statistics report · Uncategorized

Go Orlando’s strength and attendance data

I have been collecting data about club members (strengths, attendance, etc.) for a little over a month.  The process is just in its early stages, but I hope to keep it up.   I have this kind of information for about 14 people, though Josh has 26 people on his email list, so I clearly have a lot of gaps.  If you are interested and have not already given me your data, I would be curious to get your full name, any AGA rating data you have, as well as on-line info (handle, rank, server, etc.)

Still, I have enough to provide some preliminary summaries if people care.  I intentionally withhold all personal information in this summary.I’ve been collecting most of this information since just before the tournament (since Tuesday 4/29/08), five club meetings.  As I say, I have strength & attendance data for about 14 people, of which (since that date) we average an attendance of about 40% (5.6 people).  Our strengths range from 3d down to (perhaps) 22k-ish.  On any given Tuesday, you are almost certain to find at least one player 1d or above or 4k – 12k.  There’s also a small chance of finding a weaker player.  During the month I’ve collected data, no one between 3k and 1d has attended.

We’ve played over 26 matches since 4/29 at regular club meetings, and we average a bit over 2 games per person per meeting.

I downloaded AGA ratings for all rated players in the US over the last two years and divided those into “kyu” and “dan” (this is a bit questionable because the AGA rating system does not necessarily directly map into those rankings).  I created two histograms for each of these, including the mean and standard deviations for those data.  Again, the meaning of these summary data are at best complex (perhaps even dubious), as the distributions are decided non-Normal, but all’s fair in blogs and late-night number twiddling.

I then co-plotted some of the strengths of our members along the bottom of those plots.  You can see the dan plot here and the kyu plot here (these are small PDFs).  For our players, I used AGA information if I had it, otherwise I used the player’s self-professed rank, otherwise I used their ranking on an on-line service.  In one case (the 22k), I just speculated.  Of course we have many members whose strengths are not reflected, but I do not have their information (hint, hint).

Here are my conclusions:  From the kyu point of view, we have a strong club … most of our attending members are stronger than the average AGA player.   From a dan point of view, we are more or less on par.  Moreover, we have a good diversity of strength attending most club meetings; however, our general attendance is low (hint, hint).

It would be nice to get feedback from those that do not attend:  Is it the location?  Is it the availability of a plethora of on-line services and players?  Is it the day of the week, the time, etc.?  Is it the presence of annoying and obsessive number-crunchers?  That sort of thing.  No nagging, guilting, or arm twisting (okay, maybe some mild, good-spirited arm twisting) will be employed … just collecting useful information.



Leave a Reply

Your email address will not be published. Required fields are marked *