vancouver
Washington
flag msg tools
Microbadge: Wooden cube fanMicrobadge: I play with blue!Microbadge: Gamestorm attendeeMicrobadge: Component BaggerMicrobadge: Copper Board Game Collector
Re: Rating with confidence - going TOTALLY NUTS wth statisti
This is a very interesting discussion. I have only one nit to pick. In the section "The Rating Game #1 - Multiple Accounts", all the graphs with "Number of Ratings per User" as the horizonal axis seem to have a categorical axis rather than a continuous axis. In other words, they should be X-Y plots rather than line graphs. Otherwise, it was really very detailed and well done.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joe Grundy
Australia
Sydney
NSW
flag msg tools
badge
Avatar
Microbadge: Pregnant - baby on boardMicrobadge: It's a Boy!!Microbadge: Pregnant - baby on boardMicrobadge: Microbadge: Offline from The Geek for a while
Trollsfire wrote:
This is a very interesting discussion. I have only one nit to pick. In the section "The Rating Game #1 - Multiple Accounts", all the graphs with "Number of Ratings per User" as the horizonal axis seem to have a categorical axis rather than a continuous axis. In other words, they should be X-Y plots rather than line graphs. Otherwise, it was really very detailed and well done.
Sorry ... I didn't explicitly notate that those are logarithmic x scales. But it does suffer the minor visual distortion that the n is integer and hence below n=15 there's one notch per n. I should try to redo those charts to visually correct the scale for this problem.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Marshall Miller
United States
Efland
NC
flag msg tools
designer
publisher
The Warren is a roleplaying game about intelligent rabbits trying to make the best of a world filled with hazards, predators and, worst of all, other rabbits.
badge
Marshall is a Boston-based researcher and game designer.
Avatar
Microbadge: The Warren RPG fanMicrobadge: Dogs in the Vineyard RPG fanMicrobadge: Plays Games with SpouseMicrobadge: PsychologistMicrobadge: Duke University
Wow!

Keep up the really really great work.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Magnus Lundgren
Sweden
Göteborg
Unspecified
flag msg tools
designer
badge
Avatar
Microbadge: Treasure Hunter fanMicrobadge: Cthulhu fanMicrobadge: Zendo fanMicrobadge: I play with yellow!
Re: Rating with confidence - going TOTALLY NUTS wth statisti
Quote:
These are largely the users forcing their ratings into a "balanced" distribution, ie giving roughly the same number of 1s as 10s, the same number of 2s as 9s and so on, often in a triangle.
Oh, that would be me, but I swear I didn't do it on purpose and it wasn't really me, and I was drunk and come to think of it you don't have any solid evidence anyway.
But I'm sorry that I am one of those bad, bad, bad people that rate over the whole scale and happen to have a fine bell curve shape to my ratings, and I think that Mamma Mia is an abomination, and should be ritually burned on each and every gaming convention
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Ken Just
United States
Yorba LInda
California
flag msg tools
badge
Avatar
Microbadge: Copper Session ReporterMicrobadge: Railroad Tycoon fanMicrobadge: Electric Light Orchestra fanMicrobadge: Star Trek fanMicrobadge: Engineer
I wrote an article several years ago were I tried to balance a games rating with the number of people who owned the game. I tried to make the effect of ownership and ratings equal. Not being a stats geek I'm not sure how accurate my rating was but it gave, I think, a better indication of the best games for the general gamer. Here were the top 50 for Oct 2005.

Game --------------- Owners - Rating - Overall Game Rating

Puerto Rico ------------- 4791 - 8.61 -- 19710
Settlers of Catan ------- 5609 - 7.89 -- 19037
Carcassonne ------------ 5276 - 7.62 -- 16461
Euphrat & Tigris -------- 3171 - 8.36 -- 12243
Ticket to Ride ---------- 2932 - 7.78 -- 9602
Bohnanza --------------- 3272 - 7.35 -- 9325
Citadels ---------------- 3072 - 7.44 -- 9032
Lost Cities ------------- 3068 - 7.44 -- 9020
Princes of Florence ----- 2304 - 8.16 -- 8426
Acquire ----------------- 2665 - 7.62 -- 8315

El Grande --------------- 2309 - 8.10 -- 8310
Lord of the Rings ------- 2546 - 7.25 -- 7002
Memoir '44 -------------- 2002 - 7.89 -- 6795
San Juan ---------------- 2146 - 7.64 -- 6738
Power Grid -------------- 1834 - 8.16 -- 6709
RoboRally --------------- 1992 - 7.72 -- 6414
Chess ------------------- 2460 - 7.07 -- 6322
LotR - The Confrontation 1931 - 7.71 -- 6195
Modern Art -------------- 1859 - 7.82 -- 6176
Tikal ------------------- 2055 - 7.44 -- 6042

Bang! ------------------- 2315 - 7.04 -- 5880
Samurai ----------------- 1820 - 7.72 -- 5853
Through the Desert ------ 1837 - 7.67 -- 5825
Diplomacy --------------- 2084 - 7.26 -- 5752
War of the Ring --------- 1620 - 7.99 -- 5649
Magic: The Gathering ---- 1975 - 7.25 -- 5431
Cosmic Encounter -------- 1886 - 7.21 -- 5111
Ra ---------------------- 1466 - 7.94 -- 5037
Go ---------------------- 1443 - 7.95 -- 4971
Alhambra ---------------- 1783 - 7.22 -- 4850

Scrabble ---------------- 2359 - 6.55 -- 4836
Civilization ------------- 1586 - 7.41 -- 4615
Torres ------------------ 1511 - 7.44 -- 4442
HeroScape --------------- 1472 - 7.48 -- 4387
Goa --------------------- 1268 - 7.94 -- 4361
Taj Mahal --------------- 1332 - 7.73 -- 4302
Axis and Allies ---------- 2015 - 6.61 -- 4252
Formula Dé -------------- 1589 - 7.14 -- 4195
Game of Thrones, A ------ 1296 - 7.71 -- 4154
6 Nimmt! ---------------- 1593 - 7.06 -- 4078

Apples to Apples -------- 1635 - 6.94 -- 3989
Wallenstein ------------- 1107 - 8.02 -- 3901
Age of Steam ------------ 1102 - 7.95 -- 3800
Ticket to Ride Europe --- 1108 - 7.88 -- 3747
Guillotine --------------- 1667 - 6.56 -- 3434
Risk -------------------- 2938 - 5.62 -- 3291
Backgammon ------------- 1528 - 6.49 -- 3041
Twilight Imperium 3 Ed -- 878 - 7.92 -- 3001
Advanced Civilization --- 906 - 7.74 -- 2938
Maharaja: PBiI ---------- 886 - 7.68 -- 2815

http://www.boardgamegeek.com/article/676968#676968
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joe Grundy
Australia
Sydney
NSW
flag msg tools
badge
Avatar
Microbadge: Pregnant - baby on boardMicrobadge: It's a Boy!!Microbadge: Pregnant - baby on boardMicrobadge: Microbadge: Offline from The Geek for a while
JustKen wrote:
I wrote an article several years ago were I tried to balance a games rating with the number of people who owned the game.
It's quite a good idea just to pick a fixed point (4.5) above which you call the game "positive". Anything up to 6.8 (the average rating by geeks) would be a viable choice and would produce various results.

I did something similar, "Total Goodwill", but using each individual's rating average as their zero point for their contribution to the grand total, rather than 4.5. It seems to work pretty well for games that you'd be glad to own the rights.
http://www.boardgamegeek.com/geeklist/20420

Picking a fixed 4.5 value would remove any bias introduced from the trend that people with fewer games rate them more highly.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Surya is pure Eurosnoot and proud of it!
New Zealand
Napier
flag msg tools
badge
Avatar
Microbadge: Tasty Minstrel Games "Eminent Domain: Exotica and Battlecruisers" Contest participantMicrobadge: Cosmodrome Games "Smartphone, Inc." Contest ParticipantMicrobadge: Ares Games "Sails of Glory" Contest participantMicrobadge: North Star Games "Evolution" Contest participantMicrobadge: Lamp Light Games "Nautilus Industries" Contest participant
Re: Rating with confidence - going TOTALLY NUTS wth statisti
jgrundy wrote:
ie if the user has two peaks in their ratings curve, they have an "unusual" bias in their ratings
Hmmm, not always. I have a peak around 3 since I rate Munchkin a 3. And since all Munchkin games are the same, I rated them all 3. That's not shilling, that's really how I feal about the game. I guess you won't notice once I've rated 10000 games
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joe Grundy
Australia
Sydney
NSW
flag msg tools
badge
Avatar
Microbadge: Pregnant - baby on boardMicrobadge: It's a Boy!!Microbadge: Pregnant - baby on boardMicrobadge: Microbadge: Offline from The Geek for a while
Surya, thanks for saying so.

There's also someone who went through and rated every single Monopoly variant they could find, since the rules are identical.

Are the various standalone Munchkin game rules all identical? I've never played them so I wouldn't know.

You correctly point out that a prolific game family may invoke a common reaction from a particular user, so ratings spikes are "unusual" but not necessarily anything bad.

Did you know expansions are not included in the game rankings? And you can tick a box to exclude them from any Advanced Searches. (Personally I think that would be slightly better if you had to tick a box to Include Expansions, but anyway.) So the effect of rating expansions is to comment on them for people who are interested in the expansion.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Surya is pure Eurosnoot and proud of it!
New Zealand
Napier
flag msg tools
badge
Avatar
Microbadge: Tasty Minstrel Games "Eminent Domain: Exotica and Battlecruisers" Contest participantMicrobadge: Cosmodrome Games "Smartphone, Inc." Contest ParticipantMicrobadge: Ares Games "Sails of Glory" Contest participantMicrobadge: North Star Games "Evolution" Contest participantMicrobadge: Lamp Light Games "Nautilus Industries" Contest participant
Re: Rating with confidence - going TOTALLY NUTS wth statisti
jgrundy wrote:
Are the various standalone Munchkin game rules all identical? I've never played them so I wouldn't know.
The basics are the same. Some of them have a few slight twists, but it never diverts a lot. That's why you can combine them all together. Imagine that: about 20 games and expansions in one game. Try shuffling that stack of cards!
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joe Grundy
Australia
Sydney
NSW
flag msg tools
badge
Avatar
Microbadge: Pregnant - baby on boardMicrobadge: It's a Boy!!Microbadge: Pregnant - baby on boardMicrobadge: Microbadge: Offline from The Geek for a while
I finally got around to assessing the impact of the Bayesian averaging calculation on how many "Game A better than Game B" assertions are matched in the rankings.

The current calculation costs 1.2% of the possible matches.
This would improve to only costing 0.5% if the 100 phantom votes were 6.5 instead of 5.5 (noting that 6.5 works better than 6.4 or 6.6). 6.5 is between the average game average of 6.2 and the average user rating of 6.8, which is as expected.

Every additional phantom vote costs a little in matches. But 100 votes costing only 1.2% of matches gives you an idea of how much. And the impact is not linear... 30 phantom votes at 5.5 already costs 0.65%, and 200 takes it to a cost of 1.6%. If the phantom votes were 6.5, then 30 phantoms costs 0.1%, 100 phantoms costs 0.5%, and 200 would cost 0.8%

Cheers all
Joe
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Heckle Jekyll
United States
Knox
Pennsylvania
flag msg tools
badge
Avatar
Microbadge: Gaslands: I follow The Rule of Carnage!Microbadge: Coffee drinkerMicrobadge: Playing until death!Microbadge: Metal Detector enthusiastMicrobadge: I don't know what I'm doing
Nice post, great read.

Regardless of the rating system used (Bayesian average or Raw Rating) I'd prefer full CONSPICUOUS disclosure. A list ranking by Bayesian Average should so state. We should also be able to view a Raw Rating Rank.

Is there a way to view the Raw Rating Rank?

 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joe Grundy
Australia
Sydney
NSW
flag msg tools
badge
Avatar
Microbadge: Pregnant - baby on boardMicrobadge: It's a Boy!!Microbadge: Pregnant - baby on boardMicrobadge: Microbadge: Offline from The Geek for a while
idiot crow wrote:
Is there a way to view the Raw Rating Rank?
Not exactly. But you can do an Advanced Search (and apply the filters you are interested in) and fill in a ratings range eg "0 to 11 with at least 10 ratings". The results are sorted by Raw Rating by default, and you can click the headings to resort them.

The raw ratings version, and also correcting using standard deviation / confidence intervals, are both open to manipulation in the few instances where someone wants to try.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Mike Jones
United States
Gainesville
Florida
flag msg tools
Yeah it's here! Really it's right here.
Avatar
Microbadge: One does not simply recognize great geeks. Not with 10,000 stars could you do this...Microbadge: Love Spreader of the WeekMicrobadge: Level 20 BGG posterMicrobadge: Chit Chatter of the DayMicrobadge: The Geek Citizenship Recognition Program Rolls On!
Re: Rating with confidence - going TOTALLY NUTS wth statisti
Where's the Executive Summary?
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Mark C
United States
Ypsilanti
Michigan
flag msg tools
Avatar
Microbadge: I´m only half evilMicrobadge: Through the Ages: A Story of Civilization fanMicrobadge: Stanley Kubrick fanMicrobadge: Out for blood - I play without mercy
snicholson wrote:

I'd like to second Joe's request for people to rethink their ratings. Don't rate for anyone else but yourself. While I struggle with the BGG definitions because they define both quality and replayability, I find using them as subjective guidelines (focusing much more on my replay interest) to be a great help. I'm certainly giving many more 4s and 3s using that scale.

Don't do it Scott!

Here's the thing: without doing any analysis, IMO the single biggest issue with BGG ratings is self-selection.

For example, Blood Bowl, which is not to my tastes, has a healthy rating of 7.5, and is in the top-100. If I for example, were to rate it on my personal tastes, It would drop 4½ points to a 2. So the community that likes that style of game has signalled it to be a very good game --sales and other measures of popularity and quality support that. BUT, I would expect it to take a large hit due to it's niche nature if all those who have seen it rated it based on their dislike of violent, dice-rolling, football spoofs. So do we want to move towards a more generic popularity contest, or let the niche games thrive? It's a legitimate question. Perhaps the most "pure" measure of a game is time spent playing a game and yet we choose a 1-10 scale with a self-defeating guideline (the more you play, the less likely you are to want to play again, and yet playing a game a lot says you obviously liked it).

IMO the self-selection aspect of the ratings is deeply imbedded, and switching would shake up the ratings considerably, and probably not in a good way. Many best-in-class games would fall down the radar.

In the end, ratings are entrenched, and you can use whatever scale you want, but I think it's better to recognize quality the way you have been.

I tend to "bump" my ratings based on how much I've ACTUALLY played a game. Obviously, if I play it more than I anticipated, there's something there I like but didn't recognize. Conversely, games I thought would be a big hit but didn't make it to the table much say they can't be that good....unless of course you want to factor into your ratings how often you're able to convince people who don't want to play the game....
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Re: Rating with confidence - going TOTALLY NUTS wth statisti
Gamer_Dog wrote:
IMO the self-selection aspect of the ratings is deeply imbedded, and switching would shake up the ratings considerably, and probably not in a good way. Many best-in-class games would fall down the radar.
Yes. This seems a fine thing for an aggregative value, especially under Sturgeon's Law.

Quote:
In the end, ratings are entrenched, and you can use whatever scale you want, but I think it's better to recognize quality the way you have been.
Which is what I do. I rate as if the scale were exponential and the midpoint near 3.

Quote:
I tend to "bump" my ratings based on how much I've ACTUALLY played a game.
I don't as play rate is much more a strange product of individual willingness than interest in the game. At least two of the groups here regularly end up playing a game not because that's what any of them want to play, but because it is near the only title they are all actually willing to play. Their tastes and preferences tend to vary and disagree widely, leaving the LCD often the winner.

Quote:
Obviously, if I play it more than I anticipated, there's something there I like but didn't recognize.
Go figure. I played To Court The King this weekend. I'd played it at Kublacon and determined that that was a mistake I wouldn't repeat. I played it again this weekend, disliking it equally, simply because I knew my kids would accept it and it would occupy a potentially contentious time slot. TCtK has not improved in my opine even slightly due to increased play rate.

Quote:
Conversely, games I thought would be a big hit but didn't make it to the table much say they can't be that good...
No, you have the wrong audience for that game. Change the audience to suit and the game will be played more often. Context.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Gamer_Dog wrote:
snicholson wrote:

I'd like to second Joe's request for people to rethink their ratings. Don't rate for anyone else but yourself. While I struggle with the BGG definitions because they define both quality and replayability, I find using them as subjective guidelines (focusing much more on my replay interest) to be a great help. I'm certainly giving many more 4s and 3s using that scale.

Don't do it Scott!

Here's the thing: without doing any analysis, IMO the single biggest issue with BGG ratings is self-selection.

For example, Blood Bowl, which is not to my tastes, has a healthy rating of 7.5, and is in the top-100. If I for example, were to rate it on my personal tastes, It would drop 4½ points to a 2.
So... you're saying everybody should ideally rate every game, but since that's not possible you should rate a game based on how much you think everyone else would like it?

I think this notion of "objective" ratings is what leads to the entrenched (or at least skewed) rankings you spoke of. People think, "Well, I don't love Puerto Rico/Age of Steam/Princes of Florence, but other people sure do. I'd better rate it a 9."

(This is also the only reason I can see for Finding Neverland [http://www.imdb.com/title/tt0308644/] being ranked as the 213th best film of all time. Well, that and the magic words "Johnny Depp" and "Oscar nomination.")

"Niche" games should take a hit. If you've got some oddball game that only one guy out of a hundred is going to like, the other ninety-nine people shouldn't artificially boost their rating out of fear that they'd dissuade Mr. One Hundred from trying the game.

---

On another note:

clearclaw wrote:
I rate as if the scale were exponential and the midpoint near 3.
I agree with this statement more than anything else you've ever said, J C. I think of 5 as the mid-point, but that's not important. I'm glad to see this concept verbalized by someone other than me.

Tangentially, I think viewing the 1 to 10 ratings as a linear scale is the main problem with the current correlation tools. The distance between 8 and 9 is much further than the gap between 6 and 7. i.e., if someone gives a 8 to something I rate a 9, I disagree with him more than the fellow that rates a different game a 7 while I think of it as a 6.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Mike Jones
United States
Gainesville
Florida
flag msg tools
Yeah it's here! Really it's right here.
Avatar
Microbadge: One does not simply recognize great geeks. Not with 10,000 stars could you do this...Microbadge: Love Spreader of the WeekMicrobadge: Level 20 BGG posterMicrobadge: Chit Chatter of the DayMicrobadge: The Geek Citizenship Recognition Program Rolls On!
Re: Rating with confidence - going TOTALLY NUTS wth statisti
clearclaw wrote:
Gamer_Dog wrote:
IMO the self-selection aspect of the ratings is deeply imbedded, and switching would shake up the ratings considerably, and probably not in a good way. Many best-in-class games would fall down the radar.
Yes. This seems a fine thing for an aggregative value, especially under Sturgeon's Law.

Quote:
In the end, ratings are entrenched, and you can use whatever scale you want, but I think it's better to recognize quality the way you have been.
Which is what I do. I rate as if the scale were exponential and the midpoint near 3.

Quote:
I tend to "bump" my ratings based on how much I've ACTUALLY played a game.
I don't as play rate is much more a strange product of individual willingness than interest in the game. At least two of the groups here regularly end up playing a game not because that's what any of them want to play, but because it is near the only title they are all actually willing to play. Their tastes and preferences tend to vary and disagree widely, leaving the LCD often the winner.

Quote:
Obviously, if I play it more than I anticipated, there's something there I like but didn't recognize.
Go figure. I played To Court The King this weekend. I'd played it at Kublacon and determined that that was a mistake I wouldn't repeat. I played it again this weekend, disliking it equally, simply because I knew my kids would accept it and it would occupy a potentially contentious time slot. TCtK has not improved in my opine even slightly due to increased play rate.

Quote:
Conversely, games I thought would be a big hit but didn't make it to the table much say they can't be that good...
No, you have the wrong audience for that game. Change the audience to suit and the game will be played more often. Context.
I also don't think 'number of plays' equates to a good or bad game. I love El Grande, one of my personal all time favorites. But, I don't play it much as others don't share my opinion of it. That doesn't mean I like it less and should rate it higher.

Conversly, I don't care for Nacht der Magier. It's ok, but not one of my favorites. It's, however, my most played game. Others REALLY like it, I don't mind it, and it's quick and painless (and admittedly fun). But that doesn't mean I'm going to rate it high. I don't choose to play it and don't suggest it and if they were to say yes to El Grande I wouldn't be playing it.

Puerto Rice, yes a lot of people LOVE this game, it's ok. I like to play it from time to time. But, just because it's well designed for other peoples taste. Doesn't mean I should rate it a 9 or 10.

If it was purely objective and not subjectinve. Then we would nominate a divers committee and they would tell us what we should rank it. But, it's subjective. It's about how a game makes US feel.

At least that's my take on it.

1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joe Grundy
Australia
Sydney
NSW
flag msg tools
badge
Avatar
Microbadge: Pregnant - baby on boardMicrobadge: It's a Boy!!Microbadge: Pregnant - baby on boardMicrobadge: Microbadge: Offline from The Geek for a while


Someone asked: "So with all these stats, can you finally prove that Puerto Rico doesn't deserve to be #1?"

I answered: "No, that's not stats it's psychology. It's called 'anchoring'".

2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Re: Rating with confidence - going TOTALLY NUTS wth statisti
I have been trying to absorb everything written here. I am not sure whether you answered this question, but what do you think of the institutional anti-shill ratings automatically thrown at every game by BGG admins?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Prev «  1 , 2 , 3 , 4 , 5  Next »   |