One subject, in depth · bridged to Sideline US
FightIndex
Middleweight MMA · most active 2020s

Roman Dolidze

76″ reach · +1.31% reach advantage · 19th percentile

MMA record

15–5

Roman Dolidze has a reach of 76″ against a height of 75″ — a reach advantage of +1.31%, unusually short for middleweight MMA and the 19th percentile of the 139 compared here.

20102020win v Alexander Kovbel, 2016, submissionwin v Amirali Zhoroev, 2017, kowin v Jumabek Ayjigit Uulu, 2017, submissionwin v Rémi Delcampe, 2017, submissionwin v Michał Pasternak, 2018, kowin v Eder de Souza, 2018, kowin v John Allan, 2020, decisionwin v Khadis Ibragimov, 2020, kowin v Laureano Staropoli, 2021, decisionloss v Trevin Giles, 2021, decisionwin v Jack Hermansson, 2022, kowin v Phil Hawes, 2022, kowin v Kyle Daukaus, 2022, koloss v Marvin Vettori, 2023, decisionwin v Kevin Holland, 2024, kowin v Anthony Smith, 2024, decisionloss v Nassourdine Imavov, 2024, decisionloss v Anthony Hernandez, 2025, submissionwin v Marvin Vettori, 2025, decisionloss v Christian Leroy Duncan, 2026, decision
WON ABOVE·LOST BELOW BAR = STOPPAGE·BROKEN BAR = SUBMISSION·TICK = DECISION 2016–2026

SOURCE·Roman Dolidze — English Wikipedia (CC BY-SA 4.0)·derived figures computed here·REVIEWED 2026-07-03

How this record is compiled

  1. Measurements come from the fighter's English Wikipedia infobox; the bout list is parsed from the results table on the same article, read on 2026-07-03. Boxing tables are read by header name and MMA records by the fixed column order of the MMA record template — two different parsers, because reading MMA rows as a boxing table returns a fighter's boxing record in place of their real one.
  2. A bout enters the ledger with whatever fields the source states: result, opponent, method, round, year. Any field the source leaves empty stays empty here. A date that cannot be established is left blank rather than inferred, and the bout is still counted.
  3. A no contest is counted and printed as a no contest. It is never folded into the wins, the losses or the draws, and it never disappears to make a total balance.
  4. Where the article's infobox total disagrees with its own results table, both figures are published and each is labelled with where it came from. Nothing is averaged and nothing is reconciled against a second database.
  5. Percentiles rank this fighter among the 139 middleweight MMA fighters in the dataset — the most specific group of at least 30 that this fighter belongs to.
reach advantage = (reach − height) ÷ height × 100

Sources, exclusions and known limits →

Bout record

Every bout the article's own results table lists, in order. A dash is a field the source leaves empty — it is not a zero and it is not a guess.

Bout-by-bout recordAUG 2026
YearResultOpponentMethodRd
2026 L Christian Leroy Duncan Decision3
2025 L Anthony Hernandez Submission4
2025 W Marvin Vettori Decision5
2024 W Kevin Holland KO/TKO1
2024 W Anthony Smith Decision3
2024 L Nassourdine Imavov Decision5
2023 L Marvin Vettori Decision3
2022 W Jack Hermansson KO/TKO2
2022 W Phil Hawes KO/TKO1
2022 W Kyle Daukaus KO/TKO1
2021 W Laureano Staropoli Decision3
2021 L Trevin Giles Decision3
2020 W John Allan Decision3
2020 W Khadis Ibragimov KO/TKO1
2018 W Michał Pasternak KO/TKO3
2018 W Eder de Souza KO/TKO2
2017 W Amirali Zhoroev KO/TKO1
2017 W Jumabek Ayjigit Uulu Submission1
2017 W Rémi Delcampe Submission1
2016 W Alexander Kovbel Submission1

13 of these opponents have their own page here.

SOURCE·Roman Dolidze — English Wikipedia (CC BY-SA 4.0)·derived figures computed here·REVIEWED 2026-07-03

Reach in context

Reach advantage

+1.31%

76″ reach on a 75″ frame

Compared against 139 middleweight MMA fighters. Median 2.79%.

-2.9%middleweight MMA reach advantage11.1%
Measurements
Reach76″ (193 cm)65th pct
Height75″ (190.5 cm)88th pct
StanceOrthodox
DivisionMiddleweight MMA
competed at 3: heavyweight, light heavyweight, middleweight — compared here against the last
Most active decade2020s — 14 of 20 dated bouts
CountryGeorgia

Record

MMA record

15–5

20 bouts on record

Finish rate

73%

of wins ended inside the distance — middleweight MMA median 78%, 38th percentile

KO/TKO: 8 (53%)Submission: 3 (20%)Decision: 4 (27%)
  • KO/TKO 53%
  • Submission 20%
  • Decision 27%

The career, bout by bout

20 recorded bouts from 2016 to 2026. Every incumbent shows this as a table you have to read; here it is as a shape you can see.

2020 win vs Alexander Kovbel (2016)win vs Amirali Zhoroev (2017)win vs Jumabek Ayjigit Uulu (2017)win vs Rémi Delcampe (2017)win vs Michał Pasternak (2018)win vs Eder de Souza (2018)win vs John Allan (2020)win vs Khadis Ibragimov (2020)win vs Laureano Staropoli (2021)loss vs Trevin Giles (2021)win vs Jack Hermansson (2022)win vs Phil Hawes (2022)win vs Kyle Daukaus (2022)loss vs Marvin Vettori (2023)win vs Kevin Holland (2024)win vs Anthony Smith (2024)loss vs Nassourdine Imavov (2024)loss vs Anthony Hernandez (2025)win vs Marvin Vettori (2025)loss vs Christian Leroy Duncan (2026)
winlossdraw / NC2016–2026

Career span

11y

2016–2026

Longest unbeaten run

9

without a loss — draws do not break it

Busiest year

2017

3 bouts

Recorded bouts

20

20 with a date

Standing questions

What this record does and does not say

What is Roman Dolidze's record?
15–5 in MMA. That is the aggregate stated in the English Wikipedia infobox; the results table on the same article lists 20 bouts individually.
Does a no contest change a record here?
No. A no contest is not a win, a loss or a draw, and it is never folded into one. It is counted and printed separately, so a 20–0 record with two no contests is shown as exactly that rather than as 22–0 or as 20–0 with the no contests dropped. The same applies to a bout overturned after the fact where the source records the change.
What happens when sources disagree about a fight?
This site publishes one source — the fighter's English Wikipedia article, read on the date printed under every table — and it does not silently reconcile that against anyone else's database. Where the article's own infobox total and its own results table disagree, both are shown and the page says which one each figure came from. Nothing is averaged, and no field is filled in from a second source to make a total balance.

See what every site in this family answers

Published by

Amit Sharma

Auckland real-estate agent and marketer. Builds free, sourced data tools.