A signing readiness benchmark for unsigned MMA fighters. Two pathways, two formulas, two datasets that are never combined: ONE Championship (built on 37 verified signings, 2023–present) and Dana White's Contender Series (built on 165 verified invitees across Seasons 9 and 10). Every input is observable. Every output reproduces from the published formula.
Enter a fighter's stats. Get their ASCENT score with every component broken out. The two pathways use different formulas, different weights, and different datasets — a ONE score and a DWCS score are not comparable numbers.
Enter fighter stats and hit calculate
to see the ONE ASCENT score.
Enter fighter stats and hit calculate
to see the DWCS ASCENT score.
Every component traces to something measured in the signing or invite data. Where a component is weak, it says so — the streak term scores below chance against the DWCS invite target, and it's flagged rather than quietly dropped.
The dominant predictor. ONE signs finishers — fighters who end fights by KO/TKO or submission. The formula normalizes finish rate against global division averages (FightMatrix, all promotions): a heavyweight finishing 80% of fights is average for the division; a strawweight doing it is exceptional. Normalization puts them on the same scale.
Winning matters, but less than how you win. A decision-heavy undefeated record gets credit, but not as much as a finish-heavy one. This separates "good record" from "exciting record" — and ONE skews toward exciting.
Recalibrated on 2023–present signings. ONE Friday Fights and LFA are the top pipelines at 12 pts each. Pancrase earns 6, ACA and Shooto 4. This is PromotionBase + an HQ region bonus — capturing both where fighters compete and where ONE's scouts actually sign from.
Based on each region's actual share of 2023–present signings. East Asia and South America each account for 15%+ of signings (+6). North America, Central Asia, the Middle East, and Eastern Europe sit at 8%+ (+4). Everywhere else gets +2. Real signing shares — not subjective "appeal."
Younger fighters have more upside and more potential fights on a contract. Under 24 gets +3. Prime window 24–27 gets +2. 28–30 neutral. 31–33 gets -2. Over 34 gets -4 — not because older fighters can't compete, but because promotions invest in longer timelines.
Winning championships proves you perform under pressure. The bonus is 3 + number of title fight wins (uncapped). 1 title win = +4. 4 title wins = +7. No titles = no bonus. This rewards proven ability at the highest level of your current promotion.
The strongest single component. Normalized against global division finish rates (FightMatrix, N=261,967 fights) — not against the invitee pool, which is finisher-selected and would push everyone toward the ceiling. A heavyweight finishing 80% is average for the division; a women's strawweight doing it is exceptional.
Winning matters but doesn't dominate. The old sub-80% penalty was removed in v2.8.5: win rate was already being counted four separate ways, and the step thresholds punished a fighter at 78.6% three points more than one at 80.0% for a 1.4-point difference.
A square-root curve — the first few consecutive wins carry the most weight, then returns diminish. Worth flagging honestly: against the invite target specifically, streak scores below chance (AUC 0.485). It earns its place against the contract target, not this one.
Each promotion's measured share of verified Season 9 invitees. LFA alone accounts for 21% (15 pts). Fury FC and Cage Warriors sit at 6% each (11 pts). Unlisted regional shows get 3, not 0 — a fighter nobody has seen is a longer shot, not an impossibility. At N=94, two invitees moves a promotion a whole tier; that boundary is noise and is documented as such.
Derived from the age distribution of actual invitees, not hand-set. Contender Series invitees cluster hard at 27–30 (64% of the field), so that band scores highest and fighters under 25 take a penalty. This is counterintuitive for a prospect board and worth understanding before you read the rankings: it measures who gets invited, not who has the most career ahead of them.
Title wins pay on a promotion-gated curve: at a ranked feeder, +5 for the belt, +5 for the first defense, +3 for each after. Stack bonuses add for undefeated (+5), elite normalized finish rate (+4), a perfect raw finish rate (+3), and a 7+ streak (+3). Nothing here is capped, and neither is the total — a score above 100 means several components are elite at once. That's the finding, not a bug.
Every number traces back to actual outcomes, cross-referenced against Tapology, Sherdog, and official promotion records.
No hidden weights. No black-box AI. No machine learning. ASCENT is a deterministic weighted formula — the same
inputs always produce the same output, and every component is observable from a public fight record.
The calculator on this page reproduces all 210 scored rows in the DWCS
master file exactly, and the ONE board is recomputed directly from the published scoring script, so any number here
can be checked against the formula.
The benchmarks: fighters ONE actually signed since 2023 average 80.9
(median 82.4) at the moment of signing; the 165 fighters invited to Contender Series Seasons 9 and 10 score a median of
76.3 under DWCS ASCENT v2.8.5.
What this does not claim. There is no control group of fighters who were passed over, so neither score
is a probability. They measure how closely an unsigned fighter resembles the people who actually got signed or invited.
The one prospective test so far has seven resolved outcomes — far too few to claim prediction. Season 10 results
land in October 2026 and will be the first real test.
Each public board shows the top 15. Behind them sit 194 scored unsigned prospects —
157 on the ONE pathway and 37 on DWCS, kept on separate formulas and never combined — with every component broken out
and updated as fighters win, lose, and get signed.
I'm not selling anything yet. I'm trying to find out whether this is actually useful to the people who have to decide
who to back. Tell me how you make that call today and I'll run your fighters through both boards and send you what comes out.