Know Yourself + Play Your Level = Increased Satisfaction

By Kevin Klipstein

January 2011

Skill Rating Range	Skill Level Division
5.51 and above	6.0	(A+)
5.01 to 5.50	5.5	(A)
4.51 to 5.00	5.0	(B+)
4.01 to 4.50	4.5	(B)
3.51 to 4.00	4.0	(C+)
3.01 to 3.50	3.5	(C)
2.51 to 3.00	3.0	(D+)
2.01 to 2.50	2.5	(D)
2.0 and under	2.0	(E)

Last month I wrote about my league experiences over the years and how I've enjoyed getting back to playing leagues this season. Getting into the details, we're actually in first place at the time of this writing. While I am definitely not "chalking up" a division title yet, what's interesting is how the ratings tell part of the story already.

Of the 13 teams in the men's 5.0 division, only one team has an average player rating that falls within the division's specifications of 4.51 to 5.0. The average player rating across the top 10 teams is 3.88, meaning many of our competitors should actually be competing in the 4.5 division, if not the 4.0 division (no disrespect intended). The men's 4.0 division shows the same thing, where the only team with an average player rating that falls in the specified range is the first place team, and all the other teams are below the range.

We have players telling us all the time that our ratings are a joke. In our member satisfaction surveys, it's the #1 complaint adults have when describing their dissatisfaction. Looking at the league stats above, it is easy to see why.

Recently, a concerned member pointed out a tournament to me where not one player in the 5.0 division had a rating that was in that division level. When looking at the data from tournaments played across the country, a similar pattern to that of the leagues appears.

In men's 3.0 for example, the ratio of participants playing above their rating level to participants playing within it is 20%. One in five are playing up which seems about right.

However, this ratio increases to 50% for men's 4.0 and 76% for men's 5.0. That means close to half of the 4.0 tournament players should probably not be playing that division.

The average amount the participants are playing above their rating is a full point.

With the majority of all tournament players being in the 3.5-4.5 range, this has a significant, and negative, impact on tournament draws since the whole point of the ratings system is for people to compete against players at a similar level.

Most players are probably aware that we have recommended ranges to play in by ability, but can't necessarily tell you what they are. These ranges are also the basis for eligibility into the Skill Level Championships and National League Finals.

It's a frustrating cycle where players sign up for a level that's above what they should play so that they can get "good matches" and, as a result, dilute the division they enter, creating the opposite effect for those who are actually playing their level. Of course, there are players who play up because "they have always been a 5.0," even though there is a good chance this currently is not the case, and perhaps never was.

How will this be solved?

The first step is making the players, pros and tournament directors aware of this issue. We need to work with Tournament Directors to encourage them to move players who are out of level to a division more suited for them. The same is true in working with the district league administrators as we approach the inaugural National League Finals, May 20-22, in Cambridge, MA. Districts will realize that winning teams may be named champions of the 4.5 division, but their average team rating may only be 3.7!

Reconciling which team plays in which division will be a challenge and require compromise this season, and hopefully less so next season. Improvements to the algorithm will also help, but as we look into the data more deeply, we recognize that better algorithms still won't work if there is a lack of understanding and compliance with the structure.