H228: dual purpose horse in an irish summer flat handicap

Mechanism and rationale

Ireland, June to September, a Flat handicap over 12f or more: a horse with three or more hurdle runs. The summer festivals are built round dual-purpose horses whose Flat marks lag behind what they have done over hurdles.

The rule

region = 'IRE' AND race_type = 'Flat' AND is_handicap = 1 AND month BETWEEN 6 AND 9 AND distance_yards >= 2640 AND hurdle_runs >= 3

Reads: region, race_type, is_handicap, month, distance_yards, hurdle_runs

Pre-registration

Spec hash (sha256 of the canonical pre-spec)

not designed yet

Frozen

not frozen

The definition can still change.

Broad

region = 'IRE' AND race_type = 'Flat' AND is_handicap = 1 AND month BETWEEN 6 AND 9 AND distance_yards >= 2640 AND hurdle_runs >= 3

Kind

all

Origin

research/experiments/overnight/results_v2.jsonl

Refinements

[
  {
    "name": "Galway, Listowel, Killarney or Ballinrobe",
    "rule": "(region = 'IRE' AND race_type = 'Flat' AND is_handicap = 1 AND month BETWEEN 6 AND 9 AND distance_yards >= 2640 AND hurdle_runs >= 3) AND (course IN ('GALWAY', 'LISTOWEL', 'KILLARNEY', 'BALLINROBE'))"
  },
  {
    "name": "14/1 or shorter",
    "rule": "(region = 'IRE' AND race_type = 'Flat' AND is_handicap = 1 AND month BETWEEN 6 AND 9 AND distance_yards >= 2640 AND hurdle_runs >= 3) AND (sp_decimal <= 15)"
  },
  {
    "name": "ran over hurdles last time",
    "rule": "(region = 'IRE' AND race_type = 'Flat' AND is_handicap = 1 AND month BETWEEN 6 AND 9 AND distance_yards >= 2640 AND hurdle_runs >= 3) AND (last_race_type = 'Hurdle')"
  }
]

Tier

A

Result (from the archive)

info ran over hurdles last time · 15 Oct, 23:10

Bets

744

A/E

0.99

winners vs the prices

SP return

-41.5%

level stakes

Betfair SP return

-40.8%

112 bets with a BSP

Strike rate 7.4%.

Falsification

No attacks yet. Only promising ideas are attacked.

Sealed holdout (vault)

Not queried. Only a frozen idea can spend its one query against the 2020-25 holdout.

Forward record

No forward record. Shadow tracking starts after the vault and counts only races run after the freeze.

Lineage

Agent runs

• investigator v1ok · 47s · 17 Oct, 03:15

  Output (schema-checked)

  {
    "reflection": "The thread rests on one broad idea that has not held up, and one slice of it that was then published. The broad rule (Irish June–September Flat handicaps of 12f or more, horse with 3+ hurdle runs) was GREEN in earlier research at A/E 1.05 on 1890 bets, with periods 1.03, 1.08 and 0.99. On the lab's own discovery window it came out at 0.986 on 744, so the broad rule is RED. The published system keeps only four festival courses (Galway, Listowel, Killarney, Ballinrobe), which scored 1.2 on 544. That course list was picked after seeing the numbers on a failed parent, so it is a lead, not a confirmed mechanism. Two warning signs: across the import sample, BSP ROI was +40.7% but SP ROI was -23.7%, which suggests the profit comes from a few big-priced winners; and the latest period fell back to 0.99. One plumbing point: the system rule uses `prior_hurdle_runs`, the parent used `hurdle_runs`, and the system is 'not measured', so the column needs checking before going live. What we believe now: the mark-lag story is plausible but unproven. If it is real, it should show up where the Flat mark knows least, and it should survive outside long prices and outside the four named courses.",
    "answers": [],
    "questions": [
      {
        "question": "Is the festival-course lift (A/E 1.2 on 544) carried by long-priced winners rather than by the field generally?",
        "rationale": "BSP ROI +40.7% against SP ROI -23.7% points to a few big BSP winners. Split the published system by `sp_decimal` bands (<6, 6-10, 10-20, 20+) and `sp_rank`. If the lift sits only at 16/1 or bigger, it is a variance artefact, not a mark-lag edge, and the live system is fragile.",
        "observation_ids": [
          "O-a317bdab"
        ],
        "priority": 0.75
      },
      {
        "question": "Is the lift strongest where the Flat mark has the least information, i.e. horses lightly raced on the Flat but proven over hurdles?",
        "rationale": "The mechanism says the Flat mark lags behind hurdle form. That lag should be biggest when `flat_runs` is low and `hurdle_wins` >= 1, at any Irish summer course. If the whole broad population is flat, this is the one cut the mechanism itself names. Columns: `flat_runs`, `flat_wins`, `hurdle_wins`, `hurdle_runs`, `last_official_rating`.",
        "observation_ids": [
          "O-a317bdab"
        ],
        "priority": 0.65
      },
      {
        "question": "Does the published system's column `prior_hurdle_runs` exist and match `hurdle_runs`, given the system shows 'not measured'?",
        "rationale": "The parent rule used `hurdle_runs`; the system rule uses `prior_hurdle_runs`. If that column is missing or differently defined, the live system selects nothing or something else. Check in `query_catalog`.",
        "observation_ids": [
          "O-a317bdab"
        ],
        "priority": 0.5
      },
      {
        "question": "Is the festival effect really a trainer effect (a few yards aiming dual-purpose horses at these meetings) or a large-field effect?",
        "rationale": "Check how concentrated the 544 festival bets are by `trainer`, and by `field_size`. If two yards carry it, the edge is about where those yards aim their horses, not about the Flat mark lagging.",
        "observation_ids": [
          "O-a317bdab"
        ],
        "priority": 0.45
      }
    ],
    "hypotheses": [
      {
        "title": "Sharpen: proven hurdlers lightly raced on the Flat in Irish summer staying handicaps",
        "family": "season-ireland",
        "operator": "sharpen",
        "question_id": null,
        "new_question": 1,
        "parent_id": "H228-dual-purpose-horse-in-an-irish-summer-flat-handicap",
        "reasoning": {
          "observation_ids": [
            "O-a317bdab"
          ],
          "question": "Is the lift strongest where the Flat mark has the least information about a proven hurdler?",
          "mechanism": "A Flat handicap mark is set from a few Flat runs, often run before the horse matured over hurdles. A horse that has won over hurdles but has few Flat runs carries a mark that has not caught up with what it can do now, which is exactly the lag the thread claims.",
          "market_miss": "The Flat form book shows modest Flat figures. Hurdle wins sit in a different code, and Flat punters and compilers weight them lightly.",
          "prediction": {
            "direction": "positive",
            "size": "A/E 1.08 to 1.15 on 250+ bets, across all Irish courses",
            "falsified_if": "A/E at or below 1.00 on 200+ bets, or a lift only at 16/1 or bigger"
          },
          "rule_sketch": "region = 'IRE' AND race_type = 'Flat' AND is_handicap = 1 AND month BETWEEN 6 AND 9 AND distance_yards >= 2640 AND hurdle_wins >= 1 AND flat_runs <= 6"
        },
        "confounders": [
          "long-priced winner variance",
          "big yards targeting festivals",
          "course concentration at Galway"
        ],
        "expected_survival": 0.3
      },
      {
        "title": "Rival: the festival-course lift is a long-price artefact",
        "family": "season-ireland",
        "operator": "rival-explanation",
        "question_id": null,
        "new_question": 0,
        "parent_id": "SYS-dual-purpose-irish-summer-festivals",
        "reasoning": {
          "observation_ids": [
            "O-a317bdab"
          ],
          "question": "Does the festival-course lift survive when long prices are removed?",
          "mechanism": "If the Flat mark really lags behind hurdle form, the advantage should show among the horses the market fancies, not only among a few big-priced winners. The +40.7% BSP ROI against -23.7% SP ROI points to the rival: big BSP winners.",
          "market_miss": "If the mechanism is real, the market under-rates the whole profile, so shorter-priced runners should also beat their prices.",
          "prediction": {
            "direction": "positive",
            "size": "A/E 1.05 to 1.12 on 300+ bets at under 10/1",
            "falsified_if": "A/E below 1.00 at under 10/1 on 250+ bets, which would mean the system's lift is price-band variance"
          },
          "rule_sketch": "region = 'IRE' AND race_type = 'Flat' AND is_handicap = 1 AND month BETWEEN 6 AND 9 AND distance_yards >= 2640 AND hurdle_runs >= 3 AND course IN ('GALWAY', 'LISTOWEL', 'KILLARNEY', 'BALLINROBE') AND sp_decimal < 11"
        },
        "confounders": [
          "small sample after the price cut",
          "Galway week dominating the sample"
        ],
        "expected_survival": 0.35
      },
      {
        "title": "Rival: the festival lift is a large-field effect, not a course effect",
        "family": "season-ireland",
        "operator": "rival-explanation",
        "question_id": null,
        "new_question": 3,
        "parent_id": "H228-dual-purpose-horse-in-an-irish-summer-flat-handicap",
        "reasoning": {
          "observation_ids": [
            "O-a317bdab"
          ],
          "question": "Do the four named courses stand in for big, competitive handicap fields where the market prices outsiders loosely?",
          "mechanism": "Festival handicaps draw big fields. In big fields the overround spreads thinly over many runners, and dual-purpose horses ridden for a late run may beat their prices for that reason, wherever the race is run.",
          "market_miss": "Prices in big fields get compressed toward the favourites, and outsiders with hurdle stamina are lumped in with no-hopers.",
          "prediction": {
            "direction": "positive",
            "size": "A/E 1.06 to 1.12 on 300+ bets at any Irish course with 14+ runners",
            "falsified_if": "A/E at or below 1.00 on 250+ bets, which would point to something specific to the courses or meetings rather than field size"
          },
          "rule_sketch": "region = 'IRE' AND race_type = 'Flat' AND is_handicap = 1 AND month BETWEEN 6 AND 9 AND distance_yards >= 2640 AND hurdle_runs >= 3 AND field_size >= 14"
        },
        "confounders": [
          "festival courses make up most large-field races",
          "trainer concentration"
        ],
        "expected_survival": 0.25
      }
    ],
    "findings": [
      {
        "statement": "The broad Irish summer dual-purpose rule did not hold on the lab's window (A/E 0.986 on 744 against 1.05 on 1890 earlier). The live system rests on a four-course slice (1.2 on 544) chosen after seeing the numbers on a failed parent.",
        "kind": "fails",
        "observation_ids": [
          "O-a317bdab"
        ]
      },
      {
        "statement": "The imported sample's BSP ROI (+40.7%) and SP ROI (-23.7%) are far apart, so the profit is probably concentrated in big BSP winners. The most recent period fell back to 0.99.",
        "kind": "method",
        "observation_ids": [
          "O-a317bdab"
        ]
      },
      {
        "statement": "The published system rule uses `prior_hurdle_runs` while the parent used `hurdle_runs`, and the system is unmeasured. Check that column before trusting live selections.",
        "kind": "method",
        "observation_ids": [
          "O-a317bdab"
        ]
      }
    ]
  }

Timeline

1. hypothesis note H228-dual-purpose-horse-in-an-irish-summer-flat-handicap

   15 Oct, 23:10 · importer

2. H228-dual-purpose-horse-in-an-irish-summer-flat-handicap Discovery → PROMISING import: tier A: strong candidate; best [Galway, Listowel, Killarney or Ballinrobe] GREEN A/E 1.2 on 544 bets (broad RED)

   15 Oct, 23:10 · importer

3. import result for H228-dual-purpose-horse-in-an-irish-summer-flat-handicap (ran over hurdles last time): info

   15 Oct, 23:10 · importer

4. import result for H228-dual-purpose-horse-in-an-irish-summer-flat-handicap (14/1 or shorter): info

   15 Oct, 23:10 · importer

5. import result for H228-dual-purpose-horse-in-an-irish-summer-flat-handicap (Galway, Listowel, Killarney or Ballinrobe): info

   15 Oct, 23:10 · importer

6. import result for H228-dual-purpose-horse-in-an-irish-summer-flat-handicap (broad): info

   15 Oct, 23:10 · importer

7. New idea H228-dual-purpose-horse-in-an-irish-summer-flat-handicap: H228: dual purpose horse in an irish summer flat handicap (import)

   15 Oct, 23:10 · importer

Every change is an event with an actor: an agent and its version, an engine, a human, or the importer. Current stage: Promising.