MMA judging criteria explained, and why rounds go wrong

Three judges watch the same five minutes from three sides of a cage and hand in three different scores, and the argument that follows is almost always about the same thing. One fighter was on top. The other was landing. There is no line in the rulebook that resolves that, because the MMA judging criteria are a boxing instrument fitted to a sport with three separate ways to win a round, and the mismatch explained here is where most scoring controversy actually comes from.

Boxing has one currency. Punches land or they do not, and every judge is counting the same kind of event. Mixed martial arts asks a judge to compare a knee to the body with a sweep to mount, and then to compress the comparison into a single digit that can only be a nine, occasionally an eight, and almost never anything else.

That is the design problem. Everything else follows from it.

What the ten point must system was built for

The ten point must is a boxing system, and it is a good one for boxing. The winner of a round gets ten points, always, with no discretion. The only decision a judge makes about the number is what the loser receives.

Nine in the ordinary case. Eight where the round was won by a wide margin. Seven where the margin was extraordinary. Ten where the judge genuinely cannot separate the fighters, which the criteria describe as a score generally reserved for partial rounds and treat, in a completed round, as a failure to reach a verdict at all.

Mixed martial arts adopted this wholesale when it sought regulation, and the reason was pragmatic rather than technical. Athletic commissions already knew how to run boxing. A sport asking to be sanctioned by those commissions was in a far better position offering a scoring method the officials already used than proposing something new. The ten point must in its original setting is a compression system: it throws away the difference between edging a round and winning it clearly, in exchange for producing a clean total.

Boxing can afford that. Its rounds are three minutes long, there are ten or twelve of them, and the range of things that can happen inside one is narrow.

MMA cannot afford it nearly as easily. A championship round is five minutes. A title fight is five rounds. That means twenty five minutes of a bout is divided into five verdicts rather than twelve, so each round is a fifth of the fight rather than a twelfth, and a single disputed round moves the result by twice as much.

• 3Judges scoring a bout
• 10Points to the winner of a round
• 5Minutes in a round
• 5Rounds in a championship bout

Round lengths and counts under the unified rules. Non-title bouts are three rounds; championship and designated main event bouts are five.

Then add the second problem. Boxing's judges are watching one activity. MMA's judges are watching a striking contest, a wrestling contest and a submission contest, which can run simultaneously, and which produce evidence that does not convert into a common unit. Nobody has ever written down how many head kicks a reversal is worth, because nobody can.

The MMA judging criteria, in the order judges must apply them

The criteria are published by the Association of Boxing Commissions and Combative Sports and adopted, with local amendments, by the state and national commissions that regulate the sport. They are short. Their brevity is deliberate and it is also the source of the arguments.

Effective striking and grappling comes first, and it decides the overwhelming majority of rounds. Successful striking means legal strikes that result in damage. Successful grappling means grappling that has been proven productive by the damage resulting from takedowns, reversals, submission attempts and the achievement of advantageous positions. The two are explicitly equal. A judge is not permitted to prefer punches on the basis that punches are easier to see, and a judge is not permitted to prefer wrestling on the basis that wrestling looks like control.

Aggressiveness comes second, defined as consistent offensive effort. Chasing an opponent around the cage with nothing landing is not aggressiveness in the scoring sense, and the criteria have said so in one form or another since 2016. The adjective was doing all the work then and the definition is doing it now.

Fighting area control comes third, defined as dictating the pace, place and position at which the bout is contested. This is the criterion people mean when they say "octagon control" on a broadcast, and it is very nearly a decorative category, because it is only supposed to be reached when the fighters cannot be separated on anything else.

The order is not a preference. It is a gate. Under the 2016 criteria, a judge could not move from one criterion to the next unless the prior one was completely even, and the document says so in capital letters. Aggressiveness was described as Plan B. Cage control was Plan C, to be used in what the committee expected would be an extremely rare occurrence.

That tiering was loosened in the version approved in 2025. The current text says that where a judge absolutely cannot determine even a marginal advantage in striking or grappling, the judge defers to aggressiveness or fighting area control, whichever impacted the round in a more significant manner. The strict two-step ladder has become a single tiebreaker tier with two things on it, and the judge picks whichever mattered more.

That is a small edit with a real consequence. It removes the awkward position where a judge saw a clear aggressiveness advantage and a clear control advantage pointing at different fighters, and had to take the aggressiveness one simply because it sat higher on the list.

1. 1The round startsThree judges watch from three fixed positions around the cage. They do not confer, they cannot see each other's scores, and each round is scored on its own merits with nothing carried forward from earlier rounds.
2. 2Damage is assessed firstKnockdowns, heavy legal strikes to vital targets, strikes that force an opponent into pure defence, visible injury, joint attacks and chokes causing visible distress. Damage is the direct result of effective fighting rather than an action in itself.
3. 3Dominance and duration are read against itDominance means sustained supremacy of position, action or volume that forces an opponent to defend. Duration means how much of the round one fighter spent imposing it. Both point towards damage rather than substituting for it.
4. 4Striking and grappling are compared as equalsSuccessful striking and successful grappling are treated as the same kind of evidence, because both are measured by what resulted rather than by what was attempted.
5. 5Only if that is genuinely level does the judge move onAggressiveness, meaning consistent offensive effort, or fighting area control, meaning dictating pace, place and position. The judge takes whichever affected the round more.
6. 6Defence is not scored at allThe criteria state plainly that there is no scoring for defence. Good defence keeps a fighter in the contest. It does not win a round.
7. 7The winner is fixed at tenThis is not a decision. The remaining question is whether the loser receives nine, eight, seven, or ten in a genuinely even round.
8. 8Any referee penalty is subtracted afterwardsThe round is scored on merit, then the deduction is applied, which is how a round scored 10-9 arrives at the commission as 9-9.
9. 9The card is collected and totalledScores go to the commission between rounds and cannot be revised. At the final horn the three totals are added independently and read out.

The sequence set out in the ABC criteria and the unified rules. Steps four to six are where almost all public disagreement about a decision actually lives.

Damage became the primary word, and it corrected a real distortion

The most consequential change to MMA scoring in a decade is a vocabulary change, and it was approved in August 2025.

The criteria now define three interrelated constructs of fighter effectiveness and put them in order. Damage. Dominance. Duration. The document states that damage is the most highly valued component in judging a round, because it is not an action but the direct result of effective fighting, and gives priority to immediate damage over cumulative damage within a round.

Damage is then defined with a list rather than left to instinct: knockdowns from legal strikes; heavy or concussive legal strikes to vital targets; legal strikes that force an opponent to retreat, readjust or default to defence exclusively; striking or grappling that leads to visible injury such as swelling, hematomas, lacerations or bruising; grappling and submission attempts that hyperextend joints or cause rotational damage; and chokes that cause visible distress.

Read that list again and notice what it excludes. Nothing in it is a count. There is no line rewarding output, no line rewarding position, and no line rewarding effort.

The problem it was written to fix is volume scoring. A fighter who throws a great many light strikes at the arms and gloves generates an enormous number on a broadcast graphic, and broadcast graphics are the numbers the audience sees, so they leak into the judging conversation whether officials read them or not. A fighter landing fewer but heavier shots generates a smaller number and a visibly damaged opponent. Under a criteria set that emphasised counting, the first fighter had a case. Under a criteria set that says damage is the result being measured, they have much less of one.

The same correction cuts the other way in grappling. A fighter who lands a takedown and then does nothing has performed an action. Under the current wording, an action that produces no result is not what is being scored.

Naming damage explicitly also does something subtler, which is that it gives the losing side of an argument something to argue with. A judge who scores a round for the fighter who was visibly hurting the other one is now pointing at a defined term in a published document rather than at a feeling about who looked better. That is not a small gain for a sport whose officials are asked to justify nothing and whose decisions can be reviewed by nobody.

None of which means the disagreements have stopped. Damage is a result, and results are not always visible from a judge's seat within five minutes. A body shot that quietly removes a fighter's legs looks like nothing until round four. Cumulative damage is now explicitly recognised, and it is also the hardest form of damage to see while it is accumulating.

The takedown that leads nowhere is the single most argued point in the sport

If you wanted one question to explain most MMA scoring disputes, it is this: what is a takedown worth when nothing happens afterwards?

The rulebook's answer is clear and has been for years. A successful takedown is not merely a change of position. The 2016 criteria said explicitly that a takedown scores by establishing an attack from the use of the takedown, and that fighters in top and bottom positions are assessed on the effective result of their actions rather than on their position. The 2025 text goes further by defining successful grappling in terms of the damage that resulted.

So the written answer is: a takedown that leads nowhere is worth very little.

The practical answer has often been different, and everyone in the sport knows it. Takedowns are legible. They are loud, the crowd reacts, the broadcast counts them, and they visibly interrupt whatever the other fighter was doing. A judge watching a round in real time, without replays, without a scorecard of landed strikes and without the ability to rewind, is a human being watching the most conspicuous event in the round.

There is a second reason, and it is more sympathetic than it sounds. A takedown that leads nowhere still stopped something. If a striker was building a rhythm and gets put on his back three times, the round he was going to win did not happen. That is not damage and it is not dominance, but it is not nothing either, and a judge who feels that has a genuine observation with no place to record it.

Consider an invented round, constructed for the argument rather than drawn from any bout. Fighter A lands two takedowns, is stood up twice by the referee within twenty seconds each time, and attempts nothing on the ground. Fighter B lands a dozen strikes across the round, three of them heavy, and opens a cut. Under the criteria as written, the round belongs to Fighter B, comfortably, because damage is the measure and Fighter A produced none. Under the habits the sport spent fifteen years building, that round is scored for Fighter A on a great many cards.

The gap between those two answers is the reform the 2025 language is trying to close, and closing it depends entirely on whether judges are trained and assigned in a way that rewards following the text.

The other half of the answer is that the fighter on the bottom is not scoring by being there either. A guard that produces sweeps, submission attempts and strikes is scoring. A guard that produces a comfortable place to lie down is not. The rules are symmetrical about this even though the audience rarely is, and the fastest way to see it is to watch how submission attempts change the value of a position rather than counting who is on top.

Control time is evidence, not a score

Control time appears on every broadcast graphic and in no rulebook.

It is a statistic compiled by the promotion or a data provider, measuring the seconds one fighter spent in a controlling position. It is genuinely useful information. It is not a criterion, and treating it as one is the most common analytical mistake made about MMA scoring.

The criteria are explicit. Merely holding a dominant position does not by itself demonstrate dominance in a round. Dominance means forcing an opponent to defend continually with no counters, staking a dominant position and continuously threatening fight-ending attacks from it, or maintaining an overwhelming offensive pace that keeps the other fighter reacting.

The word doing the work is "continuously". Position plus threat is dominance. Position plus stillness is a statistic.

The right way to read control time is as evidence about what a fighter was able to do, not as the thing being scored. Four minutes of control with a stream of strikes and two submission attempts is overwhelming evidence of effective grappling. Four minutes of control with nothing attempted is evidence that one fighter neutralised the other, which is worth something under aggressiveness or fighting area control, and worth very little under the criterion that decides most rounds.

A useful test: strip the position out and ask what happened. If the answer contains verbs, the control scored. If the answer is a location, it did not.

This distinction also explains a class of decision the audience finds baffling, where a fighter who was on top for most of the fight loses. Where the judges have applied the criteria as written, that is the system working. Where the judges have gone the other way and rewarded the position, the audience is furious for the opposite reason. Both complaints get called a robbery, and only one of them is a departure from the rules.

The 10-8 round, and the guidance that judges give too few of them

The most common structural criticism of MMA judging from inside the sport is not that judges pick the wrong winner. It is that they do not use the whole scale.

Almost every round in MMA is scored 10-9. That is what the system expects, but the criteria have been pushing officials towards more 10-8 rounds for a decade, on the grounds that a system with one setting for "won" cannot distinguish a round won by a whisker from a round where a fighter was nearly finished.

The 2025 text tightened what a 10-8 requires and, in doing so, changed which rounds qualify. A round is scored 10-8 when the winner earns it through offensive actions that may include dominance with duration, but which must include significant damage resulting from effort or attempts that could by their nature finish the fight.

Four points sit underneath that, and each of them settles an old argument.

Cumulative damage counts as significant damage where the losing fighter is visibly losing energy, confidence, stamina or spirit. Significant damage is by itself enough to award a 10-8, with no requirement for five minutes of one-way traffic. Significant domination together with some damage, against a fighter offering little or no offence, may be enough. And significant domination such as positional control without damage is not enough.

That last line is the change. Under the previous guidance a judge was told to consider a 10-8 where a fighter showed dominance even though no impactful scoring was achieved. Under the current guidance, dominance and duration without damage do not get there. The scale has been re-pointed at what a fighter did to the other fighter rather than at how long they did it for.

Why does any of this matter arithmetically? Because in a three-round fight, a single 10-8 is the difference between a decision and a draw.

Take an invented three-round bout. Fighter A wins rounds one and two 10-9. Fighter B wins round three, and hurts A badly enough that the judge scores it 10-8. Fighter A: ten, ten, eight, which is twenty eight. Fighter B: nine, nine, ten, which is also twenty eight. A fighter who won two of the three rounds does not win the fight.

Now stretch it to five rounds, which is where the effect becomes strange. Fighter A takes the first three rounds 10-9 and is thirty points to twenty seven up. Fighter B takes round four and hurts A enough for a 10-8, then wins round five 10-9.

• Fighter A
• Fighter B

A constructed example, not a real bout. Fighter A wins rounds one to three 10-9. Fighter B wins round four 10-8 and round five 10-9. The single 10-8 turns a three rounds to two win into a draw at 47-47.

Show the numbers

Invented five round bout, running totals after each round

Item	Fighter A	Fighter B
R1	10	9
R2	20	18
R3	30	27
R4	38	37
R5	47	47

Forty seven each. A fighter who won three rounds out of five gets a draw, and every score on that card was correct under the criteria.

That is the honest case for using 10-8 more often. It is also the honest case against, because the same arithmetic means that one judge's decision about one round can hand a title fight to the fighter who lost the majority of it. A scale used inconsistently by three officials produces more variance, not less, and the sport has not yet resolved which of those it fears more.

The invented cards below make the same point from the other direction, where the only disagreement between three officials is whether round four was a 10-8.

• Fighter A
• Fighter B

Judge one4747

Judge two4847

Judge three4648

A constructed set of cards. All three officials agree on rounds one, two and five. Judge one and judge three score round four 10-8 for Fighter B; judge two scores it 10-9. Judge three also gives round three to Fighter B. One card each and one draw produces a split draw.

Show the numbers

Invented bout, one round judged three ways

Item	Fighter A	Fighter B
Judge one	47	47
Judge two	48	47
Judge three	46	48

10-7 and 10-10, the two scores nobody wants to write down

A round is scored 10-7 when one fighter overwhelms the other in both damage and domination throughout the round. The older text put it more vividly: a 10-7 is a round where the judge is considering whether the fight should have been stopped.

It is genuinely rare, and it should be, because a round that bad usually ends before the horn. A referee whose job is to stop the contest and a judge writing 10-7 are looking at the same evidence and reaching different conclusions about whether it warranted intervention, which is an uncomfortable position for both of them.

10-10 is the opposite embarrassment. The criteria describe it as the score for a round where neither fighter established an advantage, note that it is generally reserved for partial rounds, and then say something quite blunt: scoring a completed round 10-10 is seen as a failure to adjudicate a score.

The reason the score exists at all is administrative. Rounds sometimes end early because of an accidental foul, and a judge who saw fifteen seconds of nothing needs a number for it. Its use in a full five minute round is treated as an admission that the official could not tell, and officials know they will be asked about it.

There is a defensible objection to that culture. Some rounds really are even. A system that instructs judges to invent a winner in a genuinely level round manufactures a point of difference where none existed, and three officials manufacturing separately will manufacture in different directions. The sport has decided it would rather have arguments than draws, which is the same trade boxing made.

Fouls and deductions arrive at the card after the round is scored

The referee, not the judges, rules on fouls. A judge receives the ruling and applies it.

The sequence is a verbal instruction, then a formal warning, then a point deduction, and nothing obliges the referee to go through the stages in order. A deliberate and serious foul can cost a point immediately. A repeated accidental one may draw several warnings and no deduction. Two referees can officiate identical rounds and produce different numbers of deductions while both stay inside their authority.

The mechanics on the card are what matter for scoring. The judge scores the round on merit, then subtracts. A round scored 10-9 becomes 9-9 if the winner of the round loses a point, and 10-8 if the loser of the round does. In a three round fight, a single deduction is worth as much as a knockdown, which is a startling amount of leverage to hand to a real-time judgement about intent.

The foul list itself is long and covers eye gouging, biting, hair pulling, groin strikes, small joint manipulation, strikes to the back of the head and spine, throat strikes, grabbing the fence, holding the shorts or gloves, timidity, and attacks on an opponent under the referee's care. Some entries are about safety and some are about keeping the contest a contest.

Two of them have moved recently, and both are worth knowing because they change what a judge is watching.

The ban on the downward, twelve-to-six elbow strike was removed from the unified rules by the ABC in 2024, effective from the November of that year, subject as always to adoption by each individual commission. It had been criticised for years as a rule that asked a referee to judge the clock face of a moving elbow in real time, with a safety rationale that few people could defend.

The definition of a grounded fighter was changed in the same round of amendments, to a fighter with any part of the body other than the hands or feet on the canvas. That matters enormously to scoring, because whether a fighter is grounded determines whether a knee or kick to the head is a legal strike that counts as damage or a foul that costs a point. Same technique, same contact, opposite entries on the card.

Adoption is the caveat on both. The ABC writes the rules; commissions enact them, and they do not all do it on the same day. The rule set governing a bout is the one adopted where the bout takes place, which is also why anyone comparing scoring across jurisdictions has to check what version was in force. The full rule set that surrounds the scoring is administered the same way, one commission at a time.

Three rounds against five changes what a losing fighter should do

A fighter behind on the cards faces a completely different problem depending on how long the fight is, and the difference is larger than the extra ten minutes suggests.

In a three round fight, losing the first round means needing both remaining rounds. Losing the first two means needing a finish, because no combination of scores in the third can rescue the fight unless the third is a 10-8 and the earlier rounds were close enough to have gone either way on other cards. A three round fight is therefore front-loaded in a way that punishes slow starters severely, and it rewards a fighter who takes risks early far more than the audience notices.

In a five round fight, a fighter can lose the opening round and still win three of the remaining four. The fight has room for a plan that develops. Body work, leg kicks and the slow accumulation of damage that the criteria now explicitly recognise all have time to pay, because the rounds in which they pay are still to come.

There is a second, less obvious effect. Because judges score rounds independently and carry nothing forward, damage inflicted in round one only affects round three's score to the extent that it visibly affects round three. A fighter who spends fifteen minutes wrecking an opponent's lead leg gets nothing on the cards for that investment until the leg starts failing. In a three round fight, that investment often never matures. In a five round fight it does, which is one reason five round bouts are won by different kinds of fighter.

The instruction a corner gives between rounds turns on an estimate of the cards that nobody in the cage can verify. MMA is scored blind. No running total is announced, the corner has no official information, and the difference between "you need this round" and "stay behind the jab" is the difference between a fighter taking a risk and a fighter protecting a lead that may not exist.

Get the estimate wrong in the wrong direction and a fighter coasts through a final round they had to win. That mistake costs titles more often than any judge does, and it is invisible afterwards because the scorecards do not record what the corner believed.

Where the judges sit, and why that matters more in a cage than in a ring

A boxing judge has a bad view. An MMA judge has a worse one, for reasons that have nothing to do with competence.

Judges sit at cage level on separate sides. The cage itself is fencing, and fencing is not transparent at an angle. A fighter pressed against it is partly obscured to the two judges on either side of that section and clearly visible to the one opposite. Ground exchanges happen below the level of the fence padding, on a canvas raised above the judge's eyeline, with two bodies overlapping.

The consequence is that the sport's primary scoring criterion, damage, is precisely the thing an obstructed angle destroys. A short elbow landing from top position is visible from one seat and invisible from another. A judge who did not see the strike is not scoring the strike, and no amount of training changes that.

Some of the disagreement between three cards is therefore not disagreement at all. It is three honest reports of three different sets of visible events, and the only real fix is technological rather than educational. Monitors at the judges' positions have been trialled and argued about for years. The objection is that a judge watching a screen stops watching the cage and starts watching a director's choice of shot, which substitutes one incomplete view for another.

Nothing here requires anybody to be corrupt or lazy. It requires only that a cage is an opaque object with people moving inside it.

Commissions and promotions do not all do this the same way

The unified rules are unified in the sense that most jurisdictions have adopted something close to them. They are not a single global instrument.

Each commission adopts its own version, on its own timetable, with its own amendments. Regulations in one state may run a round of amendments ahead of the neighbouring state. A promotion running an event in a jurisdiction with no athletic commission at all operates under whatever regulatory arrangement is available there, which in practice often means a body imported for the event, and the rule set can differ in details that reach the cards.

Judges are assigned by the commission, not the promotion, which is a genuine structural protection and a better arrangement than boxing's, where the sanctioning body with a financial interest in the title has a say in who officiates. It is not a complete protection. The pool of qualified MMA officials is small, assignments are a livelihood, and there is no published mechanism through which a judge whose cards are consistently out of line with the criteria is removed from big fights.

There is also no appeal that works. A commission can review a decision, and in a handful of cases has changed one where a scoring error was arithmetical rather than a matter of judgement. Disagreeing with how a round was assessed is not grounds for anything. A card that everyone in the building thinks is wrong stands, and the only consequence is reputational, and reputation is not observable from outside.

The effect on careers is real even when the decisions are defensible. A run of narrow losses reads as evidence about a fighter's level to the people booking fights, and how opponents get matched is a business calculation that treats a decision loss and a knockout loss as different things. Two rounds' worth of legitimate disagreement can redirect several years of work.

What the MMA judging criteria cannot record, and the case against rounds

Here is the structural criticism, and it is stronger than the usual complaints about individual officials.

MMA is a sport in which a fight can turn completely in one second. A single counter changes everything about who was winning and everything about who is capable of continuing. A round-based system records that second as one point, the same point it awards for winning a five minute round by landing four more jabs.

Round scoring was designed for a sport of accumulation. Boxing rounds accumulate: a fighter builds a lead over three minutes and the total reflects work done. MMA rounds do not reliably accumulate, because the sport contains events that are not increments. A submission attempt that came within a degree of finishing the fight is worth, on the card, whatever a judge decides to weigh it as inside a round they also have to weigh everything else in.

Then there is independence. Every round is scored fresh, and nothing carries forward. The reasoning is sound and it is the same reasoning boxing uses: a judge permitted to score holistically will decide who is winning and then find reasons in each round to confirm it. But the cost in a five round fight is that a fighter who was nearly finished in round two arrives at round three with a clean slate on the cards and a damaged body in reality. The scorecards and the fight have diverged, and only one of them is what the audience watched.

Alternatives get proposed regularly. Scoring the fight as a whole. Half-point scoring, which would let a judge record a round won clearly without going all the way to 10-8. Open scoring, so corners know where they stand. Each fixes something and breaks something else. Half-points add resolution and add a new argument about where the halves go. Whole-fight scoring removes the round problem and reintroduces exactly the bias round independence exists to prevent.

The honest position is that no scoring system converts a fight into a number without losing information, and the question is only which information a sport is willing to lose. MMA has chosen to lose the difference between a round won narrowly and a round won clearly, and to lose the memory of what happened two rounds ago. Those are the two losses that generate almost every disputed decision.

A related point about how much of a fight is decided before it starts: the physical state a fighter arrives in has already shaped what the judges will see, and the arithmetic of making the division limit affects the rounds in which cards are usually decided more than any criterion does.

Scoring a round yourself, properly

The useful end of this is that you can do it, and doing it changes what you watch.

Score as the horn sounds and do not revise. The moment you rewrite round two after seeing round four, you have started scoring the fight rather than the round, which is the exact habit the system exists to prevent.

Ask the questions in the order the criteria set out.

Who did damage? Not who threw more, not who looked busier, not who was on top. Who landed the things that lessened the other fighter's capacity or willingness to keep going: knockdowns, heavy shots to real targets, strikes that forced pure defence, visible injury, joint attacks and chokes that clearly distressed. If that separates them, the round is decided and nothing else matters.

Who was dominant, and for how long? Only if damage was level. Dominance means the other fighter was forced to defend continuously with no counters, or was under sustained threat from a position, or could not get out from under the pace. Holding a position is not dominance. Duration only means anything attached to something being done.

Was it genuinely, completely even? Only then do aggressiveness and fighting area control come into it, and you take whichever affected the round more rather than working down a ladder.

Then set the number. Ten to the winner. Nine unless there was significant damage, in which case eight, and do not talk yourself out of the eight because eights feel dramatic. Seven if one fighter was overwhelmed in both damage and dominance for the whole round. Ten only if you genuinely cannot separate them, and be honest that this is usually a decision you are avoiding.

Then subtract any point the referee took, from whichever fighter it was taken from, after you have scored the round on merit.

Do that for a card's worth of fights and two things happen. You will agree with the officials far more often than the internet does, and your disagreements will cluster in a recognisable place: rounds where one fighter did the damage and the other did the controlling. That is the seam in the sport's scoring, it has been there since the ten point must was borrowed, and no amount of rewriting the criteria closes it while a round remains the unit of account. The rest of the archive on fighting, its rules and its arguments sits in the combat sports section.