Monday, September 9, 2013

The USGA's War On Sandbagging - Part I, The War Begins


(The United States Golf Association (USGA) has made various attempts to control players who manipulate their handicaps in order to do well in tournaments.  The name for such manipulation is “sandbagging.” If the USGA’s effort can be characterized as a war, then it is not winning.  A series of six posts examines the history and effectiveness of the USGA’s war plan.  Part I details the flaws of the USGA’s earliest attempt at controlling sandbagging.  Part II examines a proposed policy that increased the penalties for alleged sandbagging. Part III argues the current handicap system may actually encourage sandbagging.  Part IV explains why the USGA could be losing the effort to win the hearts and minds of local golfers.  Part V examines the flaws in the USGA's current war strategy.  Part VI asks if the USGA's efforts are counterproductive and suggests it may be time for the USGA to withdraw from the battlefield.)
          

            The United States Golf Association (USGA) first introduced the "reduction of handicap index for exceptional tournament scores" into its handicap system as Sec. 10-3 in 1992.  This was an opening salvo in the war against sandbagging. Under the new handicapping procedure, a player could have his index reduced if his tournament performance is significantly better than his current index would indicate.  The reduction is based on the lowest two tournament scores over the past year.[1]  The index reduction is clearly intended to restrict the prevalence of sandbagging -- players having handicaps much higher than their ability would dictate.  It is possible, however, that this new system could also penalize the ethical player who plays in many tournaments. A system that does not clearly distinguish between the ethical player and the sandbagger is worse than no system at all.

            What is the probability the ethical player is penalized under Sec. 10-3?  We start with the assumption that a player's differentials (adjusted score minus the course rating times 113 over the slope rating) are distributed normally.[2]  The player’s mean differential is X with a standard deviation of s.  A player's expected current index is then:

                          Current Index   = .96(X - .8s)

(The .96 is the bonus for excellence adjustment in the USGA handicapping system.  In the normal distribution, the average of the better half of all scores is approximately the mean minus .8 of the player’s standard deviation.)

            For an index reduction to be imposed two conditions must be met.  First, the player's second lowest tournament differential must be at least 3.0 lower than his current index.  Second, for players with six or more tournament rounds, the difference between the average of the two lowest tournament differentials and the player’s current index must be at least 5.5 to trigger an index reduction.

            To estimate the probability of an index reduction, we simplify the problem by estimating the probability of two tournament differentials being at least 5.5 below a player's current index. 

In equation form:
                                    CI - TD   > 5.5
Where,
                                            CI   = Current Index
                                           TD   = Tournament Differential
Then,
                                           TD   < .96 (X - .8 s) - 5.5

In the standard normal distribution, the distance (measured in standard deviations) to the nearest TD meeting the condition of the equation above is:

                                               D   = {X - ( 96 (X - .8s) - 5.5}/s = .04X/s + .77 + 5.5/s
Where,
                                               D  = Number of standard deviations from the mean 
                                                       differential to the nearest tournament differential 
                                                       meeting the inequality condition.

The probability of receiving an index reduction is inversely proportional to the value of the D. That is, the larger the value of D, the lower the probability of receiving a handicap index reduction.[3]       

            The probability of a player having a differential at least 5.5 below his current index is a function of his average differential, X, and the standard deviation of the distribution of differentials, s.  In a small sample of players, standard deviations were found ranging from 3.5 to 5.5.  It was also observed that standard deviations typically increase with a player's index.  Scheid has reported standard deviations of approximately 4.0.[4]  The probability of having an exceptional round (defined here as 5.5 below your current index) is shown in Table 1 below for players of different average differentials and for a range of standard deviations.
Table 1
Probability of an Exceptional Round 



Standard Deviation

Ave. Differential = 0

Ave. Differential = 20

3.5

.0096

.0051

4.5

.0233

.0150

5.5

.0384

.0274
            The probability of having an exceptional round on any one day is quite low.  If you play in enough tournaments, however, the probability of having two such rounds and incurring an index reduction is not negligible.  Table 2 presents those probabilities under our assumptions of the mean and standard deviation of the distribution of scoring differentials.[5]
            Table 2 indicates that a player has a good chance to receive an index reduction if he plays in a sufficient number of tournaments.  A player with an average differential of 20 and a standard deviation of 4.5 (a typical bogey golfer), for example, is estimated to have a 23 percent chance of a reduction if he plays in 60 tournaments.  It is likely that the USGA did not design this system for such heavy tournament participation.  Some clubs, however, are designating weekly couples events, intra-club matches, and bi-monthly club tournaments as "tournaments." By including minor weekly tournaments, the number of tournaments entered in one year can easily rise to above 60 tournaments per year.  In this case, a player can be penalized not because he is a sandbagger, but because he participated.

Table 2
Probability of Two Exceptional Rounds in N Tournaments 




Number of Tournaments(N)

Average Differential = 0

Average Differential = 20

Standard Deviation

Standard Deviation

3.5

4.5

5.5

3.5

4.5

5.5

10

.00

.02

.05

.00

.01

.03

20

.02

.08

.18

.00

.04

.10

30

.03

.15

.32

.01

.07

.20

40

.06

.24

.46

.02

.12

.30

50

.08

.32

.58

.03

.17

.40

60

.11

.41

.68

.04

.23

.49

70

.15

.48

.76

.05

.28

.57

80

.18

.56

.82

.06

.34

.65

            Almost all of the index reductions due to the random nature of scoring will be small.  In most cases it will only be one stroke.  The inability to distinguish between the ethical player and the sandbagger, however, may weaken the credibility of the index reduction procedure.[6]  To correct this deficiency the USGA could either: 1) increase the exceptional tournament performance limit for increased levels of play, or 2) narrow the definition of "tournament" to include only major events as measured by prize money, entry fee, or some other indication of importance.  The latter procedure is recommended, since the former allows the sandbagger to minimize his index reduction by competing in a large number of minor events.

             The problem with USGA policy is that it does not differentiate between a player who enters 6 tournaments and one who enters 40 tournaments a year.  When informed of this analysis, the USGA took the position that no problem was likely to exist.  Dean Knuth, USGA Director of Handicapping wrote:[7]

“The handicap reduction procedure was developed with the use of data from about one million golfers that use the USGA’s GHIN handicap service…The most active group of tournament players on GHIN are two Northern California women’s golf associations.  These most active tournament players average five tournament scores per year, and the top one percent average sixteen….For the present, we believe that taking the table to six t-scores is sufficient.”        

The war is off to an inauspicious start when the Commanding General does not recognize and obvious flaw.[8]  There was never any reporting on the success of  Sec. 10-3 (e.g., body counts of sandbaggers snared).  Sec. 10-3 was adopted to show the USGA was doing something, Whether that “something” was effective did not appear to be a concern. The USGA did try to improve the “reduction in index formula” over the years.  Unfortunately, those changes were also plagued with flaws as will be seen in Part II of the USGA’s War on Sandbagging-Shock and Awe.







[1]USGA Handicap System, United States Golf Association, Far Hills, NJ, 1991, pp. 10-12.


[2]Scheid, F. J., "On the Normality and Independence of Golf Score, with Various Applications," in Science and Golf: The Proceedings of the First World Scientific Congress, Rutledge, Chapman, and Hall, London, 1990.


[3] Note that the higher handicap player has less chance to receive a reduction than the low handicap player if both have the same standard deviation in scoring.  This stems from the current index being based on differentials reduced by .96 while tournament differentials are not similarly affected.  It is not clear why the USGA included this bias against low handicap players in handicap system.



[4]Scheid, loc. cit.


[5]If we have N tournaments, the probability of having at least two exceptional tournaments is 1 minus the probability of having no exceptional tournaments and one exceptional tournament.  In equation form:

                        E = 1.0 - {(1 - p)N + N(1 - p)N-1(p)}

where,

                        E = Probability of at least 2 exceptional rounds in N tournaments

                        p = Probability of an exceptional round


[6]If clubs use different definitions of what is a "tournament," the equity of the handicap system among clubs would be lessened.  That is, some clubs may generate more index reductions simply because they designate more events as tournaments.


[7] Letter to author, February 19, 1992.


[8] It was only two years later, however, that the USGA had to correct the procedure by adding categories for up to 40 or more tournaments.  The change greatly reduced the probability of both the  ethical player and the sandbagger receiving an index reduction for exceptional tournament scores..

Monday, August 19, 2013

Chapman Handicaps and Sec. 3-5: Proposed Changes in Allowance


This post proposes two minor changes in the USGA Handicap System.  The first change is editorial and its purpose is to clarify the procedure for adjusting Chapman Handicaps in accordance with Sec. 3.5 (Players Competing from Different Tees or Men and Women from the Same Tees).  The correct adjustment procedure is now hidden within a labyrinth of subsections. To minimize any confusion, it is suggested that all Sec. 3-5 adjustment procedures be consolidated within Sec. 3-5. 
The second change proposes to make the rounding procedure for Chapman handicaps consistent with the rounding procedure for foursome handicaps.  This change will improve the equity of competition.  Each proposed change is discussed in turn.
Clarifying the Sec. 3-5 Adjustment - There is some confusion about how the team handicap in a Chapman competition should be adjusted in accordance with Sec. 3-5.   The confusion stems from the absence of explicit instruction in the USGA Handicap System on the correct adjustment procedure.  In fact, a strict reading of Sec. 9-4 (Handicap Allowances) could lead the Committee to assign incorrect handicaps.  Here is an example:  
Step 1: Players must first determine Course Handicap (from the tees played).
                Both players are 10 handicaps on the course with the higher Course Rating.
 Step 2: Players should then apply handicap allowances for the appropriate format.
The allowance is 60 percent for the player with the lower Course Handicap and 40 percent for the player with the higher Course Handicap.  The player with the lower Course Handicap would have a 6-handicap and the player with the higher Course Handicap would have a 4-handicap.
Step 3: Not applicable.
Step 4: If players are competing from different tees,…players must apply the adjustment for the difference in USGA Course Rating from the tees played.
Our players are competing from the tees with the longer course rating.  The difference in course rating is two strokes.  The handicap, adjusted by Sec. 3.5, for the player with the higher Course Handicap is 8 (6+2).  The player with the lower Course Handicap would have an adjusted handicap of 6 (4+2.)  The combined team handicap would be 14 (8+6).
Assigning a team handicap of 14, however, would be incorrect.  Sec. 9-3ciii states that in a foursome competition, the side playing from the tees with the higher Course Rating should receive additional strokes equal to the difference in Course Ratings.  The method used for adjusting handicaps in foursome competitions should also apply to Chapman competition.  Both competitions have a target score equal to a Course Rating + Team Handicap.[1]  To equalize the competition when players are competing from different tees, the difference in Course Rating should be added to the handicap of the team playing the course with the higher Course Rating.  In our example, the team’s handicap before the Sec. 3-5 adjustment is 10 (10 x .6 + 10 x .4).  Adding the adjustment for the difference in Course Ratings, the team’s handicap should be 12.
The confusion can be cured with two editorial changes:
1.       Change the heading of Sec. 9-3ciii to Foursome,  Chapman, and Pinehurst Competitions Using Different Tees or Men and Women Using the Same Tees. Currently, Sec. 9-3ciii is solely for foursome competitions even though it is referred to in Sec. 9-4aix and Sec. 9-4bvii which are only for Chapman and Pinehurst competitions.[2]  It would also be helpful if a Chapman example was presented.

2.       The Sec. 3.5 adjustment for foursome competition is now in Sec. 9.3 (Assignment of Strokes).  The Sec. 3.5 adjustment, however, does not seem to have anything to do with the assignment of strokes. It would seem reasonable to take Sec. 9.3c (Players Competing from Different Tees or Men and Women from the Same Tees) and integrate it into Sec. 3.5 (Players Competing from Different Tees or Men and Women from the Same Tees).  Sec. 9.3ci and Sec. 9.3ii can be eliminated since they are exact copies of Sec. 3.5a and Sec. 3.5b.  A proposed Sec. 3.5c would be the revised Sec. 9-3ciii (Foursome, Chapman, and Pinehurst Competitions Using Different Tees or Men and Women Using the Same Tees) discussed above.
Placing all of the Sec. 3-5 adjustment procedures in Sec. 3.5 should reduce confusion and minimize the errors in assigning team handicaps.
Inconsistent and Inequitable Rounding Procedure - In foursome stroke play, the handicap allowance is 50 percent of the partners’ combined Course Handicap (Sec. 9-4bvi).  In essence, the procedure is the same as taking 50 percent of each player’s handicap, adding the two together and then rounding to the nearest integer.  For example, assume the team consists of a 9- and an 11- handicap.  The team handicap would be:
                Team Handicap = .5·9 + .5·11 = 4.5 + 5.5 = 10
If the handicap of each player was rounded first and then added together, the team handicap would be 11 (5 + 6).  Under this procedure, a team with two odd-numbered handicaps would always get an undeserved stroke—e.g., two 11 handicaps would play as a 12, and two 12-handicaps would also play as a 12.  Wisely, the USGA does not recommend rounding individual handicaps for computing foursome handicaps.
In Chapman, however, the USGA does recommend this procedure.  The allowance is 60 percent of the lower Course Handicap and 40 percent of the higher Course Handicap.  Though not explicitly stated, the examples in Sec. 9-4aix and Sec. 9-4bvii shows the handicaps of each player being rounded before added together. Taking the same players as in the example above, the Chapman handicap would be:
                Player A = .6·9 = 5.4 rounded to 5
                Player B = .4·11 = 4.4 round to 4
                Team Handicap = 5 + 4 = 9

Had the same rounding procedure recommended for foursome competition been followed, the Chapman handicap would have been 9.8 rounded to 10.
The best measure of a team’s ability should be its unrounded handicap (i.e., a 9.8 handicap should be a 10-handicap and not a 9-handicap).  Therefore both for equity and consistency reasons, the team handicap in Chapman should be rounded and not the handicaps of individual players.
To make this change, only the examples in Sec. 9-4aix and Sec. 9-4bvii need be changed.  The example in Sec. 9-4bvii would be changed to:
Example: On side A-B, Player A has a Course Handicap of 10 and Player B a Course Handicap of 18.  60% of Player A’s Course Handicap is 6.0 (10 x 60%) and 40% of Player B’s Course Handicap is 7.2 (18 x 40%), so the total is 13.2 which is rounded to 13.  Side A-B will receive 13 strokes. 
Questions about Four-ball Stroke Play and Four-ball Match Play Allowances - In four-ball match play, players are assigned their full handicap (Sec 9-4aiii).  In four-ball stroke play, men are assigned 90 percent of their course handicap (Sec. 9-4bii). 
If high handicap teams have an edge in stroke play, why don’t they also have an edge in match play?  And why does the USGA recommend a maximum difference in handicaps for four-ball stroke play, but not four-ball match play?
Strangely, the USGA is of no help.  The USGA has not published any research on the equity of its handicap system when applied to singles matches.  The USGA has reported summary results on the equity of multi-ball competitions (e.g., four-ball match four-ball-stroke play).  This research, however, was completed by 1978.  The research involved taking s from multi-ball events at 31 clubs. Soley plotted team scores versus team handicaps to determine appropriate allowances. Since this research was conducted before the implementation of the Slope Handicap System,  Soley’s sample contained measurement errors. He considered all players of the same handicap to be equal, though Slope Theory maintains there would a difference in ability depending on the Slope Rating of the course.   Slope Theory also holds that it is easier to have a net score under the Course Rating the higher the Slope Rating.[1]  Soley clearly did not make any adjustments to his data since Slope Ratings had not yet been invented. In any event, the USGA has little to contribute to the debate over the equity of its handicap system.

The USGA Responds -Scott Hovde, Assistant Director of Handicap and Course Rating Administration wrote the following on August 29, 2013:

Larry,
I  have looked at your proposal and reviewed with Steve (Edmondson, Director of Handicap and Course Rating Administration) once he returned to the office. We both agree your suggested proposals have merit and will be reviewed at the next Handicap Procedure Committee meeting.  Thank you for taking the time to review and contact us with this issue.  If you have any additional concerns please feel free to reach out to us.
Regards,
Scott Hovde    
It seems likely the proposed changes will be made, but it will be January 1, 2016 before it is known for sure.



[1] This tenet of Slope Theory cannot be substantiated in theory or empirically.  See “Far Hills, We Have a Problem,” www. ongolfhandicaps.blogspot.com. 









[1] The Course Rating for a Chapman competition would not be the same as the Course Rating for an individual player.  The USGA assumes the difference in Chapman Course Ratings between tees is the same as the difference in the individual Course Rating between tees.
[2] There appears to be a typographical error in Example A of Sec. 9-3iii.  Currently it reads: “…The difference in Ratings is 2.4.  After the Course Handicap of each team is determined, one-half of the difference between Ratings (1 stroke) is added to the Course Handicap of each team with the resulting figure rounded off to the nearest whole number…”  I believe one-half the difference between Ratings is 1.2 strokes not 1 stroke.