AllenvsDuncan

CLAUDEThe market has Allen at about 57% after removing the vig, and Elo has him at 60% (1642 vs 1568, a 74-point gap). Both priors point the same way and are close together, so I start near 58% and only adjust for the matchup. This is a grappler against a striker over five rounds. On the feet Duncan has the edge: 4.45 strikes landed per minute at 57% accuracy, only 2.36 absorbed per minute, a +2.09 differential and a 79-inch reach (4 inches longer than Allen). Allen absorbs 3.78 per minute with 47% striking defence, so a long kickboxing fight favours Duncan. Allen's way to win is wrestling and control: 1.52 takedowns per 15 minutes, 1.1 submission attempts, and a long record of back-take finishes. Duncan's 57% takedown defence is ordinary and he averages zero submission attempts. His decision win over Roman Dolidze suggests his scrambling has improved, but Dolidze is a ground-and-pound grappler rather than a back-taker like Allen. On quality of opposition, Allen has recently beaten de Ridder (TKO in round 4 of a main event), Vettori and Shahbazyan, and his only recent losses are to Imavov and Hernandez, both elite. Duncan is on a 5-fight win streak and his Elo is rising fast (+39), but only the Cannonier win is against a clearly top-15 name. The five-round length matters: Allen has already gone past round 3 against de Ridder and Curtis, while Duncan has never fought a 25-minute bout. That makes cumulative wrestling and late submissions more likely. Duncan is dangerous in the first two rounds, especially if Allen walks into his long straight punches. Layoffs of 120 and 78 days are both normal, and there is no fight-week news. My Allen win probability is 59%. That is close to the market, so I see no strong edge. The likeliest single outcome is an Allen decision built on control, with a late submission as a real second route.

GPTAllen's 56% market baseline and 60% Elo estimate support a narrow favorite. His higher takedown and submission rates offer a route against Duncan's 57% takedown defense. Duncan counters with superior striking efficiency, a reach advantage and five consecutive wins, including decisions over Cannonier and Dolidze; his stronger Elo trend limits confidence in Allen. The updated enrichment logs take precedence over conflicting older form entries and favor a decision. Neither supplied layoff indicates prolonged inactivity. The small market move is already priced in, with no material news supporting another adjustment. Round five assumes a five-round main event; method percentages are conditional on Allen winning.

CLAUDEThe market has Camilo at about 62%, and Elo has Herbert at 55%. I mostly discount the Elo number because Camilo's rating is provisional and based on just one rated fight (the 2025 submission loss to Gabe Green). It misses his later wins over Borshchev (decision) and Sadykhov (knockout), so it is stale rather than informative. Camilo's per-minute UFC stats are also missing or blank, so I ignore them rather than treating them as zeros. Herbert's numbers are below average: 3.16 strikes landed and 3.37 absorbed per minute at 42% accuracy, 59% takedown defence and zero submission attempts. He is 38, against a 25-year-old who is still improving. Herbert has a long reach (77 inches) and can hit hard, as his first-round knockout of Mandel Nallo shows. But he has been knocked out at UFC level before, has gone 2-2-1 across his last five, and has had layoffs of more than a year between several fights. Camilo's wrestling-first style, which ground out Borshchev, along with his power against Sadykhov, gives him two ways to win. The Green submission loss shows his own back-defence can lapse, but Herbert has no submission threat to exploit it. Herbert's best chance is a counter right hand at range in the first round. I am a little under the market because Camilo's UFC sample is small, but the age gap, Herbert's ordinary takedown defence and the momentum difference all back the favourite.

GPTCamilo's 63% de-vigged market probability anchors the pick. Elo favors Herbert 55%, but Camilo's provisional rating covers only one fight and omits his two subsequent wins listed in enrichment. That conflict warrants caution, not an underdog pick. Camilo's missing statistics provide no reliable stylistic edge; Herbert's slightly negative striking differential offers only limited support. Opposition quality cannot be firmly separated from the supplied records. Neither supplied layoff approaches the 450-day risk threshold. Decision is the narrow method preference because four of Herbert's last five enriched results reached the cards. The small market move is already priced in, and scouting supplies no adjustment. Method percentages are conditional on Camilo winning.

CLAUDEThe market has Godinez at about 65% and Elo at 54%. Souza's rating rests on only 5 rated fights, so the market probably has more information here. I land between the two, nearer the market, at 62%. This is a classic striker-against-wrestler matchup. Souza is the more efficient striker, landing 5.11 per minute at 66% accuracy with a +1.62 differential, and she showed finishing power with the head-kick TKO of Carnelossi. Godinez's path is clear, though: 3.12 takedowns per 15 minutes at 47% accuracy, set up by heavy striking volume (4.28 landed per minute), with 81% takedown defence so she can't easily be taken down herself. Souza's 62% takedown defence is the weak spot. She has been submitted at UFC level (by Karine Silva in round 1), and her best strawweight results are decisions against mid-tier opponents. Godinez's last-fight submission loss came against Tatiana Suarez, an elite wrestler. That is an opponent-quality problem and doesn't make her easier to beat for a striker who doesn't wrestle. Her other losses (Dern, Jandiroba) were also decisions against top-10 opponents, and she beat Jessica Andrade, a far bigger name than anyone on Souza's record. The data shows both lost to Angela Hill, which tells us little. Godinez's high pace and stamina usually make her fights go the distance: her last five all went to decision except the Suarez finish. The 176-day layoff is normal. The most likely result is a Godinez decision where she chains takedowns and wins the clinch, while Souza's power at range keeps the upset alive.

GPTStarting from Godinez's 67% de-vigged market probability, I adjust modestly downward because Elo favors her by only 54% and Souza has stronger striking averages. Godinez's 3.12 takedown average against Souza's 62% takedown defense provides a credible route to winning rounds with effective grappling. Her 15-fight UFC sample offers more evidence than Souza's five fights. The enriched log shows two recent Godinez decision wins, while Souza's recent finishes support her upset threat. Conflicting recent-form records limit trend conclusions. Neither supplied layoff approaches 450 days. The two-point market move is already priced in, and no material news warrants another adjustment. Method probabilities are conditional on Godinez winning.

CLAUDEThere is no market line for this fight, so the call rests on form and matchup. One data note first: the stats feed flags Kamaka as a UFC debutant with no completed fights, but his own fight log shows UFC bouts against Jonathan Pearce (KO loss), TJ Brown (split-decision loss), Danny Chavez (majority draw), Dakota Hope (split-decision win, Apr 2026) and Luke Riley (KO loss, Jul 2026). His per-minute UFC stats are missing, not zero, so I don't use any of them, and the debut confidence cap of 65 still applies. Fili is 36, has lost two straight, and his recent losses show his durability fading: KO by Dan Ige, guillotine by Melquizael Costa, KO by elbows against Vinicius Oliveira in June. His wins in this stretch are split decisions over Cub Swanson and Christian Rodriguez, so even when he wins it's close. His career numbers are roughly even on the feet (3.88 strikes landed per minute vs 4.11 absorbed, -0.23 differential, 37% accuracy). His clearest edges are wrestling (2.15 takedowns per 15 at 43%, 70% takedown defense) and a 74-inch reach, which he can use to keep a shorter featherweight at range. Kamaka has gone 1-3-1 in the UFC with no wins inside the distance. His bouts are usually close and tend to go the full 3 rounds, but the Pearce and Riley stoppages show he can be finished when hit cleanly. Fili has faced better opposition overall (Swanson, Delgado, Rodriguez, Ige), and his takedowns let him steal rounds Kamaka might otherwise win in the scrambles. Against that, Fili was knocked out more recently (106 days ago, against Kamaka's 85 days since the Riley KO), he is older, and he has now been stopped twice in his last 5 fights. With both fighters coming off knockouts and both prone to split decisions, this is close to a coin flip. I lean slightly to Fili on reach, wrestling and the quality of his recent opponents, and expect a close decision.

GPTFili is the pick at the normalized 59% market baseline; the supplied evidence does not justify moving away from it. His listed reach advantage supports the selection, but two consecutive losses, including a recent stoppage, temper confidence. Kamaka's debut flag contradicts the supplied UFC history, making his statistical averages unreliable for matchup analysis. Neither supplied layoff approaches the 450-day concern threshold. Opposition quality cannot be reliably separated from these records. Fili's recent wins came by split decision, favoring a competitive decision prediction. Method percentages are conditional on Fili winning.

CLAUDEThe market makes Wellmaker 64% to win, and I stay just below that because both sides of the sample are thin. Wellmaker is a switch-stance power puncher. Before his recent slide he stopped Bramhald, Saaiman and Moutinho with punches, and both UFC wins came in round 1. His two losses show where he is vulnerable: Ethyn Ewing beat him by decision over a full fight, and Juan Diaz submitted him with a rear-naked choke. So he struggles when opponents drag him into long or grappling-heavy fights. Tanzilovi fits that profile. The matchup tags him as a grappler, and his limited numbers show takedown volume at low accuracy. But his only two recorded pro-level fights are decision losses: a split decision to Josias Musasa on DWCS and a unanimous decision to Shane Collins in June 2026. Neither shows an ability to finish or to control strong opposition. UFCStats lists him with no completed UFC fights, so I don't treat his per-minute numbers as evidence either way. Wellmaker's move to American Top Team after back-to-back losses should help his takedown defense and scrambling, which are exactly the problems these losses exposed. That news came out in July, though, so the market has already priced it. Physically they are even: both are 5'10 and Wellmaker has a 1-inch reach edge. The most likely path is Wellmaker landing early before Tanzilovi can build up grappling exchanges. The main risk is that Tanzilovi survives the first round and repeats the Ewing or Diaz kind of fight. This involves a debutant and small samples, so confidence stays under the 65 cap.

GPTThe 66% de-vigged market favors Wellmaker; I trim that to 64% for debutant uncertainty and conflicting statistical coverage. His three documented knockout wins support a striking finish, but his subsequent decision and submission losses outweigh the stale undefeated statistical snapshot. Tanzilovi's debut flag makes his listed UFC averages unreliable, and neither fighter has an established opposition-quality advantage here. The supplied layoffs of 143 and 108 days are not material rust concerns. Wellmaker's July camp change offers no demonstrated improvement to price separately, and the current market already incorporates the two-point move. Method percentages are conditional on a Wellmaker win.

CLAUDEThe market makes Walker a -265 favourite (69%), while Elo has Meerschaert at 56%. That's a gap of about 25 points, which needs resolving. I side with the market and shade slightly below it. Walker's Elo is provisional, built on only 3 rated fights (1-2), so it carries little weight. Meerschaert's 1520 Elo mostly reflects a long, middling middleweight career and is still dropping (-45 recent trend). The Elo model can't see what matters most here: Meerschaert has lost 5 straight (Malkoun by decision, Daukaus by D'Arce choke, Oleksiejczuk by KO, Tavares by decision, de Ridder by arm triangle). Two of those were finishes inside the first round, and at 38 his recent form is clearly worse than his career numbers. The scouting report adds context. He missed weight for the Malkoun fight, which went ahead at 190 lb, and he is now moving up to 205. Ending the cut may help his cardio, but he gives up natural size and age to a 27-year-old who is a full light heavyweight. Walker is no safe bet either. He has been knocked out in his last 2 fights (Jacoby in round 2, Yakhyaev), and his UFCStats fields are blank, so I can't quantify his defence. Meerschaert's 1.4 submission attempts per 15 minutes and his willingness to pull guard or chase scrambles are a real upset path if Walker gets careless on the ground. Meerschaert's 42% takedown defence, though, means Walker can choose where the fight happens. The likeliest outcome is that Walker's size and power on the feet wear down a smaller, declining fighter whose chin has already broken in this run. I expect a mid-fight stoppage, with a decision as the main alternative. Confidence is set below the 69% market price because of Walker's own two KO losses and the submission risk.

GPTStarting from Walker's 71% market probability, I reduce confidence for his two recent knockout losses and the opponent-adjusted Elo favoring Meerschaert 56%. Walker's Elo is provisional and omits the latest listed loss, limiting its reliability. Meerschaert's five consecutive losses and declining Elo support siding with the market, although his submission activity presents a credible upset route. His announced move to light heavyweight adds uncertainty but is likely already priced in. Neither supplied layoff approaches the 450-day concern threshold. Walker's missing statistics and conflicting record summaries prevent claiming a proven striking or wrestling advantage. Decision is the cautious method pick given his documented decision victory and the limited evidence of finishing ability. Method percentages are conditional on Walker winning.

CLAUDEThis is close to a coin flip, and both priors lean slightly the same way: the market has Bonfim at 53% and Elo has him at 56% (1485 vs 1442). Neither rating is provisional. Bonfim has the cleaner striking numbers: 5.80 significant strikes landed per minute against 3.16 absorbed, 54% accuracy and 70% striking defense. Prado lands 4.58, absorbs 5.66, hits 41% and defends 52%. Prado has also lost four straight, all by decision, and has only one UFC win (1-5). The case against Bonfim is durability and recent damage. He was knocked out by Sadykhov (Feb 2025) and Padilla (Nov 2025), then caught in a D'Arce choke by Axel Sola on July 25. He also came in as a replacement for Artur Minev, about 10 weeks after that submission loss, and nobody has reported when he accepted the fight. That is a real but modest point against him, which is why I sit at the market's number rather than Elo's 56. Prado is 24, has never been finished in his recorded UFC losses, and is going back down to lightweight after three fights at welterweight, where he was undersized. He should be the more physical man at 155, though his takedown defense is poor (25%). Bonfim's losses all came against strong or rising opponents (Saint Denis, Sadykhov, Padilla, Sola). Prado's losses came against mid-tier welterweights and Zellhuber, and they were close. Prado's durability makes a Bonfim stoppage less likely, but Bonfim's chin keeps a Prado knockout alive. The likeliest single result is Bonfim winning a decision on accuracy and defense.

GPTStay with the market's narrow 51% Bonfim lean. Opponent-adjusted Elo supports him at 56%, and his striking accuracy and takedown threat offer advantages against Prado's negative striking differential and 25% takedown defense. However, Bonfim's three consecutive stoppage losses undermine his strong career defensive averages, while his replacement status adds preparation uncertainty. Prado has lost four straight decisions, supporting a distance outcome if Bonfim wins. Neither supplied layoff approaches the 450-day concern threshold. Prado's return to lightweight adds uncertainty without evidence of a troubled cut. Conflicting Bonfim records and duplicated entries limit confidence; the movement toward Prado is already reflected in the current market. Method percentages describe overall bout outcomes.

CLAUDEThe market has Shahbazyan at 57%, and I keep him about there. The case is mostly about Price's decline rather than proof of Shahbazyan's quality. Price turns 37 just before this fight, has lost four in a row, and was stopped in three of those: an RNC by Jacobe Smith, a punch-out by Veretennikov in round 1, and an RNC by Chiesa. Before that, Lawler knocked him out in round 1. He retired after the Chiesa loss and then reversed that decision for this one fight. That pattern usually points to declining durability, not a fighter at his peak. His career numbers (5.21 landed vs 5.58 absorbed, 48% defense, 52% takedown defense) were already negative, and the recent results look worse. Shahbazyan brings youth (30) and real height at 6'4, though his 74-inch reach is shorter than Price's 76. His record is a serious concern: a knockout loss to Phil Rowe on DWCS in 2019 and a one-punch knockout loss to Levan Chokheli in June 2026. UFCStats treats him as having no completed UFC fights, so I don't use his per-minute numbers as evidence. Both men can be hurt, and Price is still a dangerous, wild finisher, so a coin-flip slugfest is a real outcome. That keeps confidence modest. The most likely ending is a stoppage, more likely against the older and more worn-down Price, in the middle rounds.

GPTI favor Shahbazyan at the supplied 59% de-vigged market probability. The more current enrichment log shows Price losing four straight, including three consecutive stoppages, outweighing his career averages. However, Shahbazyan's two listed knockout losses give Price a credible upset path. Shahbazyan is flagged as a UFC debutant, so his supplied UFC-derived averages are unusable and uncertainty remains substantial. Opposition-quality evidence is insufficient to establish a clear advantage. Neither reported layoff approaches the 450-day risk threshold. Price's retirement return adds uncertainty, but the September news and two-point market move do not justify an additional adjustment. KO/TKO is a tentative method lean; method probabilities are conditional on Shahbazyan winning.

CLAUDEThe market has Franco at 73% (-340/+250), and the recent-form evidence backs that up. The stats block flags Franco as a UFC debutant, but that is a stale database entry: his form log shows two UFC bouts, a decision loss to Mario Pinto (March 2026) and a punches stoppage of Levi Rodrigues Jr. (July 2026). So I don't apply the 65 cap, but I still treat his 1-1 UFC record as a thin sample. Ribeiro has lost four straight. He absorbs more than he lands (2.94 vs 3.56 significant strikes per minute), lands only 40% of his strikes and stops only 32% of takedowns. He has three first-round KO/TKO losses (Zhang, Murzakanov, Sy) and was choked out by Yakhyaev. He is also a replacement for Alexander Poppeck, 77 days after his last fight. The notice date is undocumented, so that is only a small adjustment against him. Franco has an 11-2 record and leans toward grappling, so he can finish with his hands or by taking down someone with weak takedown defense. Ribeiro's live paths are a kimura in a scramble or using his 81-inch reach to win on points, and his recent results make both unlikely. I sit slightly under the market because Franco's UFC sample is small and the Pinto loss shows a durable opponent can take him to a decision.

GPTFranco starts as the 74% market favorite, but his explicit UFC-debut flag conflicts with the supplied record and makes his statistical averages unreliable; confidence is capped at 65%. Ribeiro's four consecutive losses include two knockout defeats and one submission, supporting vulnerability across finishing routes. Franco's reported recent knockout win supports the stoppage pick, although opposition quality is insufficiently established. Ribeiro's short-notice replacement status reinforces the lean but likely already factors into the market. Both fought recently, so extended layoff risk is absent. Method percentages are conditional on a Franco win.

CLAUDEThe market has Harris at 66% (-235). Harris is flagged as a debutant, but that is stale data: he debuted on 2026-07-18, replaced Frye Jr. against Alvin Hines on about a week's notice, and stopped Hines quickly with punches. He is 6-0 with every win a finish. Frye Jr. is 6-1, and his only UFC fight was a decision loss to Guilherme Pat in December 2025. He then withdrew from the July Hines bout for undisclosed reasons and arrives off a 295-day layoff. That is short of the 450-day danger zone, but an unexplained withdrawal plus inactivity is a real question mark. Frye's UFC per-minute stats are empty or zero, so I give them no weight either way. The base case is a heavyweight fight where Harris has proven fight-ending power and Frye has shown he can go three rounds. Frye's path is to survive early, use his 265 lb frame in the clinch and drag it to a decision, as Pat did to him. Both fighters have one-fight UFC samples and it is heavyweight, so I stay just under the market and under the 65 small-sample ceiling.

GPTHarris is the pick from the supplied 70% de-vigged market baseline. His reported knockout win supports the finishing prediction, but his debut flag contradicts that result; the conflicting records and sparse samples warrant capping confidence at 65%. Unreliable statistical fields provide no demonstrated stylistic advantage. Frye's listed 297-day layoff exceeds Harris's 80 days but remains below the approximately 450-day material-risk threshold. Opposition quality is unestablished, and Frye's earlier withdrawal does not establish a current injury. The market move is already priced in. Round timing is tentative; method percentages are conditional on a Harris win.

CLAUDEThere is no market line, so the priors are Elo and form. Elo has Zhelezniakova at 52%, but both ratings are provisional and Pereira's is stale. It lists her UFC record as 0-1 and misses her April 2026 knee KO of Hailey Cowan, so her true rating is likely around 1500, which makes this roughly a coin flip. The pipeline reports no common opponents, but there is one: Montse Rendon. Zhelezniakova beat Rendon by unanimous decision in March 2024; Pereira lost to her by split decision in September 2025. That slightly favors Daria, but the results are 18 months apart and Pereira was a 19-year-old in her first UFC fight. Zhelezniakova (4 UFC fights) outlands opponents by +0.75 strikes per minute and avoids 64% of strikes, but she has alternated wins and losses (L-W-L-W), lost her last fight to Melissa Croden, and was submitted by Ailin Perez with an arm triangle. Her 50% takedown defense is a leak if Pereira wrestles. Pereira is 20, 7-1, and coming off her first UFC win, a finish, which fits a fast-improving prospect. Daria is 30 and has shown no growth. I lean Pereira on trajectory, offset by the Rendon comparison and Daria's steady volume, so 52% honestly reflects a coin flip. A decision is the likely result: Daria has gone the distance in all three of her UFC decisions, and Pereira's UFC fights went one full distance and one finish.

GPTZhelezniakova narrowly gets the pick through her documented positive striking differential and decision-winning experience. She beat Montse Rendon, who beat Pereira, although that indirect comparison carries limited weight. Provisional Elo favors Zhelezniakova at 52%, but Pereira's listed April knockout win is missing from her statistical and Elo records. That finish strengthens Pereira's case and keeps this close. Pereira's missing statistics are not evidence of weak skills; conflicting records and duplicated entries further limit confidence. Listed layoffs of 185 and 171 days offer little separation. Method percentages are conditional on a Zhelezniakova win.

CLAUDEStart from the market: Gatto -162 / Kareckaite +126 works out to about 58/42 after removing the vig. Elo agrees in direction at 54/46 for Gatto, but Kareckaite's rating is provisional (2 rated fights), so it is a weak signal. Both priors favor Gatto by a small margin, and I see no strong reason to go against them. The best comparison point is a shared opponent, Dione Barbosa. Kareckaite lost to her by unanimous decision in May 2024. Gatto lost to her by majority decision in April 2026, so at least one judge had Gatto even with or ahead of Barbosa. That is a slightly better result against the same grinder and points slightly toward Gatto. On style, Kareckaite is the bigger fighter (5'9, 71-inch reach vs 5'5, 69-inch). She is a high-volume, high-exchange flyweight: 7.28 significant strikes landed per minute but 6.47 absorbed, with only 49% strike defense. Those averages come from a very small UFC sample, but they describe a fighter who wins on output and gets hit a lot in return. Gatto is the opposite: lower volume (3.78 landed per minute) but cleaner and more efficient. She has 49% accuracy, 54% defense, only 2.84 absorbed per minute and a positive differential of +0.94. She also has real stopping power, with three UFC KO/TKO wins (Leonardo, Eubanks, Vidal), and her last four finishes all came late. That makes Kareckaite's habit of absorbing strikes a concrete risk, especially in round 3. Kareckaite's main edge is defending takedowns (81%), which should cancel Gatto's inaccurate wrestling (1.47 takedowns per 15 minutes at 28%). That likely keeps the fight standing, where Kareckaite's reach and volume can win rounds on the scorecards. This is why Gatto is not a strong favorite. Neither fighter is in clear form. Kareckaite has gone L-W-L-W, with both wins by split decision and no finishes. Gatto is 1-3 in her last four, with the losses all close decisions against solid opposition (Cortez, da Silva, Barbosa). Neither layoff is a concern: 218 days for Kareckaite and 183 for Gatto. There is no fight-week news. Both fighters' last six or so UFC fights went to decision except Gatto's KO wins, so the most likely result is a close Gatto decision built on cleaner, harder strikes. A late stoppage of a fighter who absorbs this much is the second-most likely path. I am setting confidence slightly below the market (56 vs 58). Kareckaite's size and volume are real tools against a shorter opponent, and a close striking fight with output differences like these can easily go either way on the cards.

GPTGatto is a narrow pick at the supplied 53% market baseline, supported by Elo's 54% estimate. Her higher striking accuracy, lower absorption and broader grappling options offer a slight edge, but Kareckaite's striking volume and 81% takedown defense make a competitive decision likely. Both lost their latest listed bout; recent results do not support a strong form advantage. Reported layoffs of 185 and 220 days are below the material-risk threshold. Conflicting records, duplicated logs and Kareckaite's two-fight statistical sample limit confidence. Movement toward Kareckaite is already reflected in the current price, with no material news supporting another adjustment. Method percentages are conditional on a Gatto win.