Age boundary: 8 and 9 year olds in marathon handicap chases sit near par

Reasoning

Why this hypothesis exists: what the lab saw, the question it asked, the mechanism it proposed, what it predicted, and the result that would prove it wrong. Written before the test was run.

1. Observed
   • H-9cf148a1 (Older horses in marathon handicap chases are over-bet (mirror of the y) pointed the predicted way but failed the bar: A/E 0.97 on 148 bets (p 0.581), control 1.03. O-0ba01528
   • H-98848406 (Young marathon chasers stepping up to 3m4f+ for the first time carry t) met the sealed vault supportive: A/E 1.22 on 384 bets (p 0.0945, threshold 0.0025). O-bef6da89
2. AskedWhere does the lift stop with age: do 8 and 9 year olds in 3m4f+ handicap chases also beat their price, or is the lift limited to age seven or under? answered
3. MechanismIf the market discounts young horses because marathon trips are believed to need a hardened horse, the discount should fade by 8 and 9, the age the market sees as mature. Those runners should then be priced about right.
4. Why the market misses itThe 'needs an old horse' belief is a line about age that everyone can see. It should affect only the runners it marks out as young, not those of middle age.
5. Predictednegative, A/E 0.92 to 1.02 on 500+ bets, clearly below the young runners' 1.18
6. Falsified ifA/E at or above 1.12 on 400+ bets, which would mean the lift reaches every horse under ten and points to old horses being over-bet rather than young ones being under-bet
7. Rule sketchrace_type = 'Chase' AND is_handicap = 1 AND age >= 8 AND age <= 9 AND distance_yards >= 6160

The question's answer so far: Mostly limited to 7 or under: 8 and 9 year olds scored 0.97 on 229 (CI 0.56 to 1.55) against 1.18 for the young band. The sample is too thin to fix the boundary exactly.

Mechanism and rationale

If the market discounts young horses because marathon trips are believed to need a hardened horse, the discount should fade by 8 and 9, the age the market sees as mature. Those runners should then be priced about right.

The rule

race_type = 'Chase' AND is_handicap = 1 AND age >= 8 AND age <= 9 AND distance_yards >= 6160

Reads: race_type, is_handicap, age, distance_yards

Pre-registration

Spec hash (sha256 of the canonical pre-spec)

35a6a6804b9f6301095e32b2e8e20c3a52fc43b49c990090e3f0f465b2a5e314

Frozen

not frozen

The definition can still change.

Confounders

• Field size: big festival marathons (Grand National, Scottish/Welsh/Irish Nationals) are dominated by 8–10 year olds and have unusual price structures
• Festival marathons: a handful of high-profile races may drive the group's A/E on their own
• Trainer mix: marathon specialist yards may favour one age band
• Price band composition: 8–9 year olds may be concentrated at bigger prices where the longshot bias depresses A/E regardless of age
• Weight: middle-aged horses tend to carry higher marks and weights than young progressive horses, which could produce the gap apart from age
• Period decay: the young-stayer lift weakened in the latest period (1.05 on 125), so a narrowing gap may reflect time, not age

Control

{
  "description": "The same-race comparison is the other runners in these marathon handicap chases. If the mechanism holds, the young runners (7 or under) in those races should keep their lift of about 1.15 to 1.20, the 10+ runners should sit near par, and the 8–9 group should come out near or below 1.0. If instead the whole race field outperforms its prices, the effect is a property of marathon handicap chases (e.g. overround or longshot structure) and not of age. The control rule gives the young group in the same population for a direct like-for-like comparison of A/E.",
  "rule": "race_type = 'Chase' AND is_handicap = 1 AND age <= 7 AND distance_yards >= 6160"
}

Discovery result

fail broad rule · 16 Oct, 13:58

Bets

229

17 winners

A/E

0.97

95%: 0.56 to 1.55

SP return

-32.0%

level stakes

Betfair SP return

-8.9%

229 bets with a BSP

Strike rate 7.4%. p = 0.59. Same-race control: the other runners in these races had A/E 1.05, so the selection did not beat its own races. Zone: discovery. Evaluator eval-1.0, data 6a4e98be18e83933

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

• designer v1ok · 32s · 16 Oct, 13:57

  Output (schema-checked)

  {
    "rule": "race_type = 'Chase' AND is_handicap = 1 AND age >= 8 AND age <= 9 AND distance_yards >= 6160",
    "refinements": [
      {
        "label": "8-year-olds only (nearest the young boundary)",
        "rule": "race_type = 'Chase' AND is_handicap = 1 AND age = 8 AND distance_yards >= 6160"
      },
      {
        "label": "Fields of 13 or fewer (strips big festival marathons)",
        "rule": "race_type = 'Chase' AND is_handicap = 1 AND age >= 8 AND age <= 9 AND distance_yards >= 6160 AND field_size <= 13"
      }
    ],
    "control": {
      "description": "The same-race comparison is the other runners in these marathon handicap chases. If the mechanism holds, the young runners (7 or under) in those races should keep their lift of about 1.15 to 1.20, the 10+ runners should sit near par, and the 8–9 group should come out near or below 1.0. If instead the whole race field outperforms its prices, the effect is a property of marathon handicap chases (e.g. overround or longshot structure) and not of age. The control rule gives the young group in the same population for a direct like-for-like comparison of A/E.",
      "rule": "race_type = 'Chase' AND is_handicap = 1 AND age <= 7 AND distance_yards >= 6160"
    },
    "pass_criteria": {
      "min_bets": 400,
      "min_ae": 1.05,
      "max_p": 0.05
    },
    "forward_bar": {
      "min_bets": 100,
      "min_ae": 1,
      "max_p": 0.2
    },
    "confounders": [
      "Field size: big festival marathons (Grand National, Scottish/Welsh/Irish Nationals) are dominated by 8–10 year olds and have unusual price structures",
      "Festival marathons: a handful of high-profile races may drive the group's A/E on their own",
      "Trainer mix: marathon specialist yards may favour one age band",
      "Price band composition: 8–9 year olds may be concentrated at bigger prices where the longshot bias depresses A/E regardless of age",
      "Weight: middle-aged horses tend to carry higher marks and weights than young progressive horses, which could produce the gap apart from age",
      "Period decay: the young-stayer lift weakened in the latest period (1.05 on 125), so a narrowing gap may reflect time, not age"
    ],
    "mechanism": "If the market discounts young horses in marathon chases because 3m4f+ is believed to need a hardened, mature horse, the discount is tied to an age line. By 8 and 9 the market sees a horse as mature, so the discount should be gone and these runners should be priced about right (A/E near 1.0), clearly below the young group's ~1.18. This tests whether the live young-stayer lift is specific to age 7 and under or applies to every horse that is not old.",
    "market_miss": "The 'marathons need an old horse' belief draws a line about age that everyone can see. It should misprice only the runners it marks out as young, and should leave middle-aged 8–9 year olds, whom the market already treats as mature stayers, correctly priced.",
    "prediction": {
      "direction": "negative",
      "size": "A/E about 0.92 to 1.02 on 500+ bets, clearly below the young runners' ~1.18 in the same races",
      "falsified_if": "A/E at or above 1.12 on 400+ bets. That would mean the lift reaches every horse under ten, which points to old horses being over-bet rather than young ones being under-bet."
    },
    "notes": "This is a boundary test of the live young-stayer system, not a search for a new edge. Direction is pre-registered as negative (near par or over-bet), so a positive pass is not expected. The informative outcome is the gap against the young control. Every column used (race_type, is_handicap, age, distance_yards, field_size) is pre-off and available from July 2020 in GB and IRE. Refinements are only to be read if the broad rule shows evidence. The age-8 refinement shows whether the discount fades gradually. The field-size refinement checks the festival-marathon confounder in the same way as H-8b0e3766. No price-band slicing, per the earlier method finding."
  }

Timeline

1. Job 1724 finished

   16 Oct, 13:58 · engine:compute-1

2. H-a4349569 Discovery → KILLED fails: 229 bets, needs 400; A/E 0.967, needs 0.95 or lower; p 0.5078 (lower tail), needs 0.05; periods [0.97], each needs 0.99 or lower

   16 Oct, 13:58 · engine:evaluator

3. discovery result for H-a4349569 (broad rule): fail

   16 Oct, 13:58 · engine:evaluator

4. H-a4349569 Designed → RESEARCHdiscovery measuring on the discovery zone

   16 Oct, 13:58 · engine:orchestrator

5. compute-1 took job 1724 (evaluate, attempt 1)

   16 Oct, 13:58 · compute-1

6. Job 1722 finished

   16 Oct, 13:57 · engine:ai-1

7. Queued evaluate on the compute lane: discovery run of the broad rule

   16 Oct, 13:57 · engine:scheduler

8. H-a4349569 Proposed → RESEARCHdesigned pre-spec written and hashed

   16 Oct, 13:57 · agent:designer@v1

9. ai-1 took job 1722 (design, attempt 1)

   16 Oct, 13:57 · ai-1

10. Queued design on the ai lane: write the pre-spec

    16 Oct, 13:56 · engine:scheduler

11. New idea H-a4349569: Age boundary: 8 and 9 year olds in marathon handicap chases sit near par (test-mechanism)

    16 Oct, 13:56 · agent:investigator@v1

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