Detailed Report and Rationale on Proposed New Standards for Awarding JD awards

Tom Dingle -- November 2006

 

At the JD committee meeting in September, 2006, I was asked to look into revising the standards used for choosing athletes to be considered for JD awards.

 For your convenience after having read this report here are the links to the various files referred to in the report:

Timeline, TopTen, Synthetic Top Ten, 2006 Standards, Hypothetic Awards List for 2006, Comparison of Old and New Standards

Note added in April, 2007. I have recently taken the 2006 All Time Top Ten List and recalculated the Standards table (2007 Standards) to bring it into line with the above analysis and also making note of the changes in JD events that have been dropped (200m, 400m for 9 and 10 year olds and 200m Hurdles for 11 year olds) and those that have been added (60m for 9, 10 and 11 year olds and pole vault for 13 year olds). I also recalculated the percentages that would lead to an equitable distribution with the changes taken into account and found that they did not change significantly.

.

The current criteria (2006) are:

 

1)      Be a full competitive member of BC Athletics

2)      Compete in at least one of the three JD Championship meets in the last twelve months (Track and Field, Pentathlon or Cross Country from the previous year) AND at least two other meets from the approved list.

3)      At meets approved by the JD committee the athlete should:

a.       For 9 to 12 year olds  --   have at least one performance which is as good as or better than the #10 on the All Time Top Ten list AND during the current year have competed in a running event, a jumping event and a throwing event (can be at the same or different meets)

b.      For 13 to 15 year olds  -- have at least one performance which is as good as or better than the #10 on the All Time Top Ten list  OR have a performance within 3% of #10 and be dominant in that event (i.e. have  several performances which are at or near the top for that year and better than most others in the group).

 

 

The reason I was asked to review the standards is the fact, that in the 2005 awards selection process it was felt that there were an inordinate number of awards given in the throwing areas and to the pentathlon, events which were relatively new or where the event had been changed recently (e.g. hurdle heights or weights of throwing implements).  There was general feeling that some of the #10 performances were a bit weak. Athletes competing in long established events like the 100m or the high jump were faced with #10 performances which were almost the same as the #1 performances for that event since competition in the event had been going on since the start of the All Time Top Ten program.

 

When I was checking the 2006 nominations over the past month I had the same feeling as was expressed in 2005.

 

To quantify these feelings I did the following:

 

1)      Went back in the Top Ten booklets for the past twenty years and looked at how the #1 and #10 performances changed over the years.  In addition, I looked at the percent difference between the two to see how it had changed.

 

Details of this analysis are given in the worksheets Timeline and TopTen (click to access). 

 

·        In the TimeLine worksheet I have given the #1 and #10 values for selected events for 15 year olds and 11 year olds taken at four year intervals starting in 1989. 

 

·        For the 100m and 200m (sprints), the 1500m (distance), the High Jump and Long Jump (well established events), it is clear that both the  #1 and #10 values have changed very little over this period and the percent differences jump around a little bit depending on which value changed slightly.  The high jump differences are a bit larger on the whole than for the runs. 

 

·        The same is true for the sprint hurdles, except for the 11 year olds where the last change was made in 2000.  Here it is noticeable that there was a significant relative change in the # 10 values and hence an increase in the differences.

 

·        For the race walk, which started only being listed in 1996, the differences are quite large and it is obvious that things have not settled down yet.  In addition, the event is not very popular at the older age group (in fact there were not enough performances to have a #10 performance for several years) so it will take some considerable time for decent standards to be established.

 

·        For the shot put, where  changes in weights were made in 1998 and 2002, the effect of the change is noticeable in most cases.  Similarly for the pentathlon the effect of a change in one of the events is obvious.

 

·        In the TopTen worksheet, examination of the difference between the #1 and #10, in the 2005 tables for all events, expressed as a percentage of the #1 value, is fairly constant for the established events but is considerably larger and varies considerably more from age to age for the new and recently changed events.  This seems to indicate that the differences settle down to a small value with only slight variations only after the event has been running unchanged for many years.  On the other hand, the larger values and larger variation for the newer events indicates that the things are still in a state of flux and the # 10 value may not be a reliable standard from year to year or event to event. 

 

·        The percentage differences are around 4 to 5% for the established running events, around 6 to 8% for the established jumping events and over 12% for the throwing events, the pentathlon and those running events that are not well established.

 

·        If one takes into account the differences in the various categories using the IAAF points formulas for comparing events, then 5% difference in the runs is about the same as 6.6% in the jumps and 9% in the throws. The trends are in the right direction but the low value of the throws indicates to me that the #10 in most of the throws has not settled down the way that the runs and jumps have. 

·        These points are further confirmed if one takes as a starting point the Top Ten (not All Time Top Ten) performances in selected events from 2001 and follows them through to 2006 using the Top Ten values for each year to create a Synthetic All Time Top Ten Table with the events all starting in 2001. In none of the events with histories longer than six years have the Synthetic #1 to #10 spreads settled down to the actual values taken from the All Time Top Ten list for 2005.

 

2)      I then looked at the relative percentages of run, jump, throw and pentathlon awards given out in 2005 and 2006 and compared them to an equitable spread depending on number of events.  If we look at the total number of events that athletes can compete in, we get the following results:

 

 

Runs

Jumps

Throws

Pentathlon

 

(100m to 1500m, RW

Sprint H) x 7 years

D Hurdles x 5 years

3000 m x 3, SC x 2

LJ, HJ x 7 years

TJ x 5 years

PV x 2 years

SP x 7 years

DT, JT x 6 years

HT x 4 years

Pent. x 7 years

No. of age groups and events

 7x7 + 5 + 3 + 2

2x7 + 5 + 2

7 + 2x6 + 4

7

Number genders

2

2

2

2

Total events

118

42

46

14

Equitable Percentage

53.6%

19.1%

20.9%

6.4%

% awards for 2005

41.0%

6.6%

28.7%

23.8%

% awards for 2006

39.1%

10.9%

33.6%

16.4%

% using new standards

52.2%

19.0%

21.3%

7.5%

 

·        As you can see, when awards are compared to the number of events, the jumps are very low, and the runs somewhat less so, whereas the throws and pentathlon are well over the equitable level.  Again I think this reflects the fact that all of the throwing events are new or have been changed in the passed few years so the difference between the #10 and #1 level is much larger and it is much easier to get a #10.

 

3)      I then played around with a whole series of scenarios, using the data I had collected while checking the 2006 nominations, to see whether I could derive what seemed to be the most reasonable way of addressing the above concerns. I took as my working guidelines the following criteria to set the standards:

 

a.       A more equitable distribution of  awards to the four areas

b.      A total of about 100 athletes being given awards

 

These various scenarios included:

 

i)                    Using some percentage of the #1 value for that event

ii)                   Using some percentage of the average of the #1, #2 and #3 values.  I tried this because I thought that in some cases the #1 value was anomalous whereas the average evened out most anomalies.

iii)                 I repeated i) and ii) using the points system developed by Brian Foan. This system is based on the IAAF scoring system and the Island clubs use it to make awards to athletes in the Vancouver Island Track Series across all events.

iv)                 At first I tried use a single percentage for all events but that didn’t work out too well.  I could never get an equitable distribution.

v)                  So I tried using separate percentages for each main category.  This seemed to be much more consistent with the results I found in part 1) by examining the Top Ten Values.

 

My final choice was to use a different percentage of the average of the #1, #2 and #3 values for each of the main categories (95 % for Runs, 90.2% for Jumps, 87% for Throws and Pentathlon). See 2006 Standards Table.

·        One rationale for doing this is that it is similar to how the standards are set for the provincial teams at higher levels.  Averages of the third place finishes in the event for five years are used as the basis for the standards.  However, the cut has to be lower than the third place value here or we would end up giving out only a few awards each year.  I decided that the average of #1, #2 and #3 was a better reference point as it tended to smooth out any anomalies at the top end. 

 

·        The resulting percentage values for the four areas using the new standards, is given in the last line of the table above.  These are the values I got using the values given above.  The percentage for pentathlon is a shade higher than the ideal but I think this is reasonable given the fact that the athletes do 5 events. If desired, these values can be adjusted to give a better distribution by fine tuning the cuts used for the standards.  As an aside,  using Brian Foan's points system gave essentially the same results and I didn't want to increase the apparent complexity of any system we might put into place.

 

4)  These new standards, based on the 2005 Top Ten Tables, are given in the Standards worksheets

 

 


  • This gave me 98 award winners, close to the 93 we gave this year. There were.13 new runners added, 8 new jumpers and three new throwers.  There were 11 throws athletes dropped, 4 pentathletes, one in distance hurdles one in the racewalk, one in the pole vault and one in the long jump who just made it this year on the 3% dominance criterion. This seems to emphasize  the fact that the new standard is more rigorous in the new events than the old #10.  The addition of 3 throwers and the dropping of 2 jumpers indicates that there are some anomalies between the age and gender groups even within a given event.  Hopefully, the use of a common average for each category will compensate for this to some extent.

 

Details of the old and new award lists are given in the worksheet in the attached file along with comments.

 

For your convenience after having read this report here are the links to the various files referred to in the report:

Timeline, TopTen, Synthetic Top Ten, 2006 Standards, Hypothetic Awards List for 2006, Comparison of Old and New Standards

Note added in April, 2007. I have recently taken the 2006 All Time Top Ten List and recalculated the Standards table (2007 Standards) to bring it into line with the above analysis and also making note of the changes in JD events that have been dropped (200m, 400m for 9 and 10 year olds and 200m Hurdles for 11 year olds) and those that have been added (60m for 9, 10 and 11 year olds and pole vault for 13 year olds). I also recalculated the percentages that would lead to an equitable distribution with the changes taken into account and found that they did not change significantly.