|
Mark Newnham trained five winners from nine runners at Happy Valley on 18 March 2026 — each one winning a different way. The running positions and 47 years of history reveal a performance built on placement, not fortune.
Written by AI · Prompted by a Human
This article was researched and written entirely by an LLM (Large Language Model, a.k.a. AI). Sohil Patel provided the data, guided the analysis, and prompted the conclusions. The AI wrote the full article in under 3 minutes. Mark Newnham had a runner in all nine races at Happy Valley on 18 March 2026. Five of them won. The results were there for anyone to read the following morning, but results alone do not capture what made the evening remarkable. When the running positions are examined and the historical record consulted, a clearer picture emerges: this was not just a big night for a trainer in form. It was one of the rarest single-card performances in the 47-year history of modern Hong Kong racing. Soaring Bronco dead-heated in the opening race, halving his $44 dividend to $22 for every $10 unit. A $10 accumulator across the five winners would have returned $17,042. The Five Winners
Five Wins, Five Different WaysThe running positions tell their own story. Newnham's five winners did not benefit from a single prevailing track bias or one dominant riding pattern. They won from the front, from midfield, and from the rear of the field — five different tactical approaches, all successful on the same nine-race card. Armor Golden Eagle led at every call under A Atzeni in the Azalea Handicap over 1650 metres. Notthesillyone, ridden by R Kingscote from gate 12, sat second before taking over to win the Bellflower Handicap — racing without cover for nearly half the race. Crimson Flash, also under Atzeni, tracked in third and fourth before hitting the front in the Class 2 Daisy Handicap. Soaring Bronco, under J Orman, raced in fifth and sixth before dead-heating with Glorious Ryder in the opening Aster Handicap. And Ace War, the $1.75 favourite under Z Purton, was positioned ninth, ninth, tenth, and ninth before surging to win the 1800-metre Dahlia Handicap.
Four different jockeys rode the five winners, with Atzeni contributing a double. The range of classes was equally broad: Class 5 through Class 2, and distances from 1000 to 1800 metres. The whip counts add another dimension to the tactical variety. Orman used the whip 16 times aboard Soaring Bronco in a driving finish to the dead heat; Kingscote struck Notthesillyone seven times after racing without cover for nearly half the 1000 metres. Purton, aboard the even-money favourite Ace War, did not use the whip at all. This was not a case of one jockey in irresistible form carrying a yard. It was a trainer placing the right horse in the right race and four jockeys executing four different plans. The History: 14 Times in 47 YearsIn 3,551 HKJC race meetings from September 1979 to March 2026, a trainer has saddled five or more winners on a single card just 14 times. The all-time record belongs to Caspar Fownes, who trained six winners at Happy Valley on 20 June 2010; B Prebble rode five of them. No trainer in the modern era has managed seven. The complete list, with every winning horse, jockey, and $10 dividend:
The Accumulator TestOne way to measure the improbability of a multi-winner card is the hypothetical $10 accumulator — a rolling bet where each winner's return becomes the stake on the next. Dead-heat dividends are halved. The range across these 14 cards is vast:
Note: Dividends for races before April 2018 are derived from odds rounded to the nearest $10. Accumulators for those cards are approximations. Fownes' 2004 card at Sha Tin — which included Bold Winner at $500 for $10 — produced the largest hypothetical accumulator at nearly $2.7 million. Size's 2019 New Year's Day quintet, where all five winners were ridden by J Moreira at short prices, returned the smallest at $1,187. Newnham's $17,042 sits in the middle of the table, inflated by Notthesillyone's $262.50 dividend but tempered by the dead-heat deduction on Soaring Bronco. Several patterns emerge from the full list. G Moore achieved the feat three times as a trainer in the early 1980s, with G W Moore riding the majority of his winners. John Size's four appearances include the 2019 card where Moreira rode all five. D J Hall's five from just six runners on 1 May 2008 remains the highest strike rate of any card on this list. Newnham's quintet is notable for the range of jockeys used — four different riders — and the tactical diversity of the winning runs, from gate-to-wire to last-to-first. The Championship PictureNewnham entered the meeting with 36 winners for the 2025-2026 season, trailing championship leader Caspar Fownes by six. He left with 41, and after Fownes trained one winner on the same card, the gap stood at two. Context matters here. Newnham is only in his third Hong Kong season, having arrived from Australia in 2023. His progression has been steady:
* Season in progress, through 18 March 2026. Newnham's season started strongly — six winners from his first 43 runners in September, followed by an exceptional October that produced 14 winners from 63 runners at a 22 per cent strike rate, briefly placing him atop the table. From November through February, the numbers dropped: 15 winners across four months, a combined strike rate below 7 per cent. The quintet shifted March's tally to six winners from 45 runners, with five of those arriving in a single evening. The trainers' championship has some distance to run. What the data from 18 March establishes is that Newnham's quintet was not the product of a favourable draw, a track bias, or a single jockey in commanding form. Five winners across four classes and four distances, with running positions ranging from gate-to-wire to last-to-first. Whip counts that varied from 16 strikes in a driving dead-heat finish to zero aboard a horse that was simply too good for his field. The history books have seen it happen only 13 times before in nearly half a century. The numbers speak plainly enough. ■
About the Author
This article was researched and written entirely by an LLM (Large Language Model, a.k.a. AI). The AI wrote the full article in under 3 minutes. About the Prompter Sohil Patel is a Hong Kong racing handicapper based in San Francisco who focuses exclusively on HKJC racing. Unlike most punters who rely on publicly available form guides, Patel maintains a proprietary database of curated datasets spanning multiple seasons that allow him to find systematic edges in HK race betting. He spends more time than he should in prompting LLMs.
0 Comments
The Narrowest Track In Town: Why Happy Valley's C+3 Configuration Is Killing The Low-Draw Edge3/12/2026 The rail moves out, the track shrinks to 19.5 metres wide, and the starting gate shifts forward by a minimum of 36 metres and more than 100 on the longer trips. Thirty years of data show that when Happy Valley races on its tightest configuration, the old inside-draw playbook no longer applies.
Written by AI · Prompted by a Human
This article was researched and written entirely by an LLM (Large Language Model, a.k.a. AI). Sohil Patel provided the data, guided the analysis, and prompted the conclusions. The AI wrote the full article in under 3 minutes. Nine races at Happy Valley last Wednesday night and not one of them won by a horse drawn in gates one, two or three. All nine winners came from gate four or wider. Punters who lean on the inside draw at the city track — and there are plenty of them — would have left Wong Nai Chung Road with lighter pockets and a familiar excuse. Bad night. Fluke. It will correct itself. Maybe. But in 170 C+3 meetings at Happy Valley stretching back to October 1996, that was only the seventh time the low draws have been completely shut out. And the data from the past three racing seasons suggests something more structural is going on. What C+3 Actually MeansMost punters hear "C+3" on the race programme and think nothing of it. A course designation. Administrative detail. In reality, it is the single biggest factor shaping how a race at Happy Valley is run, and understanding it changes how you should bet. Happy Valley has seven course configurations, labelled A through C+3. The difference between them is where the inside running rail sits. ![]() On the A course, the rail is in its true position and the track is at its widest: 30.5 metres across with a 312-metre home straight. Horses drawn inside are running right next to the fence. The geometry is as favourable to low draws as Happy Valley gets. The other configurations exist to protect the ground. When the inside strip of turf gets worn from repeated racing, the HKJC moves the rail further from the true fence to give the grass time to recover. Each step narrows the usable racing surface and adjusts the home straight slightly. C+3 is the final step — the rail at its furthest point from the true inside position.
From 30.5 metres on the A course down to 19.5 on C+3. Eleven metres of racing width gone. That is a different track. The Gate Moves ForwardHere is the part most punters miss entirely. When the rail moves out, every horse travels a longer path around the turns because the circumference of the track increases. But the winning post does not move — it is fixed in place. So to keep the advertised race distance accurate, the HKJC shifts the starting gate forward, further up the track from the winning post, to compensate for the extra ground the field will cover on the bends. ![]() Happy Valley hosts five race distances — 1000, 1200, 1650, 1800 and 2200 metres — and at every one of them on C+3, the gate sits meaningfully ahead of where it would be on the A course.
At 1200 metres — the most common Happy Valley distance — the gate sits 49.3 metres ahead of where it would be on the A course. At 2200 metres, the shift is 111.7 metres. That changes the run into the first bend. It changes where horses settle in the field. It changes which draws get first use of the rail. Three things are working against the low draw on C+3. The track is narrower, so wide draws are no longer far from the fence. The home straight is longer, giving closers more room to finish over the top. And the gate is further forward, altering the tactical run to the first turn. Three Decades Of Fading EdgePut the configuration data alongside the results and the story writes itself. In the 2001–02 season, horses drawn one through three at C+3 meetings won at 18.7 per cent — close to one in five. The inside draw was a genuine weapon. By last season it had fallen to 9.7 per cent. This season it sits at 9.9. Two consecutive campaigns below 10 per cent, something that has not happened since the late 1990s. Across the most recent 20 C+3 meetings — from the start of the 2023–24 racing season through to March 2026 — the numbers look like this: 523 starters from draws one to three, 58 winners, an 11.1 per cent strike rate. For context, a random draw in a twelve-horse field gives you an 8.3 per cent baseline. The low-draw advantage has not disappeared. But it has been cut to a fraction of what it was.
The cumulative win rate across those 20 meetings: 11.1 per cent. The cumulative place rate: 32.9 per cent. Horses from the inside gates are still running into the frame regularly — roughly one in three finishes in the first three. They are just not converting. Other horses are getting past them in that 335-metre home straight. Distance MattersThe 1200-metre sprint remains the low draw's best friend at Happy Valley. Draws one to three win at 14.8 per cent over the distance, comfortably the highest of any trip on the card. At 1200, the gate shift of 49.3 metres on C+3 still leaves a relatively short run to the first turn. Speed from the inside can establish position early and the narrow track makes it difficult to roll forward from out wide. Stretch the trip and the advantage thins. At 1650 metres, the win rate drops to 11.2 per cent. The gate has moved 75 metres forward, the run to the first turn is longer, and riders from wider draws have more time and room to slot into position without burning fuel. By the time the field straightens for home, gate one and gate eight may as well have started from the same spot. The 1000-metre dash is an interesting case. You might expect the inside to dominate in a pure sprint around one turn. But at 1000 metres the field starts on the back straight with a long run before the home bend, and on C+3 the gate moves forward 36.7 metres, placing starters even further up the straight. Wide draws can use that run-in to find cover without the disadvantage they face at 1200. Seven Shutouts In Thirty YearsMarch 11 was rare, but it was not without precedent. There have been seven C+3 meetings in three decades where no horse from draws one to three managed a single win.
Even on the night itself, the place rate held at 22.2 per cent. Low-drawn horses were competitive. They just could not finish the job. The only time consecutive C+3 meetings produced zero low-draw winners was January and February 1998. Every other shutout has been a one-off followed by a rebound. History says March 11 was an outlier. When The Rail Bites, It Bites HardFor all the talk of decline, C+3 still produces days when the inside draw dominates. Thirty-three of 170 meetings — 19.4 per cent — have seen draws one to three win more than half the card. On November 24, 2010, low draws won seven of eight races, an 87.5 per cent strike rate that remains the all-time record for a single C+3 meeting. And when an outsider does emerge from a low draw, the prices can be extraordinary. Ping Hai Galaxy won at 160-1 from gate two on February 12, 2020, paying $1,603 for a ten-dollar ticket. Blue Illusion landed at 98-1 from draw two on December 3, 2025, returning $984. The low draw is not dead as a source of value. It has simply stopped being the automatic edge that punters treat it as. The Track Moved. The Market Hasn't.Here is the disconnect. The HKJC moves the rail out to C+3 and the track narrows to 19.5 metres. The starting gate shifts forward by up to 111 metres. The home straight grows. The geometry of every race on the card changes. And yet the betting market continues to price low draws as though the rail is in its true position and the track is 30 metres wide. The edge is not in the draw itself anymore. The edge is in knowing what the draw means on a given configuration — and right now, on C+3, it means less than it has at any point in the past 25 years. Happy Valley has not changed shape. But when the rail moves to C+3, it races like a different track entirely. The punters who adjust will find the ones who have not make generous donors to the pool. ∎
Written by AI · Prompted by a Human
About the Author About the Prompter Halfway through the season, a simple question — how many wins should each jockey have had? — produces an answer few expected. Written by AI · Prompted by a Human This article was researched and written entirely by an LLM (Large Language Model, a.k.a. AI). Sohil Patel provided the data, guided the analysis, and prompted the conclusions. The AI wrote the full article in under 2 minutes. EIGHTY WINNERS. Through 485 of approximately 850 races this season, Zac Purton sits atop the Hong Kong jockey premiership. The number commands respect. It should. But here is another number: 87.3. That is how many winners the betting market expected Purton to ride — based on the odds of every horse he sat on when the gates opened. Eighty wins this season versus an expectation that he should have won on nearly 88 mounts puts him at underperforming by more than seven winners. The biggest deficit of any jockey in Hong Kong this season. Dead last on a table few talk about. Now look at the top of that same table. C L Chau. Twenty-eight winners from an expected 20.7. Plus 7.3. Same number. Opposite sign. The premiership leader and the man outperforming his opportunities by the widest margin — separated by the mirror image of 7.3 wins. How Expected Wins WorkThe idea is simple. Take the market odds of every horse each jockey rode this season and calculate the implied number of wins — after adjusting for the track take. For the sake of this example, we leave the track take aside to illustrate the concept. A $2.00 favourite has a 50 per cent chance of winning. Ten rides on $2.00 favourites, five expected wins. If you win on six, you are one win ahead of what the market expects. If you win only on three, you are two winners behind. Add them all up across the season, compare the total to actual winners ridden, and the difference tells you who is winning more than they should — and who is winning less. The Problem With Raw NumbersConsider Ka Ying Rising. Five starts this season, five wins. His highest starting price? $1.05. At those odds, the horse was essentially a walkover with a saddle on it. Any jockey aboard was collecting. But Ka Ying Rising is not the only short-priced winner inflating the tally. Purton won on Little Paradise at $1.35, Super Strong Kid at $1.45, Bulb General at $1.50, and Invincible Ibis at $1.55. In total, 15 of Purton’s 80 winners this season came at odds shorter than $2.00 — rides where the market had already done most of the work before the gates opened. Expected wins strips that noise away. It asks a harder question: given the quality of your rides, are you winning more or less than the opportunities warranted? For Purton — so far this season — the answer is less. The McDonald TestHere is where it gets interesting. James McDonald rode Romantic Warrior four times this season. The prices? $1.05, $1.10, $1.40, and $1.55. All four won. Add Gold Patch at $1.55 and McDonald had five sub-$2.00 winners of his own — a higher proportion of his 16 total winners than Purton’s 15 from 80. Yet McDonald sits at minus 0.2. Dead level. Almost to the decimal. If riding short-priced horses was enough to explain the deficit, McDonald should be deep in the red too. He is not. The market gave both men cheap wins. Only one failed to bank enough of the harder ones. Who Is Beating The Market?Start with what Jerry Chau is working with. His shortest-priced winner this season was Flying Wrote at $2.30. Not a single sub-$2.00 walkover. No Ka Ying Rising. No Romantic Warrior. Where Purton’s longest-priced winner paid $13.15, Chau won on Celtic Times at $35.10, Healthy Happy at $22.15, and Lucky Twin Stars at $17.90 — horses the market had all but written off. He is not padding the scorecard. He is picking locks. Chau is not alone in outperforming. D B McMonagle and M F Poon both sit at plus 3.7 — finding winners the market did not expect. Andrea Atzeni and Hugh Bowman, each at plus 3.2, are delivering more than the odds suggested from their respective books of rides. Jerry Orman at plus 2.5 and Y L Chung at plus 2.4 round out a group punching above their weight. At the other end, the drop-off below Purton is steep but not solitary. Brenton Avdulla at minus 6.8 — eleven winners from 17.8 expected — is the second-biggest underperformer among the regulars. Matthew Chadwick at minus 5.0 and Karis Teetan at minus 3.8 fill out a bottom three that may surprise those following the premiership table alone. Then there is the middle — jockeys landing almost exactly where the market put them. C Y Ho at dead level. Lyle Hewitson one-tenth of a win below expected. Harry Bentley, three-tenths. If you believe the market is efficient, these riders are confirming it. The Full Table
Season data through 485 of approximately 850 races. Visiting jockeys with limited engagements included for completeness. What This Is NotThis is not a verdict on Purton’s season. Eighty winners at the halfway mark is elite by any measure. And the market is not infallible — odds can be wrong, and a jockey who consistently draws tough barriers or poorly-rated horses may see expected wins inflated by factors beyond their control. There is also the matter of market bias. Popular jockeys attract popular money. When Purton sits on a horse, the public bets it — and that weight of money pushes the odds shorter than they might otherwise be. Shorter odds mean higher implied probability, which means higher expected wins. If Purton’s mounts are systematically overbet by punters who follow the name rather than the form, his expected tally of 87.3 may itself be inflated. The market’s faith in the champion may, paradoxically, be the very thing that makes his numbers look worse on this table. Then again, McDonald is no stranger to public money either. And his numbers land dead level. On Sunday at Sha Tin, Jerry Chau rode his 200th Hong Kong winner. Plus 7.3. Zac Purton leads the premiership by a mile. Minus 7.3. ∎ About the Author This article was researched and written entirely by an LLM (Large Language Model, a.k.a. AI). The AI wrote the full article in under 2 minutes. About the Prompter Sohil Patel is a Hong Kong racing handicapper based in San Francisco who focuses exclusively on HKJC racing. Unlike most punters who rely on publicly available form guides, Patel maintains a proprietary database of curated datasets spanning multiple seasons that allow him to find systematic edges in HK race betting. He spends more time than he should prompting LLMs. |
Authorracequant used to cover races in India as the Racing Editor for the Times of India before he shifted his focus on Hong Kong racing Archives
March 2026
Categories |


RSS Feed