- The MSN Gaming ZoneGame Ratings
Since backgammon is a game of both luck and skill, one of the best ways to measure a player's skill is to observe their results over a very long period of time. Rating systems do that quite effectively.
The best kind of backgammon rating system is one similar to what is used in real-life rated games like chess and Scrabble. (Real-life backgammon does not have a rating system.) Ratings are based on tournament matches where a player cannot choose their opponent, where there is no way to gain points without risking losing points, and where the rating systems are programmed accurately.
GammonEmpire has a unique system where ratings are separated between money and non-money play, and more importantly, they are weighted by the stakes played for. So players can’t dump low-stakes matches and then play well for higher stakes. One other thing that GammonEmpire are doing is to track down users that sign in under multiple names from the same computer, to prevent cheating on the rating system.
|
I have developed rating systems like that for several online groups, and the best of those is warpgammon.com.
With the advent of money sites, a lot of people are looking to get low ratings, rather than high. So it’s important for a money site that has ratings to do what it can to prevent that.
As for sites where people try to play well and get honest ratings, I suggest GamesGrid and FIBS. Both sites have a lot of good players, and have sysops who will take action against people who cheat to increase their ratings.
I also think that the importance of ratings has decreased a lot since Snowie and GNUBG were created. Prior to those programs, the only way to assess your level of play was a rating developed over a period of time. But with those programs, you can see just how well you play. If your Snowie error rate is, say, 8.0, does it matter if your online rating is 1500 or 1700 or 1900? Maybe there’s a bit of ego involved in having a high rating, but players know their ability level, and that of their opponents, from the Backgammon software: Snowie and GNUBG analysis.
Followers of backgammon share an obsession common with those of other games: the need to know who the best in the game are, and how the best rank among themselves.
There is no "official" ranking that exists as for other games, such as the FIDE ELO point system in Chess.The following systems are/have been in use:
American Backgammon Tour
The American Backgammon Tour awards master points based on the results of certain select and authorized tournaments: only to prize money winners based on a formula that takes into account:
- the player grade,
- the number of players in the tournament, and
- the percentage of the total prize money paid to the particular individual.
One of the criticisms against this method is that regular participation gets an edge above high skill in this system. A more serious criticism is that the ABT is a "local" (though respected) organization, so its rankings are likely to be disputed, particularly outside USA.
Backgammon Rating on the Internet
With the coming of the Internet, and the increasing popularity of, and participation in backgammon on the web together with the availability of computers, efforts have been made to analyze the results of on-line games and rate them using statistical formulae with varying degrees of sophistication. The advantage was that in addition to rating top players, the common enthusiast could also rate herself. The earliest effort was made by Larry Kaufman and Kent Goulding to implement a variant of the Chess ELO system for backgammon (the KG rating list). Unfortunately, the people adjudged by the system to have the highest ratings tended not to be those rated highest by the experts! Understandably the KG rating system died a natural death.
Ratings and Backgammon Software
More elaborate versions of the same basic methodology are used by various sites on their backgammon game hosting servers. Sophisticated backgammon software is available today (Jellyfish,Snowie and Gnubg mentioned above), some of them free. Most of them can analyze a player's moves throughout a game and assign "error levels" that discount the element of luck from her performance and rate her level of skill. Some sites suspect some players of deliberately throwing certain games at certain times to LOWER their ratings to gain unfair advantage under certain betting situations, and have set up sophisticated auditing systems to catch the cheats. Clearly, the battle of wits between the site administrators and players will continue.
On the other hand, if a player wants to honestly rate herself she can compete against the software and read its rating of her play. She will know what to realistically expect when she plays competing human beings, perhaps for money. To that extent, there is now less of a need to get a rating from a site .Sites will however continue to sharpen their software and publish their own lists of comparative ratings among players whose games they can access.As they weed out the distortions, bias and errors in their systems, these lists will improve in quality.
Who is the best Backgammon Player
Which now leaves us with the original problem: how do we find out who the best in the world really, really are? One possible (and sensible) option is to get yourself a proper electoral college and put the matter to a vote! The idea is good, but like all ideas involving multitudes, distance and time, the devil is in the detail.To find knowledgeable voters whose opinions count, to find them in sufficient numbers distributed across the globe so that bias is eliminated, to motivate them to fill up the forms on time, and to ultimately collate and publish the results is not a task for the faint-hearted.Some people have started on this idea: more power to their elbow(s)! The lists they come up with will be neither unbiased nor error-free, but will, I suspect contain less obvious mistakes than those generated by computer software!
Meanwhile, are there any new ideas for improving the quality of player rating systems?