The StatsMan

THE STATSMAN: Stung by Yung … again

📸 Lee Ming Donald

Top Military's 233.5-to-1 Happy Valley win was one for the history books but raises questions on the level of scrutiny on big form turnarounds

By Sohil Patel

On Wednesday night, Top Military won against odds of 283.5-to-1 and became the highest priced winner to win at Happy Valley in Hong Kong racing history. When trainer Benno Yung Tin-pang was questioned in regards to the improved performance of his horse, his only explanation was the application of blinkers. Not many handicappers would have included Top Military in their top four selections. After adjusting for the Hong Kong Jockey Club takeout, Top Military’s winning chances were estimated by the betting public to be less than 0.3%.  Understandably, some racegoers were irate. Form reversals like Top Military’s hurt customer confidence and leave the Hong Kong Jockey Club officials with some tough questions to ask. 

Firstly, are the HKJC stewards asking enough tough questions when horses with no apparent recent form show sudden improvement?  Should the explanations provided by trainers and jockeys (provided with full transparency in the Racing Incident Reports) be accepted at face value?  Or do more searching questions need to be asked about the frequency and reasoning behind extreme form reversals? 

Close followers of Hong Kong racing form know this wasn’t the first time punters were stung by Yung. The trainer’s horses have a history of producing baffling form reversals and winning at big odds.  

 One of the draws of horse racing is in the balance of predictability and unpredictability.   It is wonderfully predictable in that the use of carefully considered data sets (including past performances, barrier trials, track work, jockey and trainer patterns, breeding, paddock looks and other factors) allows a diligent handicapper to accurately pinpoint a greater percentage of winners as opposed to picking a horse at random.  A random pick in any given race in Hong Kong would allow one to correctly predict winners in 8.2% of the races (one out of the average 12.2 runners in a race) – but competent handicappers do significantly better.

Horse racing is also highly unpredictable in that even the best handicapper will be more often wrong than right. In Hong Kong, a highly competitive racing jurisdiction, the leading English-language tipster on the South China Morning Post (SCMP website) is RACING POST ONLINE.  This tipster gets less than 25% of his first choices right (173 out of 718 races which is 24.09%). While a 24.09% success rate is nearly three times better than the average 8.2% when picked at random, it is still three times worse than the 75.91% of his tips that do not win. The numbers are similar when we look at published records in the Chinese press. 

 If there was a formulaic way of consistently predicting triple figure winners, then these horses would not be triple figures. In the absence of such a bankable formula, finding the next Yung-trained long shot is a difficult journey, and not one we will be embarking on here. Instead, we break down data from the current season to highlight which trainers are hardest to catch when it comes to upset results. We were not surprised to find Yung on top of the list when it came to horses displaying the biggest number of significant form swings.

To configure the list, we looked at form reversal of more than ten lengths. This could either be one of two situations: 

  1. Performs well after a bad run: This is when horses either win a race or finish a length or less off the winner and exhibits an improvement of more than ten lengths compared to its previous start (provided that both starts were in the same class). 
  2. Performs poorly after a good run: This includes horses that followed a good run (won or ran within a length of the winner) with a dismal run (difference of more than ten lengths) in the same class.

In both approaches, we have deliberately ignored all other factors such as change in jockey, change in body weight of the horse, change in weight carried, change in racing surface and other factors that could reasonably explain some of the deviation in form.  The ten-length difference benchmark reasonably allows us to do that. Also, while we could correlate observations in the Racing Incident Reports (such as significant interference) and entries in the Veterinary Database (such as mucus, bleeding, lameness) to ‘explain away’ poor performances, we have again intentionally chosen not to.  We have ignored horses that unseated their riders, were pulled up, took no part or did not finish their race from this analysis.

The graph below shows  trainers with horses that performed well after a bad run. An example of this would be Top Military, who won on Wednesday night at record-breaking odds, while in his previous start in the same class less than a month ago, had finished 12-lengths behind.



With five runners: Ka Ying Legend, Play To Win, Voyage Star, Striking Mr C, and Top Military, Yung topped this season’s chart for horses that perform well well after a bad run.

Yung started out at the HKJC as an apprentice jockey in 1978.  He then worked as an assistant trainer to John Size and got his license to train ahead of the 2013-14 season.  Some racing observers note that Yung follows a version of the John Size ‘soft’ training template (in that he does not overly push his horses in track work). Even still, Yung has not been able to decipher the template well enough to be as successful as the third year trainer Frankie Lor, another one of John Size’s former “ATs” who is often recognised as Size’s most successful protégé.

Interestingly, in March 2020, HKJC was investigating the case of Yung-trained Ka Ying Legend (one of the five in the list).  A pre-race urine sample taken from Ka Ying Legend during a race day was found to contain the prohibited substance clenbuterol.  As per HKJC protocol, the withdrawal period for medications is more conservative than in other jurisdictions.  This is also true for Clenbuterol for which the HKJC recommends a 21 clear day withdrawal for horses 3.5 years of age or older.

Ka Ying Legend had run a half-length third at massive odds (68-to-1) after running 22 and 37 lengths behind in his previous two starts.  Later, a second analysis confirmed low levels of Clenbuterol in urine samples from Ka Ying Legend.  The HKJC strives to ensure that all Hong Kong races are run with uncompromising integrity.  To maintain this standard, the state-of-the-art HKJC testing lab conducts integral pre-race, post-race and in-training testing of blood, urine and hair samples to check for the use of banned substances in horses or jockeys.  The HKJC facility is internationally recognised for the quality of its work and has never reported a “false positive” finding in official samples since it was established in 1970.  Yung was fortunate that the club deemed Ka Ying Legend’s second sample to be negative as clenbuterol was present at a level that did not exceed the international screening limit.  An investigation into the positive swab is ongoing. 

 So, other than Yung, which trainers are hardest to catch? Here are the other trainers that have had higher numbers of positive form turnarounds than the season’s baseline (shown in orange in the above graph) and the horses that have flipped the form guide. 


  • Dennis Yip:  Murray’s Partners, Hinyuen Swiftness, Reeve’s Muntjac and Iron King
  • Ricky Yiu: Alcari, Focus, Glorious Spectrum, and Loriz
  • Tony Cruz: Sunshine Warrior, California Rad, and Mr Aldan
  • Tony Millard: Yee Cheong Lucky and Amazing Chocolate twice
  • Danny Shum: Triumphant Lord, Charity Wings, and Craig’s Star
  • Peter Ho: King Mortar, Association Fans twice


The second situation, which we label ‘Performed poorly after a good run’, includes horses that followed a good run (won or ran within a length of the winner) with a dismal run (difference of more than ten lengths) in the same class. An example of this is the Francis Lui-trained Undefeated, who followed a win with a 31.5-length defeat to finish last in his next start a month later (albeit on a different surface).


Here again, Yung’s horses figure in the list at a rate that is more than twice the baseline (shown as the orange line in the graph above).  Coolceleb, Flying Monkey, Striking Mr C, Beauty Day, and Gintoki form Yung’s list of five horses that all performed terribly in their next start in the same class after finishing within a length of the winner.

The other trainers that have had higher numbers than the season’s baseline (shown in orange in the above graph) are as under:


  • Peter Ho:  Multimax, King Mortar, Amazing Agility, and Viva Council twice
  • Ricky Yiu: Glorious Spectrum, Megatron, Alcari, and Snow Airjet
  • Caspar Fownes: Joyful Trinity, Sky Gem, and Nice Fandango
  • Dennis Yip: El Jefe, Hinyuen Swiftness, and Sam’s Love
  • Danny Shum: Break Record, Young Legend, Lockheed
  • Michael Chang: The Joy of Giving, Show Mission, Mega Heart


Getting a trainer’s license in Hong Kong is even more difficult than getting a jockey’s contract.  In the past three seasons, only three new trainers (Frankie Lor, Jimmy Ting and Douglas Whyte) have been added to the roster. Former champion David Hayes will return from Australia for the start of next season. The HKJC follows a stringent process to evaluate candidates and only allows well-qualified, well-trained and (controversially) ‘younger than 70’ professionals to train in Hong Kong.  Additionally, trainers that do not reach HKJC’s performance benchmark of 16 wins for the season for three seasons are not allowed to keep their licence.  Trainers are well compensated, the leader in prize money this season is Tony Cruz, whose horses have won more than HK$111 million in prize money. Trainers get a 9.2% cut of the purse, so Cruz gets to keep HK$10.25m of this and is also able to share an additional 10.8% of the prize money with his assistant trainer and stable staff.  Even Yung, who sits in 18th spot in the trainer’s leaderboard, has seen his horses earn more than HK$25m in prizemoney this season.

The level of scrutiny from the stipendiary stewards in Hong Kong is higher than in any racing jurisdiction in the world.  The Racing Incident Report, which is published after the day’s races accurately details, among other things, incidents and circumstances that contribute to the under performance of a horse.  ‘What more could the HKJC Stewards do?” is indeed a tough question to answer. After all, it would seem unfair to penalise Yung after Top Military performs well, even if the result leaves a bad taste in punters’ mouths. In all fairness, Kim Kelly, the chief stipendiary steward, and his team are doing a fantastic job already. Horses that turn in sub-par performances are sent back to trial and in some cases horses are observed in a series of consecutive trials. 

In the specific example of Top Military’s shock win, HKJC had issued a press release before the race which communicated that they had been provided with written advice from New Zealand Thoroughbred Racing that Top Military wore blinkers in its race starts conducted in that jurisdiction (in fact Top Military had finished a neck second with blinkers on).  Perhaps, prior to approving Yung’s application for that horse, they could have mandated that it be subjected to a blinkers test and also that the horse trial at least once with the blinkers on.  This would have allowed racegoers to see the improvement in Top Military with the blinkers on as part of their pre-race analysis, instead of seeing it in a race where big sums of money were wagered.

The second area which could do with a recalibration is raising the threshold of what constitutes a valid explanation and what seems frivolous. Jockeys and trainers are questioned and their explanations are detailed in the Racing Incident Report.  While most of these explanations are detailed and allow the betting public to understand what caused the form changes, some of these seem frivolous. For the punters for whom Top Military has been the difference in a potentially lucrative double trio or six-up collect, some of the trainers’ explanations read like weak “dog at my homework”-style excuses.  Perhaps the threshold for accepting explanations could well be raised for extreme form reversals and trainers should expect more scrutiny after an upset result. Until then Yung can consider himself fortunate that he could get by with his ‘blinkers did the trick’ explanation.


THE STATSMAN: a serious statistical deep dive into Class 3 PPs on debut