Track bias is a subject frequently discussed in horse racing circles. There is leader /on pace bias, swooper bias, inside bias, outside bias and so on. Bias is often related to prevailing weather conditions where rain or lack thereof can influence the result of a race.

Rail movement where some racecourse managers wish to protect a virgin strip of ground can often cause a “fast lane” at carnival time. Fortunately, in Hong Kong the rail movement is logical, A to B to C to C+3 thus providing an unused racing surface where a strip of ground can be replenished and be in good order every month or so.

We define track bias as “The performance of a racetrack producing results that differ from the historical performance of a particular track and surface.”

To illustrate, it is an historical fact that of the roughly 900 races run at Happy Valley on the C+3 course over the last twenty years that barrier one has won 18.6% of these races and barrier twelve 3.16%.

To put this another way if your horse draws barrier one, it is, on average, about six times more likely to win than if he draws barrier twelve (18.6 / 3.16).

On the Happy Valley C Course this “rail effect” is not as strong with barrier one hitting at 12.7% and barrier twelve at 4.3%, the effect being threefold rather than sixfold. These effects are not track biases per se according to our definition of bias stated above they are just historical facts.

Hong Kong has not historically been prone to track biases (as defined above) which is a real advantage for the Jockey Club as not all jurisdictions are able to consistently achieve this.

Let us consider the meeting at Happy Valley run on the 15th February 2024 on the C Course as a strong example of track bias.

You will note from Table 1 below that on this day barriers 1,2 and 3 did not win a race and filled the quinella spot only once. Historically these three inside barriers fill a placing 32.8% of the time as against the14.8% on the 15th. If a horse benefited from being in any of the other three groups (barrier four through twelve) it’s chances of placing were enhanced significantly. You will see from table 1 that we have included the barrier statistics for last year’s meeting held on the same day at Happy Valley over the same C Course and these figures line up well with the twenty-year averages.

 

Table 1

	25/2/24		25/2/23		Last 20 years
	St:1/2/3	Place %	St:1/2/3	Place %	Place %
Barriers 1 to 3	27: 0-1-3	14.8%	27: 4-2-2	29.6%	32.8%
Barriers 4 to 6	27: 3-5-2	37.0%	27:3-2-3	29.6%	28.1%
Barriers 7 to 9	27: 4-0-3	25.9%	27: 1-3-2	22.2%	21.3%
Barriers 10 to 12	25: 2-3-1	24.0%	26: 1-2-1	15.4%	18.8%

 

One had only to watch R1 to see the favourite lead from barrier one and get “bogged down” in the straight to finish in 6th place and the quinella horses close strongly down the outside of the track from the outside barriers 8 and 10 at the juicy odds of 22 and 7.1, to be on notice that something was not right. This was to be the pattern for the entire night apart from race three where the winner ran his first section in 29.20 thus illustrating the situation where a “pace bias” will overcome a “track bias”.

We asked ourselves “Why is it so?” and found some extremely interesting facts. The times for all races on the 15th were significantly slower than standard (see Table 2), a sum total of 8.59 seconds on the night, this even though there had been no rain and we noted that the weather and temperature were very similar to the meeting one year earlier.

In fact, at this earlier meeting (see Table 3) the times were close to standard. Why then was a 1650m race run a year previously 7 ½ lengths faster than the recent meeting? Why was the 2023 meeting run on a Good to Firm track and the 2024 meeting run on a Good to Yielding track?

 

Table 2

Date	Race	Class	Distance	Time	Standard	Difference
15/2/24	1	5	1200	1.10.98	1.10.20	+0.78
15/2/24	2	4	1650	1.41.07	1.40.00	+1.07
15/2/24	3	4	1650	1.41.08	1.40.00	+1.08
15/2/24	4	4	1200	1.10.87	1.09.80	+1.07
15/2/24	5	3	1200	1.10.51	1.09.50	+1.01
15/2/24	6	4	1000	57.27	57.05	+0.22
15/2/24	7	4	1200	1.10.81	1.09.80	+1.01
15/2/24	8	3	1200	1.10.27	1.09.50	+0.77
15/2/24	9	3	1800	1.50.78	1.49.20	+1.58

 

Table 3

Date	Race	Class	Distance	Time	Standard	Difference
15/2/23	1	5	1000	56.97	57.20	-0.23
15/2/23	2	5	2200	2.16.73	2.17.40	-0.67
15/2/23	3	4	1000	56.96	57.05	-0.09
15/2/23	4	4	1650	1.39.98	1.40.00	-0.02
15/2/23	5	4	1200	1.09.83	1.09.80	+0.03
15/2/23	6	4	1650	1.40.05	1.40.00	+0.05
15/2/23	7	4	1200	1.10.25	1.09.80	+0.45
15/2/23	8	3	1200	1.09.45	1.09.50	-0.05
15/2/23	9	3	1650	1.39.61	1.39.75	-0.14

 

We decided therefore to go back to the previous three meetings run at Happy Valley, in addition to the meeting on the 15 February and found that barrier one had not won a single race over the 35 races and figured in a quinella only once!

 

Four Meetings from 17/1- 15/2/24

Date	Course	Race#	PP 1st	PP 2nd	PP 3rd
17/01/2024	C	1	3	6	1
17/01/2024	C	2	5	9	1
17/01/2024	C	3	6	9	10
17/01/2024	C	4	4	6	5
17/01/2024	C	5	5	9	3
17/01/2024	C	6	4	10	5
17/01/2024	C	7	6	11	4
17/01/2024	C	8	2	3	1

 

Date	Course	Race#	PP 1st	PP 2nd	PP 3rd
31/01/2024	A	1	6	4	8
31/01/2024	A	2	12	11	1
31/01/2024	A	3	6	8	7
31/01/2024	A	4	3	9	2
31/01/2024	A	5	9	8	10
31/01/2024	A	6	2	9	4
31/01/2024	A	7	7	2	6
31/01/2024	A	8	12	3	11
31/01/2024	A	9	6	5	11

 

Date	Course	Race#	PP 1st	PP 2nd	PP 3rd
7/02/2024	B	1	2	5	7
7/02/2024	B	2	5	2	7
7/02/2024	B	3	5	11	1
7/02/2024	B	4	9	6	2
7/02/2024	B	5	7	9	1
7/02/2024	B	6	3	12	10
7/02/2024	B	7	6	4	10
7/02/2024	B	8	2	9	8
7/02/2024	B	9	8	5	4

 

Date	Course	Race#	PP 1st	PP 2nd	PP 3rd
15/02/2024	C	1	8	10	7
15/02/2024	C	2	8	5	10
15/02/2024	C	3	7	5	9
15/02/2024	C	4	4	6	8
15/02/2024	C	5	11	5	2
15/02/2024	C	6	4	10	5
15/02/2024	C	7	12	1	4
15/02/2024	C	8	7	6	2
15/02/2024	C	9	4	11	3

 

Summary Table

Barrier	1st	2nd	3rd	Quin %
1	0	1	6	2.9%
2	4	3	4	20.0%
3	3	3	2	17.1%
4	5	2	4	20.0%
5	4	6	3	28.6%
6	6	5	1	31.4%
7	3	0	4	8.6%
8	3	2	3	14.3%
9	2	6	1	22.9%
10	0	3	5	8.6%
11	1	4	2	14.3%
12	3	1	0	11.4%

 

 

Over all courses at Happy Valley the historical quinella strike rate for barrier one is 25.7% so over 35 races barrier one should figure in the quinella spot on average nine times. This is such an unlikely occurrence, statistically speaking, that there must be an underlying change responsible for this phenomenon. The obvious reason is that the track speed on the rail is significantly slower than the ground wider out on the track seemingly at all courses and this will not be only in the home straight but right around the course. So, you can see that a leader such as Great Days (R1 on 15^th February) was effectively “gassed” by the time he had reached the home straight as he was running against the rail in what has historically been even ground but these days the worst ground.

Form Pointers will continue to monitor the bias phenomenon at Happy Valley if and when it continues but for the time being we have this advice for punters. Look at the overall race time for R1 on a given day and compare it to the published standard times and if it is around one second slow beware! Replay R1 and look for any visually discernible bias. Be selective in your betting concentrating on horses that should be favoured by the bias. Bear in mind that if the track is riding fast that the bias could well be towards leaders and on pacers particularly if they are drawn low. Have a good set of speed maps, these are available for free on various websites. And finally, once any bias is clear riders will tend to want to take advantage of the bias and this is where a “pace bias” can overcome a “track bias”.

More from the TechTeam with the Sha Tin Report on Sunday.

Happy punting!

Preview of the 2024 Golden Slipper