The Ratings Committee (RC) was involved in a number of tasks
throughout the year.  After nearly a year of the USCF testbed server
being inoperable, the RC was able to continue testing changes to the
system that had been charged by the Executive Board in 2010.
Additionally, the RC was involved in various minor activities, some of
which have led to issues that will need to be discussed in the coming
year.  We describe our accomplishments below.

The most important task of the committee was the testing of several
potential changes to the rating system that affected the K-factor in
the established rating formulas.  We tested two aspects of the
formulas simultaneously, and examined them through a simulation
analysis which we describe briefly below.  The first aspect we
examined was the formula for the .effective N. based on the player's
rating that ultimately determines the value of K in the established
rating formula.  We considered two alternatives to the current system
formula, both of which were designed to increase the value of K by as
much as 45% in the 2000-2200 rating range, but keep K relatively
unchanged at low and high rating levels.  The other aspect was to
revise the impact on K for games played under time controls of G/60 .
G/30 among strong players.  The two alternatives to the current system
we considered involved variations of lowering the value of K the
higher one's rating.  Along with these two types of changes, we also
investigated the simultaneous effect of varying the bonus factor.  In
total, we considered a total of 18 different rating formula changes to
explore, coming from 3 formulas for .effective N. (including the
current system), 3 formulas for the value of K for fast time controls
in established events (including the current system formula), and two
different bonus thresholds (3x3x2=18).  For each of the 18 systems, all
USCF-rated events were retroactively re-rated starting in 2004, and
two types of information were recorded.  The first was the final 2012
ratings of all players based on each of the 18 different systems.  To
summarize this information, these final ratings will be compared to
the ratings of the current system to check for increases or decreases
in particular rating ranges, and among scholastic and adult players
separately.  Also, Top 50 rating lists will be produced according to
these different systems.  The second was a measure of the
predictability of ratings on a game-by-game basis, and this measure
was summed for all games from 2010 onward.  The reason for only
including games starting in 2010 was that the changes to the system
were implemented starting in 2004, so that games played in 2010 and
after would likely have evidenced the effects of rating system changes
noticeably six years after implementation.  These predictability
measures will be summarized overall, and by rating ranges. Summaries
of these results are planned to be presented at the May 2012 EB
meeting.

Over the past year, the RC began testing and optimizing the Glicko-2
rating system.  Initially, a system was being tested that was an
approximation to the Glicko-2 formulas, but the properties of the
approximating system appeared to be worse than the actual Glicko-2
system.  The Glicko-2 system is being analyzed for possible adoption
by the USCF to address many of the ongoing problems with the current
system.  Currently, the system has been successfully implemented and
applied to a set of USCF game results, though certain parameters of
the system need to be optimized for predictability. While thorough
testing is necessary and essential, the advantages of such a system
include the following:  (1) Each player would have not only a rating,
but a measure of its reliability . players with reliable ratings will
typically not change by large amounts, and players with unreliable
ratings could undergo larger rating swings before stabilizing; (2)
results against opponents with unreliable ratings would have very
little impact on one's own rating; and (3) if a player had an
unusually strong performance relative to his/her rating, the player's
rating would be estimated to be less reliable, and the player's rating
would increase by larger-than-normal amounts . in effect, a bonus-type
mechanism would be an essential feature of the system.  We anticipate
presenting the results of the testing and optimizing to the EB in the
coming year. 

The RC was involved in a number of issues on a smaller scale.  One
issue involved whether extra games from a player re-entering a
tournament could be counted towards norms and titles.  The issue arose
based on tournaments that were submitted earlier this year in which
extra games were initially being counted towards norm/title
eligibility.  After discussions, it was felt that the intention of the
norm/title system was to count only games that were anticipated to be
played at the start of a tournament, and any additional games that
were .tacked on. would not count towards norm/title eligibility.
Enforcing such a rule would prevent situations where players may give
themselves extra opportunities to earn norms beyond the number of
games advertised for an event.

Another issue in which the RC participates from time to time is in the
evaluation of research proposals, and making recommendations to the
USCF office about whether to provide requesters with data in such
situations.  In one particular instance, a proposal was made to the
USCF to acquire ratings data, along with age and gender, from an
identified set of players, as part of a research proposal.  While the
RC is generally favorable about using ratings data for research, this
particular instance involved the researcher having access to
information that would compromise the privacy of certain USCF members,
a clear violation of the USCF privacy policy.  The discussion helped
the RC frame a basis by which to evaluate proposals to obtain USCF
ratings data by researchers.

The current rules for having a match rated require both players in the
match having established and published ratings when the match is held.
A situation arose this year in which players who were not both
established began a match, but by the time the match ended both
players had established ratings.  It was argued that the spirit of the
rules was meant to discourage starting a match unless the match
prerequisites were met.  To that end, the RC will work with the USCF
office to work on a clarification of the match rules so that players
would not begin matches unless they both had published ratings that
indicated they were established.

During the year, an incident came up in which a player gained 200
points over many 1-2 game events by defeating low-rated players.  It
was discovered that the player had taken advantage of a ratings rule
that says that a player must gain at least one rating point per event,
even if the player is almost certain to win every game according to
the winning expectancy formula.  The RC chair suggested that ratings
be computed and stored in decimal form, rather than as integers as is
currently carried out, and publish ratings rounded to the nearest
integer.  Such as approach would ensure that fractional gains in
rating points would be much slower to accumulate in instances where a
player constantly defeats only much weaker players.  The USCF office
will examine this possibility over the coming year.

In another unusual ratings-related occurrence, a player who recently
started tournament competition ended up with an established rating
that was clearly much higher than the player's ability.  It was noted
that the player had won all his games in rated events.  While
determining ratings based on all wins or all losses has the potential
to produce unreliable results, the RC chair suggested two approaches
going forward.  First, if there is some prior information that the
player is not at a very high or very low playing strength, then the
formulas could be revised to take that into account.  One way to do
this is to add a fictitious drawn game against a player with a rating
that might estimate the player's strength, and use this result in the
provisional rating formulas.  Second, when a player has all wins or
all losses, we should reconsider using the resulting rating from the
provisional rating formulas as the starting point for the established
rating (the source of the problem).  Instead, the formulas could be
revised by saving all the game outcomes and begin using the
established rating formula once a provisional rating was no longer
based on all wins or all losses.  This procedure would not use the
fictitious drawn game described above.  The RC will work with the USCF
office and the EB liaison over the coming year to address this issue
further. 

Each year the RC performs a set of diagnostic analyses to monitor
trends in the rating pool.  Overall rating levels have deflated from
the mid-1990s through 2000 when rating floors were decreased by 100
points without a counteracting inflationary mechanism.  With the new
rating system implemented in 2001, ratings started to re-inflate.  The
RC has the goal of restoring rating levels back roughly to where they
were at the end of 1997.  The focus of RC work has been on players
with established ratings who have been active over the current and
previous three years and who are aged 35-45 years old in the current
year.  Based on the continued decline in ratings for this group, the
RC recommended four years ago lowering the bonus point threshold from
B=10 to B=6 to accelerate the re-inflation of the rating pool.  This
change was applied retroactively to the beginning of 2008.  Over the
past three years, the average rating for this group has increased by
about 20 points in 2009, only about 3 points in 2010, and by 16 points
in 2011.  The average rating is currently about 23 points lower than
the average rating at the end of 1997.  Based on the continued
increase in ratings for this group, the RC would recommend increasing
the bonus threshold to B=8 to slow down the rating increases, and
continuing to monitor rating movement in this subset of players, but
our recommendation depends on the EB's reactions to the K-factor
analyses described at the start of this report.