Standard Race Times (Pars)
Race Pars
The Hong Kong Jockey Club (HKJC) provide standard times on their website (HKJC Standard Times) that reflect the times that could be expected by a winner (by class) on good ground. I have in the past calculated my own Pars based on the results held on my database, however I feel that it is better to use the HKJC pars, as they have access to more historical data, so their pars are probably more accurate than mine.
These Race Pars are primarily used during the Speed Rating calculation as a baseline time compared with the time achieved by each runner. During SPR calculation I prefer to compare race times with an average of these pars, as show in the All column below.
The following are the current Race Pars transferred from the HKJC website. Note that the individual Class Pars are used during Track Variant calculation. The C0 column represents average times for all Group races (G1, G2 and G3)
Sectional Pars
The average time for each section for each racecourse and race distance for the leader at the end of each section. Only races run on Good ground are included, also only races run at a Good, Good to Slow or Good To Fast paces are included. All weather races at Sha Tin are treated as being run on a different course. The Pace section (P) value shows the average total time for all except the final section of each race, these values are required to derive the pace that each race is run at.
These Sectional Pars are used to derive a Sectional Speed rating that is used to identify the performance of each runner in each section.
The following results are the current Sectional Pars.
Happy Valley
Sha Tin
All Weather
Track Variant
The purpose of Track Variant (TV) is to attempt to quantify whether or not the current average track times are riding faster or slower than the par times. Various factors affect the current times, including the following.
The current going, which is probably the greatest influence on average times.
The track, pars are not calculated by track, so we rely on TV to compensate for track differences.
Wind strength, which is not quantifiable, so we rely on TV to adjust pars where necessary.
In principal, if the race conditions produce race times that are on average faster than the par times for that race class, then the par times need to be reduced, and vice versa if slower than class par times. The amount of the par time adjustment is the Track Variant value which which can be positive or negative, this is then added to the par times during Speed Rating calculations.
The following steps are followed to calculate the Track Variant (TV) associated with a race meeting
For Each Race in the race meeting
Calculate the Average Time (AvgTime) of each runner that finished within 2 lengths of the winner. Then convert this figure to Average Speed (AvgMps) in meters per second by dividing the race distance by the AvgTime.
Get the class Par Time (ParTime) for the race course, race distance and race class. (use All class Par Time value if no race class Par Time available). Then convert this figure to Par Speed (ParMps) in meters per second, by dividing the race distance by the ParTime.
Now subtract the ParMps from the AvgMps to give the Speed Difference (MpsDiff) for that race
Now calculate the average Speed Difference (AvgMpsDiff) of each race in the meeting.
We now add the AvgMpsDiff to the ParMps calculated for each race, to create the adjusted par speed (ParMpsAdj). Then calculate the adjusted par time (ParTimeAdj) by dividing the race distance by ParMpsAdj.
The Track Variant (TV) for each race is the ParTime minus ParTimeAdj.
The following table shows the results of the above calculations for a single meeting at Sha Tin. Generally the races at this meeting were run in slow times in rain affected ground. As such the par times should be increased by the TV amount to give Speed Ratings that are not excessively low for the conditions that day.
Note that it is sometimes suggested that TV values may change during the day due to weather conditions. Whilst this is true, it is quite rare in Hong Kong and complicates the calculations, so I prefer not to consider such complications.
Past Race Speed Rating Calculation
Speed ratings are assigned to each horse in each race that finishes within 10 lengths of the winner. The SPR calculation is the difference between the overall Par Time for Course/Distance (adjusted by the current meeting Track Variant) and a horse’s race time. This difference is then multiplied by a factor that turns the time difference into a number of points or rating. I have experimented with a number of different factors and the following formula seems to give the best results for Hong Kong races.
SPR = 80 + ( (ParTime + TV - RaceTime) * 10)
The constant 80 is the baseline SPR, so if a horse runs to Par, then his Spr will be 80.
The constant 10 just causes one point to be generated for each 0.1 second difference.
Past Sectional Speed Rating Calculation
SPRs are calculated for each horse in each section is each race, using the same formula as described above, except that Track Variant is not applied to sectional Sprs. A Pace Spr is also calculated for each horse to reflect the Speed Rating for all sections excluding the final section.
These Sectional Sprs are really used as a guide to the speed of each horse in each section compared to the baseline of 80, so that one can judge if the section was run fast or slow.
Pace Points + Final Points (PpFp)
Pace Points are the Pace Spr minus 80 and Final Points are the Final Section Spr minus 80. So if Pace Points are negative then the race pace was slower than par, and if positive it was faster than par. Same with Final Points. So if Pace Points is positive and Final Points is positive we have a horse that can quicken off a fast pace which is unusual, normally they quicken off a slow pace, or slow down if the pace has been fast.
I tend to add these points figure together to give a different view of the performance of each horse.
Future Speed Rating Calculation
The speed ratings that I allocate each runner in a future race is derived from the Speed Ratings assigned to his last three runs where their Speed Rating is greater than zero. This is done by totaling 50% of the most recent Spr plus 30% of the Spr before that plus 20% of the Spr before that.