Recommend
3 
 Thumb up
 Hide
24 Posts

BoardGameGeek» Forums » BoardGameGeek Related » BGG General

Subject: The quest for the best game ever (or not) rss

Your Tags: Add tags
Popular Tags: [View All]
Matthieu Weber
Finland
Jyväskylä
flag msg tools
badge
Avatar
mbmbmbmbmb
Game ratings are not meaningful (see http://www.boardgamegeek.com/thread/292656), we all know that (sorry for those who just discovered that Santa doesn't really exist). Therefore game ranks are equally worthless. How can we objectively evalute which games are the best ? Well, objectively we can't, but following the saying ``if it's so popular, it has to be good'' (also know as ``eat sh*t, one billion flies can't be wrong'') , we can rank the games based on how many BGG users have been playing them. And with no further delay, the real top 50:

1 Race for the Galaxy
2 Carcassonne
3 Puerto Rico
4 The Settlers of Catan
5 Dominion
6 Pandemic
7 Ticket to Ride
8 Power Grid
9 Lost Cities
10 Agricola
11 Stone Age
12 Battlestar Galactica
13 Small World
14 San Juan
15 Bohnanza
16 Ra
17 Le Havre
18 Citadels
19 Alhambra
20 Galaxy Trucker
21 Thurn and Taxis
22 Caylus
23 Magic: The Gathering CCG
24 Bang!
25 Ticket to Ride: Europe
26 For Sale
27 Dominion: Intrigue
28 Tigris & Euphrates
29 Race for the Galaxy: The Gathering Storm
30 Ingenious
31 Arkham Horror
32 Coloretto
33 Saint Petersburg
34 Kingsburg
35 Category 5
36 Memoir '44
37 Through the Ages: A Story of Civilization
38 Steam
39 Roll Through the Ages: The Bronze Age
40 Zooloretto
41 TransAmerica
42 Tales of the Arabian Nights
43 El Grande
44 Hey! That's My Fish!
45 Acquire
46 The Princes of Florence
47 No Thanks!
48 Shadows over Camelot
49 Endeavor
50 Cosmic Encounter

For those who want to know the math behind the numbers, I've taken the top 50 of the most played games, sorted by unique users, of every month since Jan 1st, 2000. Games that were in the top 50 earlier but are not anymore got fake rankings of value 51 for every month between now and the last time they appeared in the top 50. Then I averaged those rankings and re-ranked the games based on that nuumber. This trick prevents games that were at the top some time ago but did not manage to stay there to get too high a rank.

Yes I know, there are shortcomings to this approach, such as ``Race for the Galaxy is a pretty quick game, so it's to be expected that it will be more played than others'', but I warned you about the philosophical underlinings of this study.

EDIT:
Since many readers complained that Dominion should be number 1, I experimented with other methods. For example, one can consider that a game that stays for a long time in the top 50 is a better game than one that appears for a couple of months and disappears again (e.g., Hacienda, Amun-Re, Shadows over Camelot), or one which has been in the top 50 for only a short time, so we can define a minimum period of time during which the game should appear in the top 50 (e.g., 12, 24 or 36 months), and if it has not appeared in the beginning of that period yet, we assign it an average rank (i.e., 25.5) for that period. If a game disappears from the top 50 before the current time, we assign it a rank of 51.

So for a period of 12 months, we have the following top 20:

1 Dominion
2 Race for the Galaxy
3 Pandemic
4 Carcassonne
5 Puerto Rico
6 The Settlers of Catan
7 Ticket to Ride
8 Power Grid
9 Agricola
10 Lost Cities
11 Stone Age
12 Battlestar Galactica
13 San Juan
14 Bohnanza
15 Small World
16 Le Havre
17 Ra
18 Citadels
19 Galaxy Trucker
20 Alhambra

For 24 months, we have:

1 Race for the Galaxy
2 Carcassonne
3 Puerto Rico
4 The Settlers of Catan
5 Ticket to Ride
6 Power Grid
7 Pandemic
8 Lost Cities
9 Agricola
10 San Juan
11 Bohnanza
12 Stone Age
13 Dominion
14 Ra
15 Citadels
16 Battlestar Galactica
17 Alhambra
18 Galaxy Trucker
19 Le Havre
20 Thurn and Taxis

And for 36 months, we have:

1 Carcassonne
2 Puerto Rico
3 The Settlers of Catan
4 Ticket to Ride
5 Power Grid
6 Lost Cities
7 Race for the Galaxy
8 San Juan
9 Bohnanza
10 Pandemic
11 Agricola
12 Ra
13 Citadels
14 Stone Age
15 Dominion
16 Alhambra
17 Thurn and Taxis
18 Battlestar Galactica
19 Magic: The Gathering CCG
20 Caylus

This last list clearly favorises games that are not rated number 1, but have stayed in the top 50 for a long time, while a shorter period (12 months) favorises games that have had a very high rank in the recent months.

Whether those games are ``better'' or not can be indefinitely discussed in absence of an objective way of rating games, but if one considers that being able to stay in the top 50 for the past 9 years is the mark of a very good game, then the last list might be of interest.
4 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Virre Linwendil Annergård
Sweden
Stockholm
flag msg tools
Forza Bajen!
badge
... and still sober
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
noursy wrote:

Yest I know, there are shortcomings to this approach, such as ``Race for the Galaxy is a pretty quick game, so it's to be expected that it will be more played than others'', but I warned you about the philosophical underlinings of this study.


It also have the problem with this: People play games that are not the best games they know because 1) They want to get people into gameing 2) They want to play games and some times that means playing not so good games etc etc
3 
 Thumb up
0.01
 tip
 Hide
  • [+] Dice rolls
Jim Cote
United States
Maine
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
As valid as any other system, I suppose. But for me, I don't play my top-rated games as much as others, because like most, I play games with other people who each have their own tastes. I often play B (7.5-ish) and sometimes C (6-ish) games because that's just what's on the menu.
6 
 Thumb up
0.01
 tip
 Hide
  • [+] Dice rolls
Maarten D. de Jong
Netherlands
Zaandam
flag msg tools
badge
Avatar
Re: The quest for the best game ever
noursy wrote:
Game ratings are not meaningful (see http://www.boardgamegeek.com/thread/292656),...

Unfortunately, game ratings are very meaningful to individuals. And with a better analysis which takes into account what a rating actually means, the rankings will carry more meaning too. See here.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Kyoko Steeple
United States
Columbia
HI
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
Your sample is also limited to BGG'ers that log their plays (and who, I suspect, are in the vast minority here on BGG.) This introduces a number of possible confounding variables that are unique to play-loggers; at best, these results extend only to users who log their plays and tell us nothing about BGG opinions at large.
5 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthieu Weber
Finland
Jyväskylä
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
virre wrote:
noursy wrote:

Yest I know, there are shortcomings to this approach, such as ``Race for the Galaxy is a pretty quick game, so it's to be expected that it will be more played than others'', but I warned you about the philosophical underlinings of this study.


It also have the problem with this: People play games that are not the best games they know because 1) They want to get people into gameing 2) They want to play games and some times that means playing not so good games etc etc


This is not incompatible with the fact that the game is at least somehow ``good'', imho.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthieu Weber
Finland
Jyväskylä
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
cymric wrote:
noursy wrote:
Game ratings are not meaningful (see http://www.boardgamegeek.com/thread/292656),...

Unfortunately, game ratings are very meaningful to individuals. And with a better analysis which takes into account what a rating actually means, the rankings will carry more meaning too. See here.


But the rating of the games is something that is made once, uner the influence of that day's mood, and often not updated to reflect changes in personal tastes, while the choice of a game for a session is something which occurs much more often and in itself contains the enthousiasm of the players for that game.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthieu Weber
Finland
Jyväskylä
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
ekted wrote:
As valid as any other system, I suppose. But for me, I don't play my top-rated games as much as others, because like most, I play games with other people who each have their own tastes. I often play B (7.5-ish) and sometimes C (6-ish) games because that's just what's on the menu.


Which makes B and C games good too, although not in the same sense as the ones in you A list. If B and C games were so terrible, you would fight against playing them, I guess.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthieu Weber
Finland
Jyväskylä
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
SleightOfHand wrote:
Your sample is also limited to BGG'ers that log their plays (and who, I suspect, are in the vast minority here on BGG.) This introduces a number of possible confounding variables that are unique to play-loggers; at best, these results extend only to users who log their plays and tell us nothing about BGG opinions at large.


That's absolutely true. On the other hand, game ratings only reflect people's preferences and games' popularity, which is influenced by so many factors (author, brand, marketing...).
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Kyoko Steeple
United States
Columbia
HI
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
No doubt. I'm just pointing out that your sample isn't representative of BGG as a whole.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Maarten D. de Jong
Netherlands
Zaandam
flag msg tools
badge
Avatar
Re: The quest for the best game ever
noursy wrote:
But the rating of the games is something that is made once,...

No.

Quote:
... under the influence of that day's mood...

How do you know? Did you do (semi-)rigorous tests to substantiate this claim?

Quote:
and often not updated to reflect changes in personal tastes...

How do you know? Did you do (semi-)rigorous tests to substantiate this claim?

Quote:
...while the choice of a game for a session is something which occurs much more often and in itself contains the enthousiasm of the players for that game.

How do you know it contains the enthousiasm of the players for that game? You only measure what gets played and recorded, any further interpretation of that number is subjective. In any case, my inmpression of this ranking method is that because you want to use play data you find reasons to discredit the ratings. (By the way, determining what gets played during a session is also very dependent on the day's mood.) I can't say I'm impressed, really.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Sven
Germany
Frankfurt am Main
Hessen
flag msg tools
badge
We will move from victims to survivors and conquer as we go.
Avatar
mbmb
Re: The quest for the best game ever
If rankings don't mean anything, why do you spend so much time on this?

Edit: Sorry, misunderstood your method. I thought it was based on the number of plays, not on the number of users logging plays...which of course introduces a whole different set of faults.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Virre Linwendil Annergård
Sweden
Stockholm
flag msg tools
Forza Bajen!
badge
... and still sober
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
noursy wrote:
cymric wrote:
noursy wrote:
Game ratings are not meaningful (see http://www.boardgamegeek.com/thread/292656),...

Unfortunately, game ratings are very meaningful to individuals. And with a better analysis which takes into account what a rating actually means, the rankings will carry more meaning too. See here.


But the rating of the games is something that is made once, uner the influence of that day's mood, and often not updated to reflect changes in personal tastes, while the choice of a game for a session is something which occurs much more often and in itself contains the enthousiasm of the players for that game.


Did I say that that raitings was good? Only way to find a good game is to play untill you find your personal favourite. Will never be a good conclusive way to find the best game.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Werner Bär
Germany
Karlsruhe
Baden
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
noursy wrote:
I've taken the top 50 of the most played games, sorted by unique users, of every month since Jan 1st, 2000. Games that were in the top 50 earlier but are not anymore got fake rankings of value 51 for every month between now and the last time they appeared in the top 50. Then I averaged those rankings and re-ranked the games based on that nuumber.

I don't understand your math. In the 'plays by unique users' list, Dominion was #9, #1, #1, #1, #1, #1, #1, #1, #1, #1, #1, #1. How is it possible that 4 games have a better average than that?
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthieu Weber
Finland
Jyväskylä
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
cymric wrote:
noursy wrote:
But the rating of the games is something that is made once,...

No.


Maybe.

cymric wrote:
Quote:
... under the influence of that day's mood...

How do you know? Did you do (semi-)rigorous tests to substantiate this claim?


Of course not. But I have yet to be presented a person who is able to distanciate herself from her emotions (Vulcains don't count, they don't exist).

cymric wrote:
Quote:
and often not updated to reflect changes in personal tastes...

How do you know? Did you do (semi-)rigorous tests to substantiate this claim?


No, But I know for a fact that people are lazy.

cymric wrote:
Quote:
...while the choice of a game for a session is something which occurs much more often and in itself contains the enthousiasm of the players for that game.

How do you know it contains the enthousiasm of the players for that game? You only measure what gets played and recorded, any further interpretation of that number is subjective.


Did you actually read the introduction to this post?

cymric wrote:
In any case, my inmpression of this ranking method is that because you want to use play data you find reasons to discredit the ratings. (By the way, determining what gets played during a session is also very dependent on the day's mood.) I can't say I'm impressed, really.


No worries, I'm not trying to impress anyone. But I did decide beforehand that the current ranking system is worthless, mainly because almost half of the games in the databse are rated 7. But yes, I wanted to play with the data, I can't deny it.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Eric Phillips
United States
Greensboro
North Carolina
flag msg tools
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
I think the BGG ratings system must be more accurate than this method.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthieu Weber
Finland
Jyväskylä
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
Werbaer wrote:
noursy wrote:
I've taken the top 50 of the most played games, sorted by unique users, of every month since Jan 1st, 2000. Games that were in the top 50 earlier but are not anymore got fake rankings of value 51 for every month between now and the last time they appeared in the top 50. Then I averaged those rankings and re-ranked the games based on that nuumber.

I don't understand your math. In the 'plays by unique users' list, Dominion was #9, #1, #1, #1, #1, #1, #1, #1, #1, #1, #1, #1. How is it possible that 4 games have a better average than that?


It's because with the current method, both Dominion and Race for the Galaxy get an extra fake ranking of 51. Since Dominion has been present for much less time than Race for the Galaxy, Carcassonne, ..., this ``bad point'' weighs in the averaging more than it does for older games. If Dominion is so great that it will stay in the top 50 for many years, as did Carcassonne and Race fr the Galaxy, it will go up.

Since i'm playing with the numbers to make them say what they are not saying, I can get very different results with relatively small changes in the parameters. E.g., if instead of giving a rank of 51 I'd give a rank of 25.5, Dominion becomes number 2.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthieu Weber
Finland
Jyväskylä
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
Fortuna wrote:
I think the BGG ratings system must be more accurate than this method.


BGG ratings are based on a Baysian average, which consists in adding 100 fake ratings of value 5.5 to the actual ratings by the users. The choice of 100 is arbitrary, as are my choices for my method. The alternative method described in http://www.boardgamegeek.com/geeklist/27254/page/2 is probably more sound than mine or the current ranking method (although it also has arbitrary values as parameter), on the condition that ratings made by the people are consistant (which I believe to be false, in average (don't bug me for proof, I don't have any, hence the ``I believe'' in my sentence)).
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dan Cain
United States
Tacoma
Washington
flag msg tools
badge
The Evil-Do That I Do-Do
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
From a publisher side, this list also just represents who has the best distribution model for their games. I believe availability has something to do with the games that are on this list.

LA
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Wendell
United States
Yellow Springs
Ohio
flag msg tools
Si non potes reperire Berolini in tabula, ludens essetis non WIF.
badge
Hey, get your stinking cursor off my face! I got nukes, you know.
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
One game of Agricola = six games of Modern Art

One game of World in Flames = 500 games of Agricola (minimum)
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Werner Bär
Germany
Karlsruhe
Baden
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
noursy wrote:
Quote:
I don't understand your math. In the 'plays by unique users' list, Dominion was #9, #1, #1, #1, #1, #1, #1, #1, #1, #1, #1, #1. How is it possible that 4 games have a better average than that?
It's because with the current method, both Dominion and Race for the Galaxy get an extra fake ranking of 51.

Ah, some playtesters logged their plays? And now it gets penalty scores for not being in the top 50 every month before it's release? Funny.
(it was released in germany on oct 22, and in USA on oct 30, and reached #9 that month)
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Joe Huber

Westborough
Massachusetts
msg tools
Re: The quest for the best game ever
virre wrote:
It also have the problem with this: People play games that are not the best games they know because 1) They want to get people into gameing 2) They want to play games and some times that means playing not so good games etc etc


I'm not disputing that for some, this is true, but I am curious to what extent it is the case.

My average plays, by rating, reflects a very strong correlation:

10: 217
9: 39
8: 14
7: 6
6: 3
5: 2
Under 5: 1

My suspicion is that a similar curve (if less extreme) is true for most gamers; even those who are less vocal about getting their favorites to the table tend to seek out opportunities to play their favorite games, and thus play them more, over time, than other options.

Interesting method for deriving the list...
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matthieu Weber
Finland
Jyväskylä
flag msg tools
badge
Avatar
mbmbmbmbmb
Re: The quest for the best game ever
Werbaer wrote:
noursy wrote:
Quote:
I don't understand your math. In the 'plays by unique users' list, Dominion was #9, #1, #1, #1, #1, #1, #1, #1, #1, #1, #1, #1. How is it possible that 4 games have a better average than that?
It's because with the current method, both Dominion and Race for the Galaxy get an extra fake ranking of 51.

Ah, some playtesters logged their plays?


No, more like a bug in the part which tries to penalize games that used to appear in the top 50 in the past, but do not anymore (such as Amun-Re, Hacienda or Shadows over Camelot).

Werbaer wrote:
And now it gets penalty scores for not being in the top 50 every month before it's release? Funny.


Not before its release, but after it last appears in the top 50 (read ``bug'', in the case of e.g., Dominion).
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
John Earles
Canada
Toronto
Ontario
flag msg tools
badge
Avatar
mbmbmbmbmb
"There are three kinds of lies: lies, damned lies, and statistics."

- British Prime Minister Benjamin Disraeli
3 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.