Hong Kong Horse Racing (HKJC) Data, Odds & Results Scraper avatar

Hong Kong Horse Racing (HKJC) Data, Odds & Results Scraper

Pricing

from $15.00 / 1,000 results

Go to Apify Store
Hong Kong Horse Racing (HKJC) Data, Odds & Results Scraper

Hong Kong Horse Racing (HKJC) Data, Odds & Results Scraper

Hong Kong horse racing (HKJC) data in clean JSON: live odds and market movers, odds movement with steamer/drifter tags and results for backtesting, race results, pre-race analysis, barrier trials, vet records and stewards' reports. 香港賽馬數據:賽果、即時賠率及賠率走勢、賽前分析、試閘、獸醫記錄、競賽事件報告。

Pricing

from $15.00 / 1,000 results

Rating

5.0

(2)

Developer

Alaricus

Alaricus

Maintained by Community

Actor stats

4

Bookmarked

106

Total users

0

Monthly active users

3 days ago

Last modified

Features

Limited permissions

Share

HKJC Scraper Actor

香港賽馬數據:賽果、即時賠率及賠率走勢、賽前分析、試閘、獸醫記錄、競賽事件報告。

What does the Hong Kong Horse Racing (HKJC) Data Scraper do?

The Hong Kong Horse Racing (HKJC) Data, Odds & Results Scraper is an all-in-one intelligence tool designed to extract high-fidelity data from the official Hong Kong Jockey Club platform. It provides professional-grade access to the world’s most lucrative horse racing ecosystem, covering everything from historical race results to deep medical veterinary history.

This Actor is built for professional handicappers, data scientists, and betting syndicate developers who require structured, normalized JSON data to power their predictive models and racing dashboards in 2026.

Key Features

  • Multi-Service Architecture: Access eight distinct data streams within a single Actor:

    • Race Results: Comprehensive finishing data, sectional times, and margins.
    • Pre-Race Analysis: Advanced predictive metrics including 400m sectional rankings, closing momentum profiles, "Fatigue" scores, and full historical sectional times for upcoming races.
    • Incident Reports: Official steward reports and racing incidents for performance analysis.
    • Barrier Trials: Morning trial performance data and results.
    • Veterinary Records: Complete health history including lame records, surgeries, and "passed" dates.
    • Odds Snapshot: Win/place odds for every runner captured 7 times before each race (3h to the jump).
    • Odds Movement & Results: Every odds snapshot enriched with market rank, implied win probability, odds movement, steamer/drifter tags and the horse's finishing result. Backtest-ready.
    • Live Market Movers: Today's remaining races with current odds and how the market has moved since the stored snapshots. Check it by hand, call it from your own code, or schedule it on race days.
  • One-Click Bulk Extraction: Features "Scrape ALL available dates" toggles to automatically harvest years of historical data without manual date entry.

  • Targeted Veterinary Search: Specific lookup mode for "Today's Declared Starters" or deep-history search by Horse Brand Numbers (e.g., K458).

Why scrape HKJC data?

Hong Kong horse racing is a data-driven sport. This tool provides the "Home of Truth" data necessary for:

  • Algorithmic Betting: Feed clean, historical datasets into ELO or machine learning models.
  • Horse Health Monitoring: Track veterinary history and "passed" trials to identify horses returning from injury.
  • Jockey & Trainer Analysis: Monitor weight allowances and performance trends across Sha Tin and Happy Valley.
  • Steward Insights: Extract incident reports to find "unlucky" runners who faced interference but didn't place.

How to use the Scraper

  1. Select Service Mode: Choose between Results, Pre-Race Analysis, Reports, Trials, Veterinary, Odds Snapshot, Odds Movement & Results, or Live Market Movers.
  2. Configure Date Logic:
    • Bulk Mode: Check the "Scrape ALL available dates" toggle to automatically discover and extract every date currently listed on the HKJC dropdowns (typically 150-200+ dates).
    • Targeted Mode: Provide a specific list of dates in YYYY-MM-DD format.
  3. Advanced Filters (Optional):
    • Race Numbers: Leave empty for all races, or specify (e.g., 1, 8, 10) for targeted race analysis.
    • Brand Numbers: For Veterinary mode, enter specific IDs (e.g., H095) to get full lifetime medical history.
  4. Click Run: The scraper handles the navigation, dropdown selection, and table extraction.
  5. Download Data: Export your structured dataset in JSON, CSV, or Excel.

Input Parameters

The parameters are organized into logical sections. Only the fields related to your selected Mode will be processed.

🛠 Input Configuration & Mode Mapping

The scraper is highly modular. Depending on the mode you select, different input parameters become active. Use the table below to configure your JSON input or UI settings.

Service Mode (mode)Parameter KeyTypeDescriptionOptions / Format
Race Results (result)all_result_dates
result_dates
result_race_numbers
Boolean
Array
Array
Toggle bulk extraction.
Target specific dates.
Target specific races.
true / false
["YYYY-MM-DD"]
[1, 5] (Empty = ALL)
Pre-Race Analysis (prerace_analysis)prerace_race_numbersArrayTarget specific upcoming races.[1, 5] (Empty = ALL)
Incident Reports (report)all_report_dates
report_dates
Boolean
Array
Toggle bulk extraction.
Target dates for reports.
true / false
["YYYY-MM-DD"]
Barrier Trials (barrier_trial)all_barrier_dates
barrier_dates
Boolean
Array
Toggle bulk extraction.
Target dates for trials.
true / false
["YYYY-MM-DD"]
Veterinary Records (veterinary)get_starters
get_full_database
brand_numbers
Boolean
Boolean
Array
Scrape upcoming runners.
Enable history search.
Target Horse IDs.
true / false
true / false
["K458", "H095"]
Odds Snapshot (odds_snapshot)all_snapshots
snapshot_dates
snapshot_race_numbers
Boolean
Array
Array
Toggle bulk extraction.
Target specific dates.
Target specific races.
true / false
["YYYY-MM-DD"]
[1, 5] (Empty = ALL)
Odds Movement & Results (odds_movement)all_movement_dates
movement_dates
movement_race_numbers
Boolean
Array
Array
Toggle all stored dates.
Target specific dates.
Target specific races.
true / false
["YYYY-MM-DD"]
[1, 5] (Empty = ALL)
Live Market Movers (live_movers)live_race_numbersArrayTarget specific races of today's meeting.[1, 5] (Empty = ALL remaining)

Pro Tip: When using the API, ensure your mode matches the parameters you are sending. For example, if mode is set to report, the scraper will ignore race_numbers.


Input and Output Examples

📸 ODDS SNAPSHOT MODE

This mode provides high-frequency, precision snapshotting of live horse racing odds markets. Instead of giving you a single data point, it automatically tracks market movements and price fluctuations at critical intervals leading up to the jump.

For every single horse in a scheduled race, you will receive 7 distinct snapshots captured at the following precise intervals:

  • 3 Hours before the official start time (Early baseline odds)

  • 2 Hours before the start time

  • 1 Hour before the start time (Market begins to heat up)

  • 45 Minutes before the start time

  • 30 Minutes before the start time

  • 15 Minutes before the start time (Peak liquidity window)

  • 0 Minutes (The exact scheduled start time to capture the official opening/closing price matrix)

Captured Data Points Per Snapshot

Every single one of the 7 timeline snapshots delivers a structured data object containing:

  • Race Metadata: Country, track name, race number, scheduled time, and track conditions.

  • Runner Profiles: Horse name, number, stall, jockey, trainer, and weight.

  • Odds Matrix: Win/Place fixed prices, tote fluctuations, and exchange trading values at that exact minute mark.

Key Benefits for Data Consumers

Track Market Sentiment: Perfect for building historical datasets to analyze "steamer" and "drifter" horses (where odds drop or spike dramatically right before the race).

Arbitrage & Value Detection: Easily back-test strategies by comparing early 3-hour baseline prices against the final closing odds at the jump.

Clean, Structured Outputs: Data is delivered directly to your dataset storage ready to be parsed into analytical models, sports trading bots, or odds comparison platforms.

Example Input

  1. The "Full Snapshot Archive" (Bulk Mode) Best for users wanting to extract the entire database of historically tracked odds snapshots.
{
"mode": "odds_snapshot",
"all_snapshots": true
}
  1. The "Targeted Snapshot" (Specific Races) Extract odds tracked specifically on a given day for exact races.
{
"mode": "odds_snapshot",
"snapshot_dates": ["2026-06-25", "2026-06-27"],
"snapshot_race_numbers": [1, 2]
}

Example Output

The output would be a list of dictionaries tracking odds at specific points in time:

{
"id": 2876,
"snapshot_id": 373,
"race_date": "2026-06-27",
"venue_code": "ST",
"race_no": 1,
"no": 7,
"standby_no": null,
"status": "Declared",
"horse_name": "NO OTHER CHOICE",
"horse_id": "HK_2024_K049",
"horse_code": "K049",
"draw": 11,
"handicap_weight": 131,
"current_weight": 1061,
"current_rating": 36,
"gear_info": "B1/TT",
"allowance": null,
"trainer_preference": 1,
"last6run": "3/9/5/3/5/10",
"trump_card": false,
"priority": false,
"final_position": 0,
"dead_heat": false,
"jockey_code": "MOJ",
"jockey_name": "J Moreira",
"trainer_code": "FC",
"trainer_name": "C Fownes",
"win_odds": 7.2,
"place_odds": 3.6,
"created_at": "2026-06-27T05:00:23.571984+00:00",
"updated_at": "2026-06-27T05:00:23.571984+00:00",
"tip_index_odds": 7.2,
"tip_index_initial": 6.3,
"tip_index_race_day": 8.8
}

📈 ODDS MOVEMENT & RESULTS MODE

This mode turns the raw odds snapshots into a backtest-ready market dataset. You get the same rows as Odds Snapshot (one row per runner per snapshot, 7 snapshots from 3 hours before the race to the jump), and every row also tells you how the market moved and how the horse finished.

Use it to answer questions like "how often do horses that shorten 20%+ in the last hour win?" or "is the tips index better than the final odds?" without building your own pipeline.

Coverage: odds snapshots are stored from the 2026-06-27 meeting onwards, and every new meeting is added automatically. Dates before that return no rows.

Example Input

  1. All stored dates
{
"mode": "odds_movement",
"all_movement_dates": true
}
  1. Specific meeting and races
{
"mode": "odds_movement",
"movement_dates": ["2026-11-14"],
"movement_race_numbers": [5]
}

Added Fields

Every Odds Snapshot field is included, plus:

FieldMeaning
snapshot_index / minutes_to_postWhich snapshot (1–7) and how many minutes before the scheduled start it was taken (180, 120, 60, 45, 30, 15, 0).
market_rankPosition in the betting market at that snapshot (favourite = 1, equal odds share a rank).
implied_win_probWin probability implied by the odds with the bookmaker margin removed (0–1).
market_overroundSum of 1/odds across the field at that snapshot (the pool's built-in margin).
odds_change_pct_prev% change in win odds since this horse's previous snapshot.
odds_change_pct_open% change in win odds since this horse's first snapshot (normally 3 hours out). Negative = backed in.
rank_change_openPlaces gained in the market since the first snapshot (positive = moved towards favouritism).
movement_tagsteamer (odds shortened 20% or more since the first snapshot), drifter (lengthened 25% or more) or stable.
finish_position / lbwOfficial finishing position and lengths behind the winner (null if the horse did not finish).
result_statusFINISHED, DEAD_HEAT, or why the horse has no position, e.g. PULLED_UP, UNSEATED_RIDER, WITHDRAWN_VETERINARY.
won / placedtrue if the horse won / finished in the top 3. false for any horse that started and lost, including those that pulled up, fell or unseated. null if the horse did not start (withdrawn) or there is no result.

Runners without odds at a snapshot (scratched, no market yet) get null market fields and are left out of that snapshot's rank and overround. Result fields are null until the race result has been stored (results are added daily after each meeting), and stay null for standby runners that did not start and for abandoned races. The snapshot-time final_position field is not included in this mode; use finish_position.

Example Output

GREEN SUPERSTAR, 2026-11-14 Race 5, 15 minutes before the jump: its odds fell from 27.0 (3 hours out) to 13.0, it moved from 9th to 5th in the market and was tagged a steamer. It finished 2nd.

{
"id": 27875,
"snapshot_id": 2142,
"race_date": "2026-11-14",
"venue_code": "ST",
"race_no": 5,
"no": 9,
"standby_no": null,
"status": "Declared",
"horse_name": "GREEN SUPERSTAR",
"horse_id": "HK_2026_M134",
"horse_code": "M134",
"draw": 5,
"handicap_weight": 125,
"current_weight": 1127,
"current_rating": 65,
"gear_info": "H1/XB1",
"allowance": null,
"trainer_preference": 1,
"last6run": "",
"trump_card": false,
"priority": false,
"dead_heat": false,
"jockey_code": "OJM",
"jockey_name": "J Orman",
"trainer_code": "SJJ",
"trainer_name": "J Size",
"win_odds": 13.0,
"place_odds": 6.1,
"created_at": "2026-11-14T06:40:26.806954+00:00",
"updated_at": "2026-11-14T06:40:26.806954+00:00",
"tip_index_odds": 13.0,
"tip_index_initial": 99.0,
"tip_index_race_day": 99.0,
"snapshot_index": 6,
"minutes_to_post": 15,
"market_overround": 1.2403,
"implied_win_prob": 0.062,
"market_rank": 5,
"odds_change_pct_prev": -48.0,
"odds_change_pct_open": -51.9,
"rank_change_open": 4,
"movement_tag": "steamer",
"finish_position": 2,
"lbw": 0.05,
"result_status": "FINISHED",
"won": false,
"placed": true
}

⚡ LIVE MARKET MOVERS MODE

This mode shows the market right now for today's races that haven't started yet. Each run fetches the current HKJC Win and Place odds and the Tips Index, and compares every runner with the stored snapshots for that race: its opening price (3 hours before the race) and its most recent snapshot. Steamers and drifters are tagged, so you can see where the money is going before the race, not after it.

How it differs from Odds Snapshot and Odds Movement

  • Odds at the moment you ask. Stored snapshots are taken at 180, 120, 60, 45, 30 and 15 minutes before the race and at the off. Between those points, and especially in the last 15 minutes, only this mode has fresh odds.
  • One row per runner. You get the current odds with the opening and previous prices as columns, not every snapshot again. Checking often does not mean paying for the same rows again.
  • Today's remaining races only, at the standard price ($15 per 1,000 rows).

How to use it

  1. Check by hand: run the Actor whenever you want a look, for example 10 minutes before a race with "live_race_numbers": [5]. Results are ready within moments.
  2. From your own code: call the Actor through the Apify API from your model, bot or dashboard at the times you choose (e.g. 30, 15 and 5 minutes before each race) and read the run's dataset.
  3. Hands-off (optional): create an Apify Schedule for this Actor, e.g. every 15 minutes from 09:00 to 23:59 Hong Kong time, and add an integration (Slack, email, a webhook to your own Telegram bot, or Zapier/Make into Google Sheets) to receive each run's output automatically. Hong Kong race days vary, so scheduling every day is simplest: non-race days return 0 rows.

Good to know

  • Best window: from 3 hours to about 5 minutes before a race. Each run takes a few seconds to start and fetch, and HKJC is a pool (pari-mutuel) market where final dividends are set after betting closes, so this mode is for watching the market, not for last-second betting.
  • "Previous" means our latest stored snapshot, not your previous run. To track change between your own runs, compare consecutive runs using captured_at.
  • Checking one race often? Use live_race_numbers so each run returns only that race (about 14 rows).
  • Only races that haven't started are returned, so runs get smaller as the meeting goes on. Finished races appear in Odds Movement & Results after the nightly results sync.
  • Every runner of every remaining race is returned (standby horses included, with empty odds), with the same fields as Odds Movement minus the result fields. odds_updated_at tells you when HKJC last updated the Win pool.
  • On days without a meeting, or once every race has started, the run finishes with 0 rows, so no rows are charged.

Example Input

{
"mode": "live_movers",
"live_race_numbers": []
}

Added Fields

Every Odds Movement field except the results (finish_position, lbw, result_status, won, placed), plus:

FieldMeaning
open_win_oddsThe runner's opening price: its earliest stored snapshot with odds (normally 3 hours before the race).
prev_win_oddsThe runner's latest stored snapshot with odds (180, 120, 60, 45, 30 or 15 minutes before the race).
odds_change_pct_open / odds_change_pct_prev% change of the current odds versus open_win_odds / prev_win_odds.
minutes_to_postMinutes until the scheduled start, at the time of the run.
odds_updated_atWhen HKJC last updated the Win pool.
captured_atWhen this row was captured (UTC).

A runner with no stored odds yet has no opening price, so its movement fields are empty.

Example Output (abridged)

GREEN SUPERSTAR, 2026-11-14 Race 5, 15 minutes before the jump (this is the free-plan sample):

{
"race_date": "2026-11-14",
"race_no": 5,
"horse_name": "GREEN SUPERSTAR",
"horse_code": "M134",
"win_odds": 13.0,
"place_odds": 6.1,
"minutes_to_post": 15,
"implied_win_prob": 0.062,
"market_rank": 5,
"open_win_odds": 27.0,
"prev_win_odds": 25.0,
"odds_change_pct_open": -51.9,
"odds_change_pct_prev": -48.0,
"rank_change_open": 4,
"movement_tag": "steamer",
"odds_updated_at": null,
"captured_at": "2026-11-14T06:40:00+00:00"
}

📊 PRE-RACE ANALYSIS MODE

Example Input

  1. The "Upcoming Card" Analysis Automatically pulls upcoming race data and calculates momentum and fatigue profiles based on historical sectional times.
{
"mode": "prerace_analysis"
}
  1. Targeted "Feature Race" Analysis Specify an exact race number to pull historical form metrics just for that race.
{
"mode": "prerace_analysis",
"prerace_race_numbers": [7]
}

Example Output

The output would be a list of dictionaries, such as the following:

{
"mode": "prerace_analysis",
"race_no": 1,
"date": "2026-04-26",
"racecourse": "SHA TIN",
"track": "TURF",
"course": "A Course",
"race_class": "Class 4",
"rating_range": "060-040",
"distance": 1200,
"race_name": "FWD INSURANCE ACT PRIVATE HANDICAP",
"post_time": "12:30 PM",
"horse_name": "BEAUTY GEMINI",
"brand_no": "L129",
"runner_no": 12,
"rank": null,
"fitness_rating": 1,
"weight": 122,
"body_weight": 1124,
"jockey": "Y L Chung",
"trainer": "A S Cruz",
"horse_comments": "",
"draw": 4,
"age": 3,
"horse_url": "https://racing.hkjc.com/en-us/local/information/horse?horseid=HK_2025_L129",
"horse_id": "L129",
"last_run_date": "2026-01-25",
"last_run_energy": 70,
"last_run_wide": "",
"last_run_venue": "SHA TIN",
"last_run_course": "A+3",
"last_run_distance": 1000,
"last_run_going": "GOOD",
"last_run_finish_pos": 12,
"last_run_draw": 4,
"last_run_weight": 127,
"last_run_body_weight": 1110,
"last_run_jockey": "A Badel",
"last_run_odds": 118.0,
"last_run_total_starters": 14,
"last_run_pace": "Good",
"last_run_comments": "Pace Good; Jumped awkwardly, positioned 1 to 2 lengths behind the leader on the middle track, peaked 300M, weakened thereafter.",
"last_run_race_url": "https://racing.hkjc.com/en-us/local/information/localresults?racedate=2026/01/25&Racecourse=ST&RaceNo=1",
"last_run_race_no": 1,
"last_run_track_type": "TURF",
"last_run_class": 4,
"last_run_lbw": 10.25,
"last_run_lbw_raw": "10-1/4",
"last_run_gear_raw": "TT",
"last_run_gear": [
{
"equipment": "TONGUE_TIE",
"action": "RETAINED"
}
],
"last_run_finish_time": 58.37,
"last_run_sec1_pos": 5,
"last_run_sec1_margin": 1.5,
"last_run_sec1_margin_raw": "1-1/2",
"last_run_sec1_time_rank": 5,
"last_run_sec2_pos": 9,
"last_run_sec2_margin": 2.0,
"last_run_sec2_margin_raw": "2",
"last_run_sec2_time_rank": 11,
"last_run_sec3_pos": 12,
"last_run_sec3_margin": 10.25,
"last_run_sec3_margin_raw": "10-1/4",
"last_run_sec3_time_rank": 12,
"last_run_position_change": -3,
"last_run_closing_profile": "Faded",
"last_run_sec1_time_total": 13.16,
"last_run_sec1_time_split1": null,
"last_run_sec1_time_split2": null,
"last_run_sec2_time_total": 21.13,
"last_run_sec2_time_split1": 10.25,
"last_run_sec2_time_split2": 10.88,
"last_run_sec3_time_total": 24.08,
"last_run_sec3_time_split1": 11.46,
"last_run_sec3_time_split2": 12.62,
"last_run_final_200m_fatigue": 1.16,
"last_run_fatigue_profile": "Collapsing",
"2nd_last_run_date": "2026-01-01",
"2nd_last_run_energy": 82,
"2nd_last_run_wide": "2W3W",
"2nd_last_run_venue": "SHA TIN",
"2nd_last_run_course": "B+2",
"2nd_last_run_distance": 1200,
"2nd_last_run_going": "GOOD",
"2nd_last_run_finish_pos": 14,
"2nd_last_run_draw": 7,
"2nd_last_run_weight": 128,
"2nd_last_run_body_weight": 1099,
"2nd_last_run_jockey": "Z Purton",
"2nd_last_run_odds": 5.0,
"2nd_last_run_total_starters": 14,
"2nd_last_run_pace": "Good to slow",
"2nd_last_run_comments": "Pace Good to slow; (2W3W) Sluggish start, settled in rear of mid-field, forced to steady on heels 550m out, left behind home straight.",
"2nd_last_run_race_url": "https://racing.hkjc.com/en-us/local/information/localresults?racedate=2026/01/01&Racecourse=ST&RaceNo=2",
"2nd_last_run_race_no": 2,
"2nd_last_run_track_type": "TURF",
"2nd_last_run_class": 4,
"2nd_last_run_lbw": 5.0,
"2nd_last_run_lbw_raw": "5",
"2nd_last_run_gear_raw": "TT1",
"2nd_last_run_gear": [
{
"equipment": "TONGUE_TIE",
"action": "FIRST_TIME"
}
],
"2nd_last_run_finish_time": 70.98,
"2nd_last_run_sec1_pos": 10,
"2nd_last_run_sec1_margin": 4.0,
"2nd_last_run_sec1_margin_raw": "4",
"2nd_last_run_sec1_time_rank": 10,
"2nd_last_run_sec2_pos": 12,
"2nd_last_run_sec2_margin": 4.25,
"2nd_last_run_sec2_margin_raw": "4-1/4",
"2nd_last_run_sec2_time_rank": 12,
"2nd_last_run_sec3_pos": 14,
"2nd_last_run_sec3_margin": 5.0,
"2nd_last_run_sec3_margin_raw": "5",
"2nd_last_run_sec3_time_rank": 13,
"2nd_last_run_position_change": -2,
"2nd_last_run_closing_profile": "Faded",
"2nd_last_run_sec1_time_total": 24.62,
"2nd_last_run_sec1_time_split1": null,
"2nd_last_run_sec1_time_split2": null,
"2nd_last_run_sec2_time_total": 22.69,
"2nd_last_run_sec2_time_split1": 11.3,
"2nd_last_run_sec2_time_split2": 11.39,
"2nd_last_run_sec3_time_total": 23.67,
"2nd_last_run_sec3_time_split1": 11.38,
"2nd_last_run_sec3_time_split2": 12.29,
"2nd_last_run_final_200m_fatigue": 0.91,
"2nd_last_run_fatigue_profile": "Collapsing"
}

🧠 Understanding Pre-Race Predictive Metrics

The Pre-Race Analysis mode is designed to give you an edge by automatically fetching and calculating historical sectional times for up to 7(sometimes less if the horse has not run that many times) of a horse's most recent runs. This allows you to evaluate how a horse ran its past races, rather than just where it finished.

Here are the most powerful fields to look out for:

1. Sectional Time Ranks (secX_time_rank)

Instead of manually comparing raw sectional times, the scraper calculates how fast a horse ran a specific section relative to the other horses in that exact same race.

  • Example: "last_run_sec3_time_rank": 1 means the horse ran the fastest final section of anyone in that race, regardless of its overall finishing position.

2. Position Change & Closing Profiles

This metric measures how many horses a runner passed (or was passed by) in the critical final 400m stretch: Position at 400m - Position at Finish.

closing_profileValue Range (position_change)Description
Strong Closer> +4Passed 5 or more horses in the final 400m. A massive engine late in the race.
Closer+2 to +4Made up solid ground in the straight.
Even-1 to +1Held its position to the line.
Faded-4 to -2Lost ground and was passed by a few horses.
Quitter< -4Dropped 5 or more positions in the final 400m. Hit a wall.

3. Momentum Indicator: Final 200m Fatigue

In Hong Kong, the final 400m is split into two 200m halves (split1 and split2). The Fatigue metric calculates the difference between these halves: Split 2 - Split 1. It is a true physics-based indicator of deceleration.

fatigue_profileValue (final_200m_fatigue)Description
Sprinting< 0.0sThe horse accelerated! The final 200m was run faster than the previous 200m. Elite finish.
Sustained0.0s to 0.3sMaintained momentum perfectly through the line with minimal deceleration.
Tiring0.3s to 0.7sStandard deceleration. The horse was working hard but starting to slow.
Collapsing> 0.7sThe horse completely emptied out and hit a physical wall before the finish line.

Pro Tip: Combine these metrics to find hidden gems! A horse might finish 8th and look bad on traditional form guides, but if it had a position_change of +6 (Strong Closer) and a fatigue_profile of "Sprinting", it was flying home late and might be a massive value in its next start!

🏁RESULT MODE

Example Input

  1. The "Historical Harvest" (Bulk Mode) Best for users building a new database who want to scrape every race result currently listed on the HKJC website in one go.
{
"mode": "result",
"all_result_dates": true
}
  1. The "Targeted Handicapper" (Specific Races) Ideal for analyzing specific races on a given day (e.g., just the Feature Race or the first leg of a Triple Trio).
{
"mode": "result",
"result_dates": ["2026-03-15"],
"race_numbers": [1, 5]
}
  1. The "Full Meeting" Scrape The standard way to get every result from a specific race day. By omitting race_numbers, the scraper automatically processes all races for that date.
{
"mode": "result",
"result_dates": ["2026-03-15", "2026-03-18"]
}

Example Output

The output would be a list of dictionaries, such as the following:

{
"race_number": 2,
"season_race": 516,
"venue": "SHA TIN",
"date": "2026-03-15",
"race_title": "SOUTH WALL HANDICAP",
"going": "GOOD TO FIRM",
"track_type": "TURF",
"track_type_status": "C+3 Course",
"class": "Class 5",
"distance": 1200,
"max_rating": 40,
"min_rating": 0,
"prize": 875000,
"race_time": [
23.96,
46.47,
69.22
],
"race_sectional_time": [
{
"main": 23.96,
"subs": []
},
{
"main": 22.51,
"subs": [
11.13,
11.38
]
},
{
"main": 22.75,
"subs": [
11.21,
11.54
]
}
],
"place": 9,
"status": "FINISHED",
"horse_number": 11,
"horse_name": "VERBIER",
"horse_id": "J187",
"horse_url": "https://racing.hkjc.com/en-us/local/information/horse?horseid=HK_2023_J187",
"gear": [
{
"equipment": "BLINKERS",
"action": "RETAINED"
},
{
"equipment": "TONGUE_TIE",
"action": "RETAINED"
}
],
"jockey_name": "L Ferraris",
"jockey_id": "FEL",
"jockey_url": "https://racing.hkjc.com/en-us/local/information/jockeyprofile?jockeyid=FEL&Season=Current",
"trainer_name": "C Fownes",
"trainer_id": "FC",
"trainer_url": "https://racing.hkjc.com/en-us/local/information/trainerprofile?trainerid=FC&Season=Current",
"actual_weight": 125,
"declared_horse_weight": 1261,
"draw": 8,
"length_behind_winner_raw": "5-3/4",
"length_behind_winner": 5.75,
"running_position": [
11,
12,
9
],
"finish_time_raw": "1:10.14",
"finish_time": 70.14,
"win_odds": 15.0,
"sectional_time": [
{
"position": 11,
"margin_raw": "6-1/2",
"margin": 6.5,
"time": 25.0,
"sub_splits": []
},
{
"position": 12,
"margin_raw": "5-3/4",
"margin": 5.75,
"time": 22.39,
"sub_splits": [
10.89,
11.5
]
},
{
"position": 9,
"margin_raw": "5-3/4",
"margin": 5.75,
"time": 22.75,
"sub_splits": [
11.13,
11.62
]
}
],
"comment": "Waited with towards rear, dropped to last place top of home straight, no impression."
}

📋REPORT MODE

Example Input

  1. The "Historical Archive" (Bulk Reports) Automatically finds and extracts every Steward's Incident Report currently available in the HKJC dropdown archive. This is ideal for building a comprehensive database of racing interference and track incidents.
{
"mode": "report",
"all_report_dates": true
}
  1. Targeted Meeting Reports Scrapes official incident reports for a specific date (or set of dates). Use this to analyze the steward's findings immediately after a race meeting has concluded.
{
"mode": "report",
"report_dates": ["2026-03-15", "2026-03-18"]
}

Example Output

The output would be a dictionary, such as the following:

{
"date": "2026-03-15",
"start_time": "12:30",
"venue": "SHA TIN",
"track_type": "MIXED",
"track_type_status": "AWT / TURF C+3 COURSE",
"going_details": [
{
"status": "GOOD",
"races": [
1,
5,
8
]
},
{
"status": "GOOD TO FIRM",
"races": [
2,
3,
4,
6,
7,
9,
10,
11
]
}
],
"penetrometer": [
{
"value": 2.71,
"time": "08:00 am"
},
{
"value": 2.71,
"time": "11:30 am"
}
],
"clegg_hammer": [
{
"value": 9.06,
"time": "08:00 am"
},
{
"value": 9.08,
"time": "11:30 am"
}
],
"stewards": [
{
"name": "Dr Henry H L Chan",
"position": "Chairman"
},
{
"name": "Mr Benedict Sin",
"position": "Steward"
},
{
"name": "Mr M Van Gestel",
"position": "Chief Stipendiary Steward"
},
{
"name": "Mr T Bailey",
"position": "Stipendiary Steward"
},
{
"name": "Mr T Vassallo",
"position": "Stipendiary Steward"
},
{
"name": "Mr K C Y Kwok",
"position": "Stipendiary Steward"
},
{
"name": "Mr J C H Ho",
"position": "Stipendiary Steward"
}
],
"race_number": 2,
"season_race": 516,
"race_title": "SOUTH WALL HANDICAP",
"section": 1,
"race_class": "Class 5",
"distance": 1200,
"place": 11,
"dead_heat": false,
"horse_number": 1,
"horse": {
"name": "WINNING CIGAR",
"id": "K422",
"url": "https://racing.hkjc.com/en-us/local/information/horse?horseid=HK_2024_K422"
},
"drawn": 7,
"jockey": {
"name": "H Bentley",
"id": "BHW",
"url": "https://racing.hkjc.com/en-us/local/information/jockeyprofile?jockeyid=BHW&Season=Current",
"allowance": 0
},
"incident": "Jumped only fairly. Between the 300 Metres and the 200 Metres had difficulty obtaining clear running."
}

🏇 BARRIER TRIAL MODE

Example Input

  1. Bulk Trial Discovery Fetches every available morning barrier trial result from the archive. Essential for tracking long-term horse preparation patterns and fitness cycles.
{
"mode": "barrier_trial",
"all_barrier_dates": true
}
  1. Recent Form Check Scrapes results for the most recent morning trials. This helps in identifying horses that showed strong speed or "finished strongly" in non-betting trials before their next race.
{
"mode": "barrier_trial",
"barrier_dates": ["2026-04-02"]
}

Example Output

The output would be a dictionary, such as the following:

{
"batch_number": 2,
"venue": "CONGHUA",
"track_type": "ALL WEATHER TRACK",
"distance": 1200,
"going": "GOOD",
"overall_time": 70.88,
"sectional_times": [
24.9,
22.7,
23.2
],
"video_url": "https://racing.hkjc.com/contentAsset/videoplayer_v4/video-player-iframe_v4.html?type=brts&date=20260402&rc=ch&no=02&lang=eng&rf=http://racing.hkjc.com/en-us/local/information/btresult?Date=2026/04/02&pageid=racing/local&jumpTime=47",
"date": "2026-04-02",
"horse_name": "BEAUTY CRESCENT",
"horse_id": "H334",
"horse_url": "https://racing.hkjc.com/en-us/local/information/horse?horseid=HK_2022_H334",
"jockey": "M Kellady",
"trainer": "A S Cruz",
"draw": 2,
"gear": [
{
"equipment": "BLINKERS",
"action": "RETAINED"
}
],
"lbw_raw": "8-1/4L",
"lbw": 8.25,
"running_position": [
4,
4,
4
],
"finish_time_raw": "1.12.21",
"finish_time": 72.21,
"result": "",
"comment": "Limited response when asked; not as expected."
}

🏥VETERINARY MODE

Example Input

  1. Targeted Horse Search (By Brand Number) The fastest way to get the full medical history for specific horses. Simply provide the unique HKJC Brand Numbers.
{
"mode": "veterinary",
"brand_numbers": ["K458", "H095"]
}
  1. Full Veterinary Database Export This is the "Power User" mode. It triggers the scraper to crawl the entire HKJC veterinary database to extract health records for every horse currently registered in the system. More than 700 horses with multiple reports.
{
"mode": "veterinary",
"get_full_database": true
}
  1. "Today's Starters" Health Check Automatically identifies all horses declared for the next upcoming race meeting and retrieves their medical history. Perfect for last-minute filtering of runners with recent health issues.
{
"mode": "veterinary",
"get_starters": true
}

Example Output

The output would be a dictionary, such as the following:

{
"horse_name": "SETANTA",
"horse_id": "G095",
"vet_reports": [
{
"record_date": "2021-09-06",
"record_details": "Castration.",
"passing_date": null
},
{
"record_date": "2025-08-11",
"record_details": "Eight years of age or above at season end.",
"passing_date": "2025-09-02"
}
]
}

💰 Pricing: Pay-Per-Event (PPE)

Free Tier Limitation: If you are testing this Actor as a free user on Apify, execution is strictly limited to extracting a maximum of 10 records per run. This allows you to evaluate the data structure and quality for free before committing. Free runs always return the same fixed sample for each mode (for example Results: 2024-04-10, Race 1), whatever dates or filters you enter; your own dates and filters require a paid plan. To unlock unlimited extraction capabilities and bypass this hard limit, you must run the Actor as a paying user.

This Actor uses a transparent Pay-Per-Event pricing model. You only pay for the individual records pushed to the dataset. For Results, Reports, and Trials, data is denormalized so that each horse's performance or incident is a unique, billable record.

  • Price per 1,000 results: $15.00
  • Price per single record (1 Horse): $0.015
  • Odds Movement & Results (premium): $25.00 per 1,000 rows, billed as two events per row: the standard result ($15.00 / 1,000) plus an odds-movement-row add-on ($10.00 / 1,000). One row is one runner at one snapshot, so a full meeting is roughly 1,100 rows (~$27.50).
  • Live Market Movers: $15.00 per 1,000 rows (standard price). Each run returns every runner of the races still to come.

📊 Cost Examples:

Service ModeEstimated RecordsEstimated Cost
Full Race Meeting (Results)~150 horse results~$2.25
Full Meeting Incident Reports~140-150 reports~$2.10
Morning Barrier Trials~50 - 100 horses$0.75 – $1.50
Lifetime Veterinary History1 Horse (Full History)$0.015
Full Meeting Odds Snapshot~1,100 rows (runners × 7 snapshots)~$16.50
Full Meeting Odds Movement & Results~1,100 rows at $25 / 1,000~$27.50
Live Market Movers, every 15 min for a meeting~2,000 rows (shrinks as races finish)~$30

Record Volume Warning: The total number of records (and thus the cost) depends on the specific event. For example, a Barrier Trial morning may have only 50 horses, while a peak season trial day could have over 100 horses. Similarly, race cards vary between 8 and 11 races per meeting.


💡 Pro Tip: Data Export

Because the data is denormalized (flattened), you can export these results directly to Excel or CSV from the Apify platform. Every row is a complete record containing both the horse's specific data and the meeting's metadata (Date, Venue, etc.), making it ready for immediate analysis in Pandas, Excel, or Google Sheets.

🛠 Troubleshooting

  • Date Formats: Ensure dates are YYYY-MM-DD. If the UI shows a validation error, double-check your format.
  • Empty Results: If you search for "Today's Starters" before the HKJC has officially declared the meeting (usually 48 hours before), the result list will be empty.
  • Race Numbers: If you enter [12] for a meeting that only has 9 races, no data will be returned for that specific race number.
  • Odds Modes: Odds Snapshot and Odds Movement only return meetings from 2026-06-27 onwards. In Odds Movement, result fields (finish_position, won, placed) stay empty until results are synced nightly at 00:30 Hong Kong time, and are always empty for standby horses and abandoned races.
  • Live Market Movers: Returns 0 rows on days without a meeting and after the last race has started. Movement fields stay empty until the race's first stored snapshot (3 hours before the race) exists.

✉️ Feedback & Custom Work

I am committed to the ongoing development of this Actor. Future updates and feature additions (such as Dividend/Payout scraping or Weather analysis) will be prioritized based on user reviews, demand, and specific requests.

  • Support: I actively monitor the performance of this tool to ensure it adapts to HKJC website updates. If you find any data inconsistencies or parsing errors, please open an issue in the Actor's "Issues" tab.
  • Reviews: If this data powers your winning model or saves you hours of manual work, please leave a 5-star review! Your feedback directly influences the development roadmap.
  • Customization: Need a specialized data extractor, a custom database integration, or a unique "Actor-as-a-Product" solution? I am available for custom Actor development.

Contact me: bd.pascari@gmail.com