• Hi Guest Just in case you were not aware I wanted to highlight that you can now get a free 7 day trial of Horseracebase here.
    We have a lot of members who are existing users of Horseracebase so help is always available if needed, as well as dedicated section of the fourm here.
    Best Wishes
    AR

Introducing AI Deep Dive: Horse Racing Podcasts!

▸ SOUTHWELL THURSDAY 21 MAY 2026 — V15 EARLY DOORS | LIVE STRUCTURE
Southwell built through the V15 Early Doors audit lens, with AU hierarchy kept ahead of market noise.
Structure-first card using uploaded layers only, not tipping language.
• AU integrity checked race-by-race across all eight races
• Smart Stats folded in for hot/cold handlers, headgear, class drops, stable switches, and weighted-to-win flags
• Market layer used for compression, support, and caution only
• Winner-first structure retained before forecast and TOTE construction
• Caution markers logged where market weakness, headgear, stable switch, or class-drop volatility appeared
• Forecast combos built outward from the AU anchor, not from market rank alone
• TOTE anchors bound to the same V15 Win Pick in each race
• Charter discipline enforced: model structure, not result prediction
Read the full card... Southwell Thursday 21 May 2026 – V15 Overlay Blog | Tactical Forecasts + Smart Stats Integration | Horse Racing Hobbyist

📌 FINAL SUMMARY SECTION

🔵 Top Win Picks
• Race 1: Sisterandbrother
• Race 2: Shalaa Asker
• Race 3: King Of Thebes
• Race 4: Royal Fanfare
• Race 5: Crown Ranger
• Race 6: Adelaide Bay
• Race 7: South Kensington
• Race 8: Forever Noah

🟡 Forecast Combos
• Race 1: Sisterandbrother → Arth's Gold / He's Our Cracker
• Race 2: Shalaa Asker → Westgate Warrior / Fantasy Obsessor
• Race 3: King Of Thebes → Lone Warrior / Eljowhary
• Race 4: Royal Fanfare → Dream Camp / Lightning Glory
• Race 5: Crown Ranger → Choreography / Parioca
• Race 6: Adelaide Bay → Fort Augustus / Just King High
• Race 7: South Kensington → Megaphone / No News
• Race 8: Forever Noah → Bomb Squad / Daytona Lady

🟢 EW/Combo Value Inclusions
• Arth's Gold
• He's Our Cracker
• Westgate Warrior
• Fantasy Obsessor
• Lone Warrior
• Eljowhary
• Dream Camp
• Lightning Glory
• Choreography
• Parioca
• Fort Augustus
• Just King High
• Megaphone
• No News
• Bomb Squad
• Daytona Lady

🎲 TOTE Combos Recap
• Race 1: Sisterandbrother + Arth's Gold / He's Our Cracker
• Race 2: Shalaa Asker + Westgate Warrior / Fantasy Obsessor
• Race 3: King Of Thebes + Lone Warrior / Eljowhary
• Race 4: Royal Fanfare + Dream Camp / Lightning Glory
• Race 5: Crown Ranger + Choreography / Parioca
• Race 6: Adelaide Bay + Fort Augustus / Just King High
• Race 7: South Kensington + Megaphone / No News
• Race 8: Forever Noah + Bomb Squad / Daytona Lady

⚠️ Caution Marker List
• He's Our Cracker – first-time hood evidenced from uploaded layers
• Shalaa Asker – market weakness versus AU evidenced from uploaded market layer
• King Of Thebes – market weakness versus AU evidenced from uploaded market layer
• Topathemorning – class-drop volatility evidenced from uploaded Smart Stats layer
• Parioca – first-time cheekpieces evidenced from uploaded Smart Stats layer
• Adelaide Bay – stable switch evidenced from uploaded Smart Stats layer
• Megaphone – market weakness versus AU evidenced from uploaded market layer
• Forever Noah – market weakness versus AU evidenced from uploaded market layer

🔒 Charter Reminder:

This is a tactical overlay system, not a tipping blog.
No simulation.

VALIDATION & TRUST LAYER

AU integrity

• Race 1: AU integrity evidenced — Sisterandbrother led uploaded points totals with 13pts.
• Race 2: AU integrity evidenced — Shalaa Asker led uploaded points totals with 11pts.
• Race 3: AU integrity evidenced — King Of Thebes led uploaded points totals with 11pts.
• Race 4: AU integrity evidenced — Royal Fanfare led uploaded points totals with 16pts.
• Race 5: AU integrity evidenced — Crown Ranger, Parioca, Choreography, and Sea Idol tied on 3pts; Crown Ranger retained by Rated to Win support and market compression.
• Race 6: AU integrity evidenced — Adelaide Bay led uploaded points totals with 11pts.
• Race 7: AU integrity evidenced — South Kensington led uploaded points totals with 14pts.
• Race 8: AU integrity evidenced — Forever Noah led uploaded points totals with 11pts. : SEE THE BLOG! ;)
 
▸ HAYDOCK THURSDAY 21 MAY 2026 — V15 EARLY DOORS | LIVE STRUCTURE

A clean audit-based Haydock structure built from uploaded AU figs, Smart Stats, racecard layers and market evidence.
This is tactical overlay work only, not tipping language.

• AU integrity checked race-by-race across the uploaded points layers
• Smart Stats support included for hot and cold jockey-trainer handling
• Market data used only for compression and caution, not to override AU alignment
• BF LTO, headgear, class-drop and weighted-to-win flags handled from uploaded evidence only
• Forecast structures built winner-first from the declared V15 anchor
• Caution markers isolated where the uploaded layers supported them
• TOTE Exacta and Trifecta anchors kept bound to the same V15 Win Pick
• Charter discipline maintained: model ≠ result, structure first, no simulation

Read the full card... Haydock Thursday 21 May 2026 – V15 Overlay Blog | Tactical Forecasts + Smart Stats Integration | Horse Racing Hobbyist

📌 FINAL SUMMARY SECTION

🔵 Top Win Picks
• Race 1: DREAMS ADOZEN
• Race 2: FLORIDA DREAMS
• Race 3: SECRET OF LIFE
• Race 4: LION OF MALI
• Race 5: KEEP KICKING ON
• Race 6: STONEACRE JOE
• Race 7: BRUMMELL

🟡 Forecast Combos
• Race 1: DREAMS ADOZEN → TREASURE ISLANDS / FLEURMAN
• Race 2: FLORIDA DREAMS → SCARLET LEGEND / OUTBACK LEGEND
• Race 3: SECRET OF LIFE → POINTE BLEU / PRIMA BAILARINA
• Race 4: LION OF MALI → EUPHONIA / TEKITOFF
• Race 5: KEEP KICKING ON → CHARTED COURSE / REGAL TIGER
• Race 6: STONEACRE JOE → ANGELS' SHARE / THERE'S A CHANCE
• Race 7: BRUMMELL → WHEELS OF FIRE / TRILBY

🟢 EW/Combo Value Inclusions
• TREASURE ISLANDS
• FLEURMAN
• SCARLET LEGEND
• OUTBACK LEGEND
• POINTE BLEU
• PRIMA BAILARINA
• EUPHONIA
• TEKITOFF
• CHARTED COURSE
• REGAL TIGER
• ANGELS' SHARE
• THERE'S A CHANCE
• WHEELS OF FIRE
• TRILBY

🎲 TOTE Combos Recap
• Race 1: DREAMS ADOZEN + TREASURE ISLANDS / FLEURMAN
• Race 2: FLORIDA DREAMS + SCARLET LEGEND / OUTBACK LEGEND
• Race 3: SECRET OF LIFE + POINTE BLEU / PRIMA BAILARINA
• Race 4: LION OF MALI + EUPHONIA / TEKITOFF
• Race 5: KEEP KICKING ON + CHARTED COURSE / REGAL TIGER
• Race 6: STONEACRE JOE + ANGELS' SHARE / THERE'S A CHANCE
• Race 7: BRUMMELL + WHEELS OF FIRE / TRILBY

⚠️ Caution Marker List
• FLEURMAN – Beaten favourite last time out
• FLORIDA DREAMS – Market weakness versus AU
• PRIMA BAILARINA – Beaten favourite last time out
• KEEP KICKING ON – Market weakness versus AU
• STONEACRE JOE – Cold jockey / cold trainer
• BRUMMELL – Cold jockey / cold trainer
• TRILBY – Beaten favourite last time out and first-time cheekpiece

🔒 Charter Reminder:

This is a tactical overlay system, not a tipping blog.
No simulation.

VALIDATION & TRUST LAYER

AU integrity

• Race 1: AU integrity evidenced — DREAMS ADOZEN led uploaded points totals with 12pts.
• Race 2: AU integrity evidenced — FLORIDA DREAMS led uploaded points totals with 12pts.
• Race 3: AU integrity evidenced — SECRET OF LIFE led uploaded points totals with 15pts.
• Race 4: AU integrity evidenced — LION OF MALI led uploaded points totals with 14pts.
• Race 5: AU integrity evidenced — KEEP KICKING ON led uploaded points totals with 9pts.
• Race 6: AU integrity evidenced — STONEACRE JOE led uploaded points totals with 8pts.
• Race 7: AU integrity evidenced — WHEELS OF FIRE and BRUMMELL tied on 10pts; BRUMMELL retained by Rated to Win tie-break support.: SEE THE BLOG! ;)
▸ HAYDOCK THURSDAY 21 MAY 2026 — V15 EARLY DOORS | LIVE STRUCTURE
Post-race critique now added for the Haydock V15 Early Doors card.
Calm audit only: AU structure, Smart Stats, forecast zones, caution handling and TOTE logic reviewed.
• AU figs reviewed race-by-race against the declared V15 structure
• Smart Stats layers checked against jockey, trainer, headgear and weighted-to-win flags
• Forecast zones reviewed through the locked Win Pick, Partner A and Partner B structure
• Caution markers kept separate from outcome language and handled as chaos-control evidence
• TOTE Exacta and Trifecta logic assessed only through the hardened locked rules
• Market compression reviewed as support or caution, not as an override
• Model integrity separated from betting outcome throughout the critique
Read the full card and post-race critique: Haydock Thursday 21 May 2026 – V15 Overlay Blog | Tactical Forecasts + Smart Stats Integration | Horse Racing Hobbyist

📉 Cumulative Outcome Analysis

Structured Yankee:
• Stake: £3.30
• Returns: £0.00
• Outcome: Lost

V15 Win Picks:
• Race 1: Dreams Adozen — unplaced
• Race 2: Florida Dreams — unplaced
• Race 3: Secret Of Life — 1st
• Race 4: Lion Of Mali — unplaced
• Race 5: Keep Kicking On — 3rd
• Race 6: Stoneacre Joe — unplaced
• Race 7: Brummell — unplaced

Win Pick strike:
• 1 winner from 7 races.

Exacta outcomes:
• Race 1: FAILED
• Race 2: FAILED
• Race 3: FAILED
• Race 4: FAILED
• Race 5: FAILED
• Race 6: FAILED
• Race 7: FAILED

Boxed Trifecta outcomes:
• Race 1: FAILED
• Race 2: FAILED
• Race 3: FAILED
• Race 4: FAILED
• Race 5: FAILED
• Race 6: FAILED
• Race 7: FAILED

No TOTE payout is printed because no V15 Exacta or boxed Trifecta met the locked landing conditions.

No TOTE P/L bracket is printed because no relevant V15 TOTE bet was declared landed.

The day produced repeated partner-side evidence without enough Win Pick conversion. That is a model-structure issue, not a payout issue.

🔧 Model Reflection & Refinement Notes

The primary failure was anchor failure.

Race 3 showed the cleanest model hold: Secret Of Life was the AU-led Win Pick and won. The Exacta still failed because Shushu, not Pointe Bleu or Prima Bailarina, finished second.

Race 4, Race 5, Race 6, and Race 7 showed partner-side or secondary-structure strength, but the winner-first layer was not strong enough.

Market weakness versus AU mattered:
• Florida Dreams was market-weak versus AU and failed.
• Keep Kicking On was market-weak versus AU and failed to win.
• Brummell was retained by AU tie-break but failed, while Trilby won from the partner/caution zone.

The useful refinement is not to abandon AU.

The useful refinement is to harden AU against market weakness when the race has live compression against the AU leader.

Carry forward:
• Keep AU as the primary structural driver.
• Keep winner-first discipline.
• Treat AU leaders with clear market weakness as caution anchors, especially where a partner has stronger live compression.
• Do not force strong place-shape runners into the win slot without AU support.
• Do not ignore partner-side evidence when it is repeatedly confirmed by market, Smart Stats, or caution-layer interaction.
• Preserve the distinction between model integrity and betting outcome.

Charter discipline enforced.

No simulation.
No inferred payouts.
No unsupported race-shape commentary.
 
V15 POST-LOCK LEARNING BLOCK Apply only after the MASTER LOCKED PROMPT – Critique & Debrief (HARDENED – DATA-ONLY) is complete. Purpose: Convert the completed critique into carry-forward learning for future V15 Early Doors blog builds.

What Held

Secret Of Life held as a clean Win Pick.

Some partner-side evidence held structurally:
Euphonia won from Partner A.
Outback Legend won from Partner B.
There’s A Chance won from Partner B.
Trilby won from Partner B.
Charted Course and Keep Kicking On filled second and third in Race 5.

What Failed

Anchor failure dominated the card.

Dreams Adozen failed as the Race 1 anchor.
Florida Dreams failed as the Race 2 anchor.
Lion Of Mali failed as the Race 4 anchor.
Keep Kicking On failed as the Race 5 anchor.
Stoneacre Joe failed as the Race 6 anchor.
Brummell failed as the Race 7 anchor.

Partner failure occurred where the correct Win Pick was found but the forecast partners did not support the structure.

Ordering failure occurred where live runners were inside the forecast zone but not in the required Win Pick-led order.

Forecast exposure was clear across the card: partner-side strength repeatedly appeared without anchor conversion.

The Yankee structure was unsuitable for this card because the selected win legs did not match a stable enough winner-first profile.

Build Lesson

AU leaders with clear market weakness must be treated as caution anchors, not clean anchors.

Carry Forward

Retain AU-first hierarchy.
Retain winner-first discipline.
Tighten market weakness versus AU.
Downgrade AU leaders where live market compression strongly favours a partner.
Keep strong place-shape runners in forecast, TOTE, and combo roles unless AU support justifies the win slot.
Monitor tied AU leaders more aggressively before applying Rated to Win tie-breaks.
Avoid Yankee-style escalation when anchor integrity is uneven across the card.

Classification

Anchor Failure

Discipline Rule

Do not promote a market-weak AU leader into a clean anchor without printing and acting on the caution.
 
▸ SOUTHWELL THURSDAY 21 MAY 2026 — V15 EARLY DOORS | LIVE STRUCTURE
Southwell built through the V15 Early Doors audit lens, with AU hierarchy kept ahead of market noise.
Structure-first card using uploaded layers only, not tipping language.
• AU integrity checked race-by-race across all eight races
• Smart Stats folded in for hot/cold handlers, headgear, class drops, stable switches, and weighted-to-win flags
• Market layer used for compression, support, and caution only
• Winner-first structure retained before forecast and TOTE construction
• Caution markers logged where market weakness, headgear, stable switch, or class-drop volatility appeared
• Forecast combos built outward from the AU anchor, not from market rank alone
• TOTE anchors bound to the same V15 Win Pick in each race
• Charter discipline enforced: model structure, not result prediction
Read the full card... Southwell Thursday 21 May 2026 – V15 Overlay Blog | Tactical Forecasts + Smart Stats Integration | Horse Racing Hobbyist

📌 FINAL SUMMARY SECTION

🔵 Top Win Picks
• Race 1: Sisterandbrother
• Race 2: Shalaa Asker
• Race 3: King Of Thebes
• Race 4: Royal Fanfare
• Race 5: Crown Ranger
• Race 6: Adelaide Bay
• Race 7: South Kensington
• Race 8: Forever Noah

🟡 Forecast Combos
• Race 1: Sisterandbrother → Arth's Gold / He's Our Cracker
• Race 2: Shalaa Asker → Westgate Warrior / Fantasy Obsessor
• Race 3: King Of Thebes → Lone Warrior / Eljowhary
• Race 4: Royal Fanfare → Dream Camp / Lightning Glory
• Race 5: Crown Ranger → Choreography / Parioca
• Race 6: Adelaide Bay → Fort Augustus / Just King High
• Race 7: South Kensington → Megaphone / No News
• Race 8: Forever Noah → Bomb Squad / Daytona Lady

🟢 EW/Combo Value Inclusions
• Arth's Gold
• He's Our Cracker
• Westgate Warrior
• Fantasy Obsessor
• Lone Warrior
• Eljowhary
• Dream Camp
• Lightning Glory
• Choreography
• Parioca
• Fort Augustus
• Just King High
• Megaphone
• No News
• Bomb Squad
• Daytona Lady

🎲 TOTE Combos Recap
• Race 1: Sisterandbrother + Arth's Gold / He's Our Cracker
• Race 2: Shalaa Asker + Westgate Warrior / Fantasy Obsessor
• Race 3: King Of Thebes + Lone Warrior / Eljowhary
• Race 4: Royal Fanfare + Dream Camp / Lightning Glory
• Race 5: Crown Ranger + Choreography / Parioca
• Race 6: Adelaide Bay + Fort Augustus / Just King High
• Race 7: South Kensington + Megaphone / No News
• Race 8: Forever Noah + Bomb Squad / Daytona Lady

⚠️ Caution Marker List
• He's Our Cracker – first-time hood evidenced from uploaded layers
• Shalaa Asker – market weakness versus AU evidenced from uploaded market layer
• King Of Thebes – market weakness versus AU evidenced from uploaded market layer
• Topathemorning – class-drop volatility evidenced from uploaded Smart Stats layer
• Parioca – first-time cheekpieces evidenced from uploaded Smart Stats layer
• Adelaide Bay – stable switch evidenced from uploaded Smart Stats layer
• Megaphone – market weakness versus AU evidenced from uploaded market layer
• Forever Noah – market weakness versus AU evidenced from uploaded market layer

🔒 Charter Reminder:

This is a tactical overlay system, not a tipping blog.
No simulation.

VALIDATION & TRUST LAYER

AU integrity

• Race 1: AU integrity evidenced — Sisterandbrother led uploaded points totals with 13pts.
• Race 2: AU integrity evidenced — Shalaa Asker led uploaded points totals with 11pts.
• Race 3: AU integrity evidenced — King Of Thebes led uploaded points totals with 11pts.
• Race 4: AU integrity evidenced — Royal Fanfare led uploaded points totals with 16pts.
• Race 5: AU integrity evidenced — Crown Ranger, Parioca, Choreography, and Sea Idol tied on 3pts; Crown Ranger retained by Rated to Win support and market compression.
• Race 6: AU integrity evidenced — Adelaide Bay led uploaded points totals with 11pts.
• Race 7: AU integrity evidenced — South Kensington led uploaded points totals with 14pts.
• Race 8: AU integrity evidenced — Forever Noah led uploaded points totals with 11pts. : SEE THE BLOG! ;)
▸ SOUTHWELL THURSDAY 21 MAY 2026 — V15 EARLY DOORS | LIVE STRUCTURE

Southwell is now logged through the V15 Early Doors audit path.
Structure-first review, using the card, bet slip, and official results only.

• AU figs checked against the declared Win Pick structure
• Smart Stats reviewed for headgear, class drops, stable switches, and weighted-to-win flags
• Forecast zones assessed without rewriting the pre-race card
• TOTE structure reviewed under exacta and boxed trifecta rules only
• Caution markers retained where market weakness, headgear, or volatility was evidenced
• Betting outcome kept separate from model integrity
• No unsupported payout or P/L bracket added
• Charter discipline preserved through the critique layer

Read the full card and post-race critique: Southwell Thursday 21 May 2026 – V15 Overlay Blog | Tactical Forecasts + Smart Stats Integration | Horse Racing Hobbyist

📉 Cumulative Outcome Analysis

Win Picks:

Race 1 – Sisterandbrother: 2nd
Race 2 – Shalaa Asker: unplaced
Race 3 – King Of Thebes: unplaced
Race 4 – Royal Fanfare: unplaced
Race 5 – Crown Ranger: unplaced
Race 6 – Adelaide Bay: 2nd
Race 7 – South Kensington: 1st
Race 8 – Forever Noah: 2nd

V15 Win Pick strike:
1 winner from 8 races.

Exacta outcomes:
Race 1 – FAILED
Race 2 – FAILED
Race 3 – FAILED
Race 4 – FAILED
Race 5 – FAILED
Race 6 – FAILED
Race 7 – FAILED
Race 8 – FAILED

Boxed Trifecta outcomes:
Race 1 – FAILED
Race 2 – FAILED
Race 3 – FAILED
Race 4 – FAILED
Race 5 – FAILED
Race 6 – FAILED
Race 7 – FAILED
Race 8 – FAILED

No V15 Exacta LANDED.

No V15 Boxed Trifecta LANDED.

No TOTE Exacta payout printed.

No TOTE Trifecta payout printed.

The uploaded official results did include Tote Exacta and Tote Trifecta dividends, but the V15 structures did not meet the required landed conditions. Therefore no TOTE P/L brackets are printed.

The strongest structural positives were Race 7, where the Win Pick won, and Races 1, 4, 6, and 8, where two V15 forecast horses appeared within the first three or four. Those were not enough to trigger any TOTE landed status.

The most significant structural failures were Race 2, Race 3, and Race 5, where the Win Pick failed and the race winner came from outside the selected V15 anchor position. Race 5 specifically exposed tie-break risk where Choreography won but was used only as Partner A.

The human-side betting slip produced a cashed-out return from Cayman Tai, but the full double did not complete because Daytona Lady lost.

🔧 Model Reflection & Refinement Notes

Model integrity and betting outcome must remain separate.

The V15 build followed AU discipline, but AU-led winner selection underperformed on the card.

Market weakness versus AU was correctly flagged in several races, including Shalaa Asker, King Of Thebes, Megaphone, and Forever Noah. The critique shows those warnings were not cosmetic. They should remain hard caution markers, especially where the AU leader is not strongly supported by active market compression.

Race 2 is the clearest refinement point. Cayman Tai was present in the wider AU layer but was not retained in the V15 forecast combo. The human bet slip identified the correct winner. That does not justify hindsight rewriting, but it does support tighter handling when an AU leader is market-weak and a compressed alternative has form, market, and panel presence.

Race 5 shows that low-point ties need stricter tie-break discipline. Crown Ranger was retained over Choreography, but Choreography won. In tied novice or juvenile races, the tie-break must be treated as fragile unless one runner has a decisive named AU panel edge.

Race 6 shows the structure found the right two-runner spine but ordered it incorrectly. Fort Augustus and Adelaide Bay were the live pair, but the anchor was wrong. This is an ordering failure rather than a selection-pool failure.

Race 7 confirms the value of winner-first discipline. South Kensington was a clean Win Pick and won. The failure was partner selection, not anchor quality.

Race 8 confirms that market-weak AU anchors can still run well without winning. Forever Noah finished 2nd and Bomb Squad finished 3rd, but Albert Cee won from outside the V15 three. That is a forecast exposure and anchor failure.

Refinement:
Market weakness versus AU should remain a stronger caution, especially when the selected AU leader is not also supported by leading market compression. In tied or low-density AU races, partner protection should not be mistaken for Win Pick confidence.

Charter discipline held.

No simulation.

No unsupported payout.

No TOTE P/L bracket printed without a landed qualifying structure.
 
AJ comment:
I will start the ED build using this thread in about one hour. As a one-off experiment, please consider the following track information for Bath. This is useful. It should act as a track suitability and combo-shape filter, not a new master ranking system. It helps stop us from forcing a paper-top runner into the win slot when the course shape says the horse may be tactically or physically vulnerable. Bath Track Filter: left-handed, undulating, sharp home turn, stiff uphill finish, often quick ground. Upgrade handy, balanced runners with firm-ground/course evidence. Downgrade early-committing pace without stamina, slow starters, and runners lacking bend/quick-ground proof. Current GoingGood to Firm (Good in places)Current Weather today at: BathTodaySunny spellsForecast: Sunny spellsTemp: 12.7° - 24.7°Rainfall: 0.12mmWind: 8.5 mph SWTomorrowSunnyForecast: SunnyTemp: 10.7° - 22.4°Rainfall: noneWind: 7.4 mph WLast Update: 22/05/2026 07:00

GPT_HRE:
Agreed. For today’s Bath ED build, I’ll treat this as a one-off Bath-specific track suitability layer, not as a replacement for AU or the main V15 hierarchy.

For the build later, I’ll apply it like this:

Bath Track Filter — active for today only

Bath profile:
  • Left-handed
  • Undulating
  • Sharp home turn
  • Stiff uphill finish
  • Good to Firm, Good in places
  • Dry/sunny profile, very little rainfall
  • Likely emphasis on balance, tactical efficiency, quick-ground handling and ability to finish
Upgrade
  • Handy, balanced runners
  • Horses with Bath form or similar undulating/stiff-track evidence
  • Proven Good to Firm / quick-ground performers
  • Runners who travel smoothly around bends
  • Horses with enough stamina to finish after committing
Downgrade
  • Early pace without clear stamina
  • Slow starters
  • Long-striding types who may need time to organise
  • Runners lacking firm-ground proof
  • Paper-top horses whose profile looks tactically vulnerable at Bath
  • Horses likely to get competitive too soon and weaken up the climb
For combos, I’ll use it to shape:

Win Pick discipline:
Only upgrade into the win slot if the horse fits the V15 evidence and passes the Bath filter.

Forecast/TOTE structure:
Strong finishers, course-suitable place types, and tactical closers can be treated as combo-strengtheners even if they are not clean win picks.

Caution markers:
Any short-priced or AU-strong runner with Bath concerns should be flagged rather than blindly trusted.

Clean addition. It should help today’s card without contaminating the wider V15 ED method.
 
▸ BATH FRIDAY 22 MAY 2026 — V15 EARLY DOORS | LIVE STRUCTURE

Bath is handled through the V15 Early Doors audit frame, with AU figs, Smart Stats and market structure kept in their proper order.
This is a structure-first racecard, not a tipping post.

• AU integrity checked race-by-race before forecast structure
• Smart Stats used for hot/cold jockey-trainer handling and supported course markers
• Market data used only for compression, caution and tie-break context
• Bath track filter applied as a suitability layer, not a replacement ranking system
• Caution markers include beaten favourite LTO, stable switch, headgear and market weakness versus AU
• Forecast and TOTE structures are built outward from the Win Pick anchor
• Charter discipline remains active: no simulation, no hindsight, no result claims

Read the full card... Bath Friday 22 May 2026 – V15 Overlay Blog | Tactical Forecasts + Smart Stats Integration | Horse Racing Hobbyist

📌 FINAL SUMMARY SECTION

🔵 Top Win Picks
• Race 1: Basilette
• Race 2: Madrisa
• Race 3: Jaan Ki Tukri
• Race 4: Silver Wraith
• Race 5: Control Room
• Race 6: Southbank
• Race 7: Up The Anti
• Race 8: Amathus

🟡 Forecast Combos
• Race 1: Basilette → Cloudy Rose / Everest
• Race 2: Madrisa → Supplicate / Eloquencia
• Race 3: Jaan Ki Tukri → Flight Signal / Paper View
• Race 4: Silver Wraith → Toolatetonegotiate / Moe's Legacy
• Race 5: Control Room → Guernsey Lady / Who Is Alice
• Race 6: Southbank → Legendsoftheland / Secret Handsheikh
• Race 7: Up The Anti → Newyorkstateofmind / Scent Of May
• Race 8: Amathus → Rosemary's Rose / Danehill Star

🟢 EW/Combo Value Inclusions
• Cloudy Rose
• Everest
• Supplicate
• Eloquencia
• Flight Signal
• Paper View
• Toolatetonegotiate
• Moe's Legacy
• Guernsey Lady
• Who Is Alice
• Legendsoftheland
• Secret Handsheikh
• Newyorkstateofmind
• Scent Of May
• Rosemary's Rose
• Danehill Star

🎲 TOTE Combos Recap
• Race 1: Basilette + Cloudy Rose / Everest
• Race 2: Madrisa + Supplicate / Eloquencia
• Race 3: Jaan Ki Tukri + Flight Signal / Paper View
• Race 4: Silver Wraith + Toolatetonegotiate / Moe's Legacy
• Race 5: Control Room + Guernsey Lady / Who Is Alice
• Race 6: Southbank + Legendsoftheland / Secret Handsheikh
• Race 7: Up The Anti + Newyorkstateofmind / Scent Of May
• Race 8: Amathus + Rosemary's Rose / Danehill Star

⚠️ Caution Marker List
• Big Win – beaten favourite LTO and stable switch evidenced from uploaded layers
• Jaan Ki Tukri – class-drop volatility evidenced from uploaded layers
• Toolatetonegotiate – beaten favourite LTO evidenced from uploaded layers
• Double Naughty – beaten favourite LTO and stable switch evidenced from uploaded layers
• Secret Handsheikh – market weakness versus AU and headgear evidenced from uploaded layers
• Newyorkstateofmind – headgear and long-distance travel evidenced from uploaded layers
• Amathus – market weakness versus AU and headgear evidenced from uploaded layers

🔒 Charter Reminder:

This is a tactical overlay system, not a tipping blog.
No simulation.

VALIDATION & TRUST LAYER

AU integrity

• Race 1: AU integrity evidenced — Basilette led uploaded points totals with 10pts.
• Race 2: AU integrity evidenced — Madrisa led uploaded points totals with 16pts.
• Race 3: AU integrity evidenced — Jaan Ki Tukri led uploaded points totals with 16pts.
• Race 4: AU integrity evidenced — Silver Wraith led uploaded points totals with 12pts.
• Race 5: AU integrity evidenced — Guernsey Lady and Control Room tied on 8pts; Control Room retained by market-compression tie-break support.
• Race 6: AU integrity evidenced — Southbank led uploaded points totals with 15pts.
• Race 7: AU integrity evidenced — Up The Anti led uploaded points totals with 12pts.
• Race 8: AU integrity evidenced — Amathus led uploaded points totals with 10pts. : SEE THE BLOG! ;)
 
AJ comment:
I will start the ED build using this thread in about one hour. As a one-off experiment, please consider the following track information for Bath. This is useful. It should act as a track suitability and combo-shape filter, not a new master ranking system. It helps stop us from forcing a paper-top runner into the win slot when the course shape says the horse may be tactically or physically vulnerable. Bath Track Filter: left-handed, undulating, sharp home turn, stiff uphill finish, often quick ground. Upgrade handy, balanced runners with firm-ground/course evidence. Downgrade early-committing pace without stamina, slow starters, and runners lacking bend/quick-ground proof. Current GoingGood to Firm (Good in places)Current Weather today at: BathTodaySunny spellsForecast: Sunny spellsTemp: 12.7° - 24.7°Rainfall: 0.12mmWind: 8.5 mph SWTomorrowSunnyForecast: SunnyTemp: 10.7° - 22.4°Rainfall: noneWind: 7.4 mph WLast Update: 22/05/2026 07:00

GPT_HRE:
Agreed. For today’s Bath ED build, I’ll treat this as a one-off Bath-specific track suitability layer, not as a replacement for AU or the main V15 hierarchy.

For the build later, I’ll apply it like this:

Bath Track Filter — active for today only

Bath profile:
  • Left-handed
  • Undulating
  • Sharp home turn
  • Stiff uphill finish
  • Good to Firm, Good in places
  • Dry/sunny profile, very little rainfall
  • Likely emphasis on balance, tactical efficiency, quick-ground handling and ability to finish
Upgrade
  • Handy, balanced runners
  • Horses with Bath form or similar undulating/stiff-track evidence
  • Proven Good to Firm / quick-ground performers
  • Runners who travel smoothly around bends
  • Horses with enough stamina to finish after committing
Downgrade
  • Early pace without clear stamina
  • Slow starters
  • Long-striding types who may need time to organise
  • Runners lacking firm-ground proof
  • Paper-top horses whose profile looks tactically vulnerable at Bath
  • Horses likely to get competitive too soon and weaken up the climb
For combos, I’ll use it to shape:

Win Pick discipline:
Only upgrade into the win slot if the horse fits the V15 evidence and passes the Bath filter.

Forecast/TOTE structure:
Strong finishers, course-suitable place types, and tactical closers can be treated as combo-strengtheners even if they are not clean win picks.

Caution markers:
Any short-priced or AU-strong runner with Bath concerns should be flagged rather than blindly trusted.

Clean addition. It should help today’s card without contaminating the wider V15 ED method.
AJ:
Did you include the one-off experiment filter?"Agreed. For today’s Bath ED build, I’ll treat this as a one-off Bath-specific track suitability layer, not as a replacement for AU or the main V15 hierarchy."

GPT_HRE:
Yes — partly, but not as cleanly or explicitly as it should have been.

I did include the Bath one-off filter in the build logic in places, especially through:

Race 1
  • Basilette was strengthened by recent Bath winning evidence, firm-ground suitability, and ability to finish strongly.
  • Cloudy Rose and Everest were treated more as combo/forecast shapes because they had staying-on or late-finishing profiles rather than clean win-anchor dominance.
Race 3
  • Paper View was retained partly because of Bath course evidence, but not forced above the stronger AU/market structure.
Race 4
  • Silver Wraith was reinforced by Bath course-and-distance evidence, sound-surface suitability, and tactical reliability.
  • Moe’s Legacy and Toolatetonegotiate were kept as combo runners because they had Bath/sprint-course evidence but carried caution markers.
Facebook post
  • I explicitly referenced it here:
    “Bath track filter applied as a suitability layer, not a replacement ranking system.”
Where I should tighten it:

I did not print a dedicated Bath Track Filter validation line in the main blog or Validation & Trust Layer. So while the filter influenced the shape, it was not documented clearly enough as a standalone one-off experimental layer.

The cleanest fix would be to add this to the Validation & Trust Layer, not rewrite the whole build:

Bath Track Filter — One-Off Experimental Layer

• Applied as a suitability and combo-shape filter only.
• AU hierarchy remained primary.
• Upgraded evidence: Bath form, firm-ground suitability, balance around the turn, ability to finish after the stiff climb.
• Downgraded evidence: slow starts, early-committing pace without stamina, weak firm-ground proof, and tactical vulnerability around Bath’s sharp home turn.
• Used to support Win Pick discipline and forecast/TOTE partner shape, not to override AU points.

So the honest answer is: yes, it was used, but it should have been surfaced more explicitly in the audit trail.
 
▸ PONTEFRACT FRIDAY 22 MAY 2026 — V15 EARLY DOORS | LIVE STRUCTURE

Audit-based race structure for Pontefract, built from uploaded AU layers, Smart Stats, racecard evidence and market context.
No tipping language, no simulation, no result claims.

• AU integrity checked race-by-race against uploaded points totals
• Win Picks bound through forecast and TOTE anchor structure
• Smart Stats used for hot/cold jockey-trainer handling
• BF LTO, class-drop, headgear and weighted-to-win flags isolated where evidenced
• Market compression used only as support, not as an override
• Caution markers retained where uploaded layers show volatility
• Forecast and TOTE partners built outward from the AU anchor
• Charter discipline enforced: model structure, not betting instruction

Read the full card... Pontefract Friday 22 May 2026 – V15 Overlay Blog | Tactical Forecasts + Smart Stats Integration | Horse Racing Hobbyist

📌 FINAL SUMMARY SECTION

🔵 Top Win Picks
• Race 1: NAVAL LIGHT
• Race 2: RICKY PUNTING
• Race 3: JOJO RABBIT
• Race 4: TROJAN SOLDIER
• Race 5: HAVACHOC
• Race 6: NO KNEE NEVER

🟡 Forecast Combos
• Race 1: NAVAL LIGHT → VIPER / KING'S ROAD DANDY
• Race 2: RICKY PUNTING → TOO MANY SPIRITS / ADORAMUS
• Race 3: JOJO RABBIT → BETWEENTHESTICKS / MRS TRUMP
• Race 4: TROJAN SOLDIER → SEA OF KINGS / VAGUELY ROYAL
• Race 5: HAVACHOC → DUBAI VENTURE / MUHIB
• Race 6: NO KNEE NEVER → ARKENSTAAR / POWERFUL RESPONSE

🟢 EW/Combo Value Inclusions
• VIPER
• KING'S ROAD DANDY
• TOO MANY SPIRITS
• ADORAMUS
• BETWEENTHESTICKS
• MRS TRUMP
• SEA OF KINGS
• VAGUELY ROYAL
• DUBAI VENTURE
• MUHIB
• ARKENSTAAR
• POWERFUL RESPONSE

🎲 TOTE Combos Recap
• Race 1: NAVAL LIGHT + VIPER / KING'S ROAD DANDY
• Race 2: RICKY PUNTING + TOO MANY SPIRITS / ADORAMUS
• Race 3: JOJO RABBIT + BETWEENTHESTICKS / MRS TRUMP
• Race 4: TROJAN SOLDIER + SEA OF KINGS / VAGUELY ROYAL
• Race 5: HAVACHOC + DUBAI VENTURE / MUHIB
• Race 6: NO KNEE NEVER + ARKENSTAAR / POWERFUL RESPONSE

⚠️ Caution Marker List
• NAVAL LIGHT – Class-drop volatility
• BETWEENTHESTICKS – Beaten favourite LTO + market weakness versus AU
• SEA OF KINGS – Beaten favourite LTO + first-time headgear
• HAVACHOC – Beaten favourite LTO + first-time headgear
• NO KNEE NEVER – Market weakness versus AU

🔒 Charter Reminder:

This is a tactical overlay system, not a tipping blog.
No simulation.

AU integrity

• Race 1: AU integrity evidenced — NAVAL LIGHT led uploaded points totals with 17pts.
• Race 2: AU integrity evidenced — RICKY PUNTING led uploaded points totals with 12pts.
• Race 3: AU integrity evidenced — JOJO RABBIT and BETWEENTHESTICKS tied on 8pts; JOJO RABBIT retained by market-lead compression.
• Race 4: AU integrity evidenced — TROJAN SOLDIER led uploaded points totals with 14pts.
• Race 5: AU integrity evidenced — HAVACHOC led uploaded points totals with 16pts.
• Race 6: AU integrity evidenced — NO KNEE NEVER led uploaded points totals with 8pts. : SEE THE BLOG! ;)
 
▸ BATH FRIDAY 22 MAY 2026 — V15 EARLY DOORS | LIVE STRUCTURE

Bath is handled through the V15 Early Doors audit frame, with AU figs, Smart Stats and market structure kept in their proper order.
This is a structure-first racecard, not a tipping post.

• AU integrity checked race-by-race before forecast structure
• Smart Stats used for hot/cold jockey-trainer handling and supported course markers
• Market data used only for compression, caution and tie-break context
• Bath track filter applied as a suitability layer, not a replacement ranking system
• Caution markers include beaten favourite LTO, stable switch, headgear and market weakness versus AU
• Forecast and TOTE structures are built outward from the Win Pick anchor
• Charter discipline remains active: no simulation, no hindsight, no result claims

Read the full card... Bath Friday 22 May 2026 – V15 Overlay Blog | Tactical Forecasts + Smart Stats Integration | Horse Racing Hobbyist

📌 FINAL SUMMARY SECTION

🔵 Top Win Picks
• Race 1: Basilette
• Race 2: Madrisa
• Race 3: Jaan Ki Tukri
• Race 4: Silver Wraith
• Race 5: Control Room
• Race 6: Southbank
• Race 7: Up The Anti
• Race 8: Amathus

🟡 Forecast Combos
• Race 1: Basilette → Cloudy Rose / Everest
• Race 2: Madrisa → Supplicate / Eloquencia
• Race 3: Jaan Ki Tukri → Flight Signal / Paper View
• Race 4: Silver Wraith → Toolatetonegotiate / Moe's Legacy
• Race 5: Control Room → Guernsey Lady / Who Is Alice
• Race 6: Southbank → Legendsoftheland / Secret Handsheikh
• Race 7: Up The Anti → Newyorkstateofmind / Scent Of May
• Race 8: Amathus → Rosemary's Rose / Danehill Star

🟢 EW/Combo Value Inclusions
• Cloudy Rose
• Everest
• Supplicate
• Eloquencia
• Flight Signal
• Paper View
• Toolatetonegotiate
• Moe's Legacy
• Guernsey Lady
• Who Is Alice
• Legendsoftheland
• Secret Handsheikh
• Newyorkstateofmind
• Scent Of May
• Rosemary's Rose
• Danehill Star

🎲 TOTE Combos Recap
• Race 1: Basilette + Cloudy Rose / Everest
• Race 2: Madrisa + Supplicate / Eloquencia
• Race 3: Jaan Ki Tukri + Flight Signal / Paper View
• Race 4: Silver Wraith + Toolatetonegotiate / Moe's Legacy
• Race 5: Control Room + Guernsey Lady / Who Is Alice
• Race 6: Southbank + Legendsoftheland / Secret Handsheikh
• Race 7: Up The Anti + Newyorkstateofmind / Scent Of May
• Race 8: Amathus + Rosemary's Rose / Danehill Star

⚠️ Caution Marker List
• Big Win – beaten favourite LTO and stable switch evidenced from uploaded layers
• Jaan Ki Tukri – class-drop volatility evidenced from uploaded layers
• Toolatetonegotiate – beaten favourite LTO evidenced from uploaded layers
• Double Naughty – beaten favourite LTO and stable switch evidenced from uploaded layers
• Secret Handsheikh – market weakness versus AU and headgear evidenced from uploaded layers
• Newyorkstateofmind – headgear and long-distance travel evidenced from uploaded layers
• Amathus – market weakness versus AU and headgear evidenced from uploaded layers

🔒 Charter Reminder:

This is a tactical overlay system, not a tipping blog.
No simulation.

VALIDATION & TRUST LAYER

AU integrity

• Race 1: AU integrity evidenced — Basilette led uploaded points totals with 10pts.
• Race 2: AU integrity evidenced — Madrisa led uploaded points totals with 16pts.
• Race 3: AU integrity evidenced — Jaan Ki Tukri led uploaded points totals with 16pts.
• Race 4: AU integrity evidenced — Silver Wraith led uploaded points totals with 12pts.
• Race 5: AU integrity evidenced — Guernsey Lady and Control Room tied on 8pts; Control Room retained by market-compression tie-break support.
• Race 6: AU integrity evidenced — Southbank led uploaded points totals with 15pts.
• Race 7: AU integrity evidenced — Up The Anti led uploaded points totals with 12pts.
• Race 8: AU integrity evidenced — Amathus led uploaded points totals with 10pts. : SEE THE BLOG! ;)
▸ BATH FRIDAY 22 MAY 2026 — V15 EARLY DOORS | LIVE STRUCTURE

Bath moved through the V15 Early Doors audit frame with AU figs, Smart Stats and market structure kept in sequence.
The post-race critique keeps the focus on structure, evidence and charter discipline.

• AU integrity checked race-by-race against uploaded points totals
• Smart Stats used for supported hot/cold, headgear, weighted-to-win and stable-switch markers
• Forecast zones reviewed separately from Win Pick authority
• TOTE structure assessed under strict Exacta and boxed Trifecta rules
• Caution and chaos control tracked through BF LTO, headgear, class drops, market weakness and withdrawals
• Bath track filter treated as a one-off suitability layer, not a replacement ranking system
• Betting outcome kept separate from model integrity

Read the full card and post-race critique: Bath Friday 22 May 2026 – V15 Overlay Blog | Tactical Forecasts + Smart Stats Integration | Horse Racing Hobbyist

📉 Cumulative Outcome Analysis

Structured Yankee:
FAILED

Stake:
£3.30

Return:
£0.00

TOTE Exacta outcomes:
No V15 Exacta landed under the enforced winner-first exacta rule.

TOTE Trifecta outcomes:
Race 2 boxed trifecta landed only.

TOTE Trifecta: £8.90 (P/L: +£2.90)

Overall:
The card produced one valid boxed trifecta return from the V15 forecast structure.

The broader card was marked by repeated anchor failure:

Race 1: Win Pick 3rd.
Race 2: Win Pick 3rd.
Race 3: Win Pick 2nd.
Race 4: Win Pick 2nd.
Race 5: Win Pick 2nd.
Race 6: Win Pick unplaced.
Race 7: Win Pick withdrawn.
Race 8: Win Pick unplaced.

The strongest model output was not winner selection. It was forecast-zone identification across several races.

The weakest model output was converting AU-led anchors into race winners.

🔧 Model Reflection & Refinement Notes

The AU hierarchy remained auditable, but winner-first execution was below the required standard.

Race 2 confirmed that the three-runner structure can land when the AU cluster is sound, but the win ordering was wrong.

Race 3 and Race 4 showed the same pattern: the winner was inside the selected structure, but not in the anchor position.

Race 5 showed that the caution layer correctly identified Double Naughty as a danger profile, but the model did not elevate that danger into the winning layer.

Race 6 showed that Partner A was live enough to win, while the selected anchor failed.

Race 7 cannot be repaired through post-race logic. The uploaded result lists Up The Anti as withdrawn.

The Bath track filter helped identify some course-suitable and combo-relevant runners, but it was not strong enough as applied to improve win-pick strike rate.

Carry-forward refinement:
AU points leadership should remain primary, but where Bath-style evidence highlights strong course/tactical suitability in a partner or caution runner, that evidence should be forced into a sharper win-anchor challenge before lock.

Classification:
Forecast structure partially held.
Winner-first performance failed.

Charter discipline:
Model ≠ Result.
Betting outcome separated from model integrity.
No unsupported payout printed.
No failed TOTE payout printed.
 
Back
Top