UFC Analysis

Sportsbet-priced upcoming bouts shown in Melbourne time. UFCgd is now operating in dog-bets-only mode. Model View audits the favourite first, then only promotes an underdog when the favourite looks overpriced, the dog has a credible fight path, and the historical dog-edge bucket supports the price.

Priced Bouts

Release-Grade Markov Rows

False Fav / Dog Rows

Clean Rows

Model Price Movement - refreshed 15/07/2026 16:51 Brisbane

Fresh Odds API pull across 47 current MMA events and 509 current price rows. Movement is measured against the saved Model selection ledger.

Bout	Selection	Ledger Price	Current Best	Sportsbet	Pinnacle	Betfair	Move	Model Note
Mackenzie Dern vs Gillian Robertson	Gillian Robertson	3.00 TAB	3.02 Pinnacle	2.72	3.02	-	Steady 0.7%	No urgent market signal. Edge remains model-led rather than price-move-led.

Results & Model Learnings - run 2026-07-19T18:02:50+1000

Settled UFC Oklahoma City selections. Completed fights are no longer member-facing best bets; they are now learning inputs.

Settled Bet Ideas

Record

0-3

Profit

-3.00u

Proven Sharpness

70/100

Bout	Selection	Result	Profit
Christian Leroy Duncan vs Jared Cannonier	Jared Cannonier	LOSS	-1.00u
Dricus Du Plessis vs Kamaru Usman	Kamaru Usman	LOSS	-1.00u
Fatima Kline vs Tabatha Ricci	Tabatha Ricci	LOSS	-1.00u

Implemented Learnings

Learning
Price-only dogs are downgraded to monitor unless the Markov fight path also leans dog.
Conflict-with-V2 dogs cannot be release-grade without a clear control/grappling route.
Old-fighter risk is a hard downgrade when the selected dog is 35+ and the younger favourite has the cleaner pace/damage profile.
Dog-edge historical bucket is no longer sufficient on its own; it must be supported by age, durability, path and opponent-quality checks.
Recent form remains capped, but favourite-side damage plus youth now overrides stale veteran name recognition.

Future Best Bets / Settled Learning Board

Completed July 19 rows have been settled and moved into model learning. Future rows remain for monitoring.

16/08/2026 13:00 · Mackenzie Dern vs Gillian Robertson

Gillian Robertson

Gillian Robertson has the better value path; Mackenzie Dern looks too short.

Best Price 3.02 Pinnacle

Edge +15.4 pts Strong edge · Fair 1.92

Confidence Strong watch Qualifies (4/4)

Opened 3.00 at TAB; now 3.02 at Pinnacle. Price is broadly steady.

Member View

Decision-first board. Technical profile detail is in Fight Path Evidence.

Dog-Bets-Only Guardrail

Historical dog bucket has positive ROI.
Dog edge is at least 5 percentage points.
Sample quality is usable or clean.
The dog has a credible fight path, not just a cheap price.
Favourite edges are pass/filter-only, not selections.

Qualifying Bouts

19/07/2026 11:45

Dricus Du Plessis vs Kamaru Usman

Qualifies (4/4)

Model ViewKamaru Usman

Playable SideKamaru Usman at 3.10 @ Betfair | fair 1.79

Dog Edge24.5 pts

Market CheckBest of 36 books

Historical SupportDog edge 5%+: 54 bets, 74.07% hit, 12.04% ROI

Kamaru Usman is not being priced like a real chance, but his clearest route is to slow the fight down, win clinch/wrestling minutes and make Dricus Du Plessis work harder than the market expects. Model's non-market read has the fight path with Kamaru Usman, overall profile with Kamaru Usman, recent form with Dricus Du Plessis, and finishing threat with Dricus Du Plessis. The fresh market has shortened Usman from the saved 3.50 ledger price to 3.10 best on Betfair, so the original number has beaten the current market. That is a good sign for the read, but it also means the new entry is less attractive and still needs fight-week confirmation.

Selection Guardrail

PASS Historical dog bucket
PASS Dog edge >= 5 pts
PASS Usable sample
PASS Dog has a credible fight path

19/07/2026 09:45

Fatima Kline vs Tabatha Ricci

Qualifies (4/4)

Model ViewFatima Kline

Playable SideTabatha Ricci at 4.46 @ Pinnacle | fair 2.03

Dog Edge21.7 pts

Market CheckBest of 31 books

Historical SupportDog edge 5%+: 54 bets, 74.07% hit, 12.04% ROI

Fatima Kline may still win, but the market is charging a premium price for her. Model has Fatima Kline only around 50.7%, not a 1.29 type of favorite. Tabatha Ricci's route is to turn this into a grinding fight, extend control time and keep rounds close. The non-market profile splits this fight: form/finish lean Fatima Kline, but fight path gives Tabatha Ricci enough of a route that 4.46 versus fair 2.03 becomes the value. The key market point is that Pinnacle has drifted Ricci wider than the saved 4.25 ledger price, so Model should treat this as the highest-priority confirmation watch rather than a blind release.

Selection Guardrail

PASS Historical dog bucket
PASS Dog edge >= 5 pts
PASS Usable sample
PASS Dog has a credible fight path

16/08/2026 13:00

Mackenzie Dern vs Gillian Robertson

Qualifies (4/4)

Model ViewGillian Robertson

Playable SideGillian Robertson at 3.02 @ Pinnacle | fair 1.92

Dog Edge15.4 pts

Market CheckBest of 26 books

Historical SupportDog edge 5%+: 54 bets, 74.07% hit, 12.04% ROI

Model makes Gillian Robertson the more likely winner at about 52.2%. The simple read: Gillian Robertson has the better repeatable path to bank minutes, especially if the fight spends time in clinch, grappling or top-control positions. The overall non-market profile says: Form dog | Finish fav | Profile even | Fight path dog. The market is still treating Mackenzie Dern like the clear side, and Robertson is broadly steady from 3.00 to 3.02 at Pinnacle.

Selection Guardrail

PASS Historical dog bucket
PASS Dog edge >= 5 pts
PASS Usable sample
PASS Dog has a credible fight path

Value / Watchlist

Time	Bout	Model View	Action	Price	Edge	Best Book	Guardrail	Historical Support	Confidence	Why
16/08/2026 13:00	Mackenzie Dern vs Gillian Robertson	Gillian Robertson	Dog candidate: Gillian Robertson	Gillian Robertson 3.02 @ Pinnacle \| fair 1.92	+15.4 pts dog	Pinnacle	Qualifies (4/4)	Dog edge 5%+: 54 bets, 74.07% hit, 12.04% ROI	Strong watch	Model makes Gillian Robertson the more likely winner at about 52.2%. The simple read: Gillian Robertson has the better repeatable path to bank minutes, especially if the fight spends time in clinch, grappling or top-control positions. The overall non-market profile says: Form dog \| Finish fav \| Profile even \| Fight path dog. The market is still treating Mackenzie Dern like the clear side, so 3.02 at Pinnacle remains too generous.

Monitor / Pass

Time	Bout	Model View	Action	Price	Edge	Best Book	Guardrail	Historical Support	Confidence	Why
19/07/2026 07:00	Seok Hyun Ko vs Jean-Paul Lebosnoyani	Jean-Paul Lebosnoyani	Monitor only	Seok Hyun Ko 1.5 @ BetOnline.ag \| fair 2.58	25.4 pts	BetOnline.ag	Does not qualify (2/4)	No strong historical bucket	Low	Model leans Jean-Paul Lebosnoyani, but there is not enough reliable UFC evidence yet. Profile check: Form dog \| Finish even \| Profile dog \| Fight path dog. Treat this as a watch-only fight: look for weigh-in signals, late market movement and any camp news before taking it seriously.
19/07/2026 07:00	Alden Coria vs Stewart Nicoll	Alden Coria	Monitor only	Alden Coria 1.12 @ PlayUp \| fair 1.48	20.6 pts	PlayUp	Does not qualify (2/4)	No strong historical bucket	Low	Model leans Alden Coria, but there is not enough reliable UFC evidence yet. Profile check: Form fav \| Finish fav \| Profile fav \| Fight path fav. Treat this as a watch-only fight: look for weigh-in signals, late market movement and any camp news before taking it seriously.
19/07/2026 10:45	Chase Hooper vs Mitch Ramirez	Chase Hooper	Monitor only	Chase Hooper 1.33 @ PlayUp \| fair 1.59	10.0 pts	PlayUp	Does not qualify (1/4)	No strong historical bucket	Low	Model leans Chase Hooper, but there is not enough reliable UFC evidence yet. Profile check: Form fav \| Finish fav \| Profile fav \| Fight path fav. Treat this as a watch-only fight: look for weigh-in signals, late market movement and any camp news before taking it seriously.
19/07/2026 07:00	Alberto Montes vs Tommy McMillen	Tommy McMillen	Monitor only	Alberto Montes 2.0 @ Caesars \| fair 2.19	6.7 pts	Caesars	Does not qualify (1/4)	No strong historical bucket	Low	Model leans Tommy McMillen, but there is not enough reliable UFC evidence yet. Profile check: Form even \| Finish dog \| Profile even \| Fight path even. Treat this as a watch-only fight: look for weigh-in signals, late market movement and any camp news before taking it seriously.
19/07/2026 07:00	Jose Delgado vs Austin Bashi	Jose Delgado	Monitor only	Jose Delgado 1.96 @ 1xBet \| fair 1.78	6.7 pts	1xBet	Does not qualify (1/4)	No strong historical bucket	Low	Model leans Jose Delgado, but there is not enough reliable UFC evidence yet. Profile check: Form fav \| Finish fav \| Profile fav \| Fight path dog. Treat this as a watch-only fight: look for weigh-in signals, late market movement and any camp news before taking it seriously.
18/07/2026 19:00	Levi Rodrigues vs Felipe Franco	Levi Rodrigues	Monitor only	Levi Rodrigues 1.56 @ Caesars \| fair 1.78	6.6 pts	Caesars	Does not qualify (0/4)	No strong historical bucket	Low	Model leans Levi Rodrigues, but there is not enough reliable UFC evidence yet. Profile check: Form even \| Finish even \| Profile fav \| Fight path even. Treat this as a watch-only fight: look for weigh-in signals, late market movement and any camp news before taking it seriously.
26/07/2026 02:00	Magomed Ankalaev vs Khalil Rountree	Magomed Ankalaev	Monitor only	Magomed Ankalaev 1.33 @ Paddy Power \| fair 1.49	5.4 pts	Paddy Power	Does not qualify (0/4)	No strong historical bucket	Low	Model leans Magomed Ankalaev, but there is not enough reliable UFC evidence yet. Profile check: Form fav \| Finish fav \| Profile fav \| Fight path fav. Treat this as a watch-only fight: look for weigh-in signals, late market movement and any camp news before taking it seriously.
16/08/2026 14:00	Islam Makhachev vs Ian Garry	Islam Makhachev	Monitor only	Islam Makhachev 1.29 @ DraftKings \| fair 1.31	2.4 pts	DraftKings	Does not qualify (0/4)	No strong historical bucket	Low	Model leans Islam Makhachev, but there is not enough reliable UFC evidence yet. Profile check: Form fav \| Finish fav \| Profile fav \| Fight path fav. Treat this as a watch-only fight: look for weigh-in signals, late market movement and any camp news before taking it seriously.

Priority Rows

event_time_melbourne	favorite	underdog	sportsbet_fav_decimal	sportsbet_dog_decimal	best_fav_book	best_fav_price	best_dog_book	best_dog_price	book_count	market_fav_pct	markov_fav_pct	fav_edge_pts	markov_dog_pct	dog_edge_pts	verdict	action_tier	markov_lean	model_alignment	data_quality	driver
19/07/2026 11:45	Dricus Du Plessis	Kamaru Usman	1.36	2.93	DraftKings	1.39	Betfair	3.10	36	70.0	44.1	-26.0	55.9	24.5	False Favorite / Value Dog	Dog Watch: Split Model	Kamaru Usman	Conflict with V2	clean	control momentum dog; damage chain fav
19/07/2026 09:45	Fatima Kline	Tabatha Ricci	1.29	3.90	DraftKings	1.31	Pinnacle	4.46	31	73.7	50.7	-23.0	49.3	21.7	Favorite Overpriced / Dog Value	Dog Watch: Price Only	Fatima Kline	Conflict with V2	usable	control momentum dog; damage chain fav; recent form fav
16/08/2026 13:00	Mackenzie Dern	Gillian Robertson	1.44	2.72	DraftKings	1.51	Pinnacle	3.02	26	65.4	47.8	-17.6	52.2	15.4	False Favorite / Value Dog	Dog Watch: Markov-Only	Gillian Robertson	Markov-only	clean	control momentum dog; damage chain fav

Full Upcoming Markov Board

event_time_melbourne	favorite	underdog	sportsbet_fav_decimal	sportsbet_dog_decimal	best_fav_book	best_fav_price	best_dog_book	best_dog_price	book_count	market_fav_pct	markov_fav_pct	fav_edge_pts	markov_dog_pct	dog_edge_pts	verdict	action_tier	markov_lean	model_alignment	data_quality	driver
19/07/2026 07:00	Seok Hyun Ko	Jean-Paul Lebosnoyani	1.48	2.66	BetOnline.ag	1.5	BetOnline.ag	2.7	26	64.3	38.8	-25.4	61.2	23.6	Watch: Thin Markov Sample	No Bet / Monitor	Jean-Paul Lebosnoyani	Aligned with V2	thin	control momentum dog; recent form dog; posterior skill dog
19/07/2026 11:45	Dricus Du Plessis	Kamaru Usman	1.36	2.93	DraftKings	1.39	Betfair	3.10	36	70.0	44.1	-26.0	55.9	24.5	False Favorite / Value Dog	Dog Watch: Split Model	Kamaru Usman	Conflict with V2	clean	control momentum dog; damage chain fav
19/07/2026 09:45	Fatima Kline	Tabatha Ricci	1.29	3.90	DraftKings	1.31	Pinnacle	4.46	31	73.7	50.7	-23.0	49.3	21.7	Favorite Overpriced / Dog Value	Dog Watch: Price Only	Fatima Kline	Conflict with V2	usable	control momentum dog; damage chain fav; recent form fav
19/07/2026 07:00	Alden Coria	Stewart Nicoll	1.08	8.0	PlayUp	1.12	Ladbrokes	8.25	30	88.1	67.5	-20.6	32.5	20.0	Watch: Thin Markov Sample	No Bet / Monitor	Alden Coria	Markov-only	thin	posterior skill fav; recent form fav; damage chain fav
16/08/2026 13:00	Mackenzie Dern	Gillian Robertson	1.44	2.72	DraftKings	1.51	Pinnacle	3.02	26	65.4	47.8	-17.6	52.2	15.4	False Favorite / Value Dog	Dog Watch: Markov-Only	Gillian Robertson	Markov-only	clean	control momentum dog; damage chain fav
19/07/2026 11:15	Christian Leroy Duncan	Jared Cannonier	1.35	3.57	PlayUp	1.38	Betsson / Betfair	3.90	33	70.4	55.3	-15.0	44.7	13.5	Favorite Overpriced / Dog Value	Dog Price Watch	Christian Leroy Duncan	Aligned with V2	clean	posterior skill fav; recent form fav; damage chain fav
19/07/2026 10:45	Chase Hooper	Mitch Ramirez	1.3	3.52	PlayUp	1.33	BetOnline.ag	3.7	31	73.0	63.1	-10.0	36.9	8.5	Watch: Thin Markov Sample	No Bet / Monitor	Chase Hooper	Aligned with V2	thin	posterior skill fav; recent form fav; control momentum fav
19/07/2026 07:00	Alberto Montes	Tommy McMillen	1.83	2.01	Caesars	2.0	1xBet	2.08	32	52.3	45.6	-6.7	54.4	4.7	Watch: Thin Markov Sample	No Bet / Monitor	Tommy McMillen	Aligned with V2	thin	damage chain dog; control momentum fav
19/07/2026 07:00	Jose Delgado	Austin Bashi	1.86	1.98	1xBet	1.96	Caesars	2.1	28	51.6	56.2	4.7	43.8	-6.7	Watch: Thin Markov Sample	No Bet / Monitor	Jose Delgado	Aligned with V2	thin	control momentum dog; damage chain fav; posterior skill fav
18/07/2026 19:00	Levi Rodrigues	Felipe Franco	1.51	2.56	Caesars	1.56	Ladbrokes	2.7	31	62.9	56.3	-6.6	43.7	4.6	Watch: Thin Markov Sample	No Bet / Monitor	Levi Rodrigues	Markov-only	thin	posterior skill fav
26/07/2026 02:00	Magomed Ankalaev	Khalil Rountree	1.3	3.4	Paddy Power	1.33	1xBet	3.75	30	72.3	66.9	-5.4	33.1	3.7	Watch: Thin Markov Sample	No Bet / Monitor	Magomed Ankalaev	Markov-only	thin	posterior skill fav; recent form fav; damage chain fav
16/08/2026 14:00	Islam Makhachev	Ian Garry	1.25	3.85	DraftKings	1.29	1xBet	4.12	24	75.5	76.5	1.0	23.5	-2.4	Watch: Thin Markov Sample	No Bet / Monitor	Islam Makhachev	Markov-only	thin	posterior skill fav; recent form fav; control momentum fav

Fight Path Evidence

event_time_melbourne	favorite	underdog	fav_ufc_fights	dog_ufc_fights	fav_markov_fights	dog_markov_fights	weight_overall	weight_path	weight_finish	weight_recent	weight_certainty	recent_side_quality	recent_profile	finish_profile	overall_profile	markov_path	control_diff	damage_diff	deficit_diff	recovery_diff	driver	notes
19/07/2026 07:00	Seok Hyun Ko	Jean-Paul Lebosnoyani	0	1	0	1	0.466	0.318	0.136	0.08	0.834	-0.0	-1.0	0.0	-0.395	-0.6	-1.0	0.0	0.0	0.0	control momentum dog; recent form dog; posterior skill dog	Not release-grade until more reliable market/news confirmation.
19/07/2026 11:45	Dricus Du Plessis	Kamaru Usman	10	19	10	19	0.474	0.326	0.174	0.026	0.685	0.055	0.04	0.175	-0.48	-0.129	-0.35	0.11	0.001	0.047	control momentum dog; damage chain fav	Release only if price remains available and weigh-in/news checks are clean.
19/07/2026 09:45	Fatima Kline	Tabatha Ricci	4	11	4	11	0.463	0.335	0.184	0.018	0.619	-0.195	0.189	0.23	-0.015	-0.034	-0.22	0.189	-0.152	-0.091	control momentum dog; damage chain fav; recent form fav	Release only if price remains available and weigh-in/news checks are clean.
19/07/2026 07:00	Alden Coria	Stewart Nicoll	2	3	2	3	0.453	0.331	0.14	0.076	0.29	0.0	1.0	0.1	1.284	0.181	-0.111	0.5	-0.222	0.0	posterior skill fav; recent form fav; damage chain fav	Not release-grade until more reliable market/news confirmation.
16/08/2026 13:00	Mackenzie Dern	Gillian Robertson	16	20	16	20	0.451	0.376	0.141	0.032	0.78	-0.252	-0.045	0.084	-0.039	-0.215	-0.404	0.087	0.054	0.025	control momentum dog; damage chain fav	Release only if price remains available and weigh-in/news checks are clean.
19/07/2026 11:15	Christian Leroy Duncan	Jared Cannonier	9	20	9	20	0.479	0.317	0.152	0.052	0.697	-0.343	0.289	0.1	0.324	0.092	-0.019	0.166	-0.031	0.172	posterior skill fav; recent form fav; damage chain fav	Release only if price remains available and weigh-in/news checks are clean.
19/07/2026 10:45	Chase Hooper	Mitch Ramirez	13	2	13	2	0.468	0.348	0.141	0.042	0.499	-0.007	0.586	0.087	0.881	0.247	0.372	0.115	0.154	0.0	posterior skill fav; recent form fav; control momentum fav	Not release-grade until more reliable market/news confirmation.
19/07/2026 07:00	Alberto Montes	Tommy McMillen	1	1	1	1	0.45	0.343	0.159	0.047	0.15	-0.0	0.0	-1.2	0.032	0.0	0.5	-1.0	0.0	0.0	damage chain dog; control momentum fav	Not release-grade until more reliable market/news confirmation.
19/07/2026 07:00	Jose Delgado	Austin Bashi	4	2	4	2	0.457	0.339	0.151	0.052	0.394	0.016	0.207	0.679	0.463	-0.221	-0.833	0.75	-0.167	0.0	control momentum dog; damage chain fav; posterior skill fav	Not release-grade until more reliable market/news confirmation.
18/07/2026 19:00	Levi Rodrigues	Felipe Franco	0	1	0	1	0.483	0.33	0.141	0.045	0.952	-0.0	0.0	0.0	0.524	0.0	0.0	0.0	0.0	0.0	posterior skill fav	Not release-grade until more reliable market/news confirmation.
26/07/2026 02:00	Magomed Ankalaev	Khalil Rountree	14	0	14	0	0.466	0.318	0.136	0.08	1.0	0.344	0.824	0.34	1.096	0.259	0.283	0.336	0.095	0.071	posterior skill fav; recent form fav; damage chain fav	Not release-grade until more reliable market/news confirmation.
16/08/2026 14:00	Islam Makhachev	Ian Garry	18	0	18	0	0.466	0.318	0.136	0.08	1.0	0.312	0.956	0.199	1.97	0.495	0.72	0.176	0.011	0.0	posterior skill fav; recent form fav; control momentum fav	Not release-grade until more reliable market/news confirmation.

Model Method

How individual fighter traits currently feed the dog-only guardrail and what needs to be learned next.

Current Model Coefficients

Coefficients are now dynamic per bout. They move by fighter depth, Markov history, uncertainty, opponent quality and whether recent form agrees with longer-term evidence.

Overall profileDynamic, usually largestBaseline fighter quality: skill posterior, opponent quality, rank context, striking, takedown and control profile. Gets more weight when both fighters have deeper, lower-uncertainty records.Fight pathDynamicRepeatable round pattern: control, damage, deficit risk, recovery and submission-control states. Gets more weight when both fighters have enough round-state history and the path gap is meaningful.Finishing threatDynamicAbility to turn damage states into fight-ending threat. Gets more weight when finish/damage evidence is reliable.Recent formCapped at 10%Recency-weighted win trend. Cut down when it comes against weaker opposition or conflicts with longer-term profile/fight path.

Anti-Recency-Bias Rules

Avoid prospect hypeRecent form weight is reduced when the recent-form side has weaker opponent quality.Avoid small-sample recency biasRecent form weight is reduced when either fighter has limited UFC depth.Respect long-term profileRecent form weight is reduced when it disagrees with overall fighter profile.Respect repeatable fight pathRecent form weight is reduced when it disagrees with the Markov fight path.Reward reliable traitsOverall/profile and path weights rise when both fighters have deeper, cleaner histories.

Fight Path Coefficients

Control rate0.55 inside fight pathRewards fighters who repeatedly bank grappling, clinch or top-control minutes.Damage rate0.35 inside fight pathRewards fighters who repeatedly create striking/knockdown advantage states.Bad deficit rate-0.30 inside fight pathPenalises fighters who often fall into losing striking/position states.Recovery rate0.20 inside fight pathRewards fighters who can recover after early deficit states.Submission control0.15 inside fight pathRewards control that also creates submission threat.

Overall Profile Coefficients

Skill mean0.65 inside overall profileFighter-level posterior ability from historical results.Opponent quality0.20 inside overall profileRewards proven form against stronger opposition.Rank context0.025 inside overall profileSmall adjustment for current ranking quality.Significant strike differential0.010 inside overall profileSmall adjustment for striking volume/efficiency advantage.Takedown differential0.080 inside overall profileRewards repeatable wrestling advantage.Control differential0.045 inside overall profileRewards ability to hold position and drain minutes.

Future Improvements

Learn coefficients from settled selectionsReplace hand weights with coefficients fitted to CLV, ROI and closing-market agreement.Fighter archetype priorsSeparate pressure wrestlers, submission grapplers, counter strikers, low-volume veterans and damage-first finishers.Age and durability decayOlder fighters with strong historic skill should be discounted when durability or pace is declining.Five-round vs three-round adjustmentControl paths, cardio, late damage and comeback probability change by fight length.Camp/news and weigh-in stateLate illness, bad weight cuts and opponent changes should move confidence, not just price.Book-specific CLV trackingIdentify which books lag and whether best-price edges actually beat the close.Path validation after fightScore whether the predicted route happened, even if the bet won or lost.

Self Learning

What the latest settled results changed in the model. This is deliberately conservative: poor early outcomes reduce proven sharpness and tighten future release gates.

Process Sharpness

82/100

Proven Sharpness

70/100

Settled Bet Ideas

Settled P/L

-3.00u

Current Model Read

First settled UFCgd dog cohort lost. The process remains useful as a market-favourite filter, but release criteria must be tightened before the model can claim high proven sharpness.

Implemented Rules

Rule
Price-only dogs are downgraded to monitor unless the Markov fight path also leans dog.
Conflict-with-V2 dogs cannot be release-grade without a clear control/grappling route.
Old-fighter risk is a hard downgrade when the selected dog is 35+ and the younger favourite has the cleaner pace/damage profile.
Dog-edge historical bucket is no longer sufficient on its own; it must be supported by age, durability, path and opponent-quality checks.
Recent form remains capped, but favourite-side damage plus youth now overrides stale veteran name recognition.

Next Learning Queue

Priority	Task	Purpose
High	Tighten release gate	Require dog path plus age/durability support before member-facing release.
High	Backtest old-dog veto	Quantify whether the old-fighter risk should fully veto or merely reduce stake.
High	Re-score price-only dogs	Separate real false favourites from underdogs that are simply long for good reason.

Selection History

Append-only ledger for qualifying and watchlist selections. Settled rows stay here for audit, CLV and post-fight learning.

Total Selection Snapshots

Settled Snapshots

Open Snapshots

Latest Snapshot

2026-07-16T10:56:20+1000

Recent Selection History

Recorded	Fight Date	Bout	Selection	Book	Price	Guardrail	Result	Profit
2026-07-13T18:59:00+1000	19/07/2026 11:45	Dricus Du Plessis vs Kamaru Usman	Kamaru Usman	TAB	3.5	True	LOSS	-1.00u
2026-07-13T18:59:00+1000	19/07/2026 09:45	Fatima Kline vs Tabatha Ricci	Tabatha Ricci	Ladbrokes	4.25	True	LOSS	-1.00u
2026-07-13T18:59:00+1000	16/08/2026 13:00	Mackenzie Dern vs Gillian Robertson	Gillian Robertson	TAB	3.0	True
2026-07-13T18:59:00+1000	19/07/2026 11:15	Christian Leroy Duncan vs Jared Cannonier	Jared Cannonier	BetOnline.ag	3.7	False	LOSS	-1.00u
2026-07-16T10:56:20+1000	19/07/2026 09:45	Fatima Kline vs Tabatha Ricci	Tabatha Ricci	Ladbrokes	4.8	True	LOSS	-1.00u
2026-07-16T10:56:20+1000	19/07/2026 11:45	Dricus Du Plessis vs Kamaru Usman	Kamaru Usman	BetOnline.ag	3.05	True	LOSS	-1.00u
2026-07-16T10:56:20+1000	19/07/2026 11:15	Christian Leroy Duncan vs Jared Cannonier	Jared Cannonier	Betfair	4.0	True	LOSS	-1.00u
2026-07-16T10:56:20+1000	16/08/2026 13:00	Mackenzie Dern vs Gillian Robertson	Gillian Robertson	Pinnacle	3.02	True

Selection Trend Count

Selection	Snapshots
Kamaru Usman	2
Tabatha Ricci	2
Gillian Robertson	2
Jared Cannonier	2