How skilled can one human swordsperson be, in terms of technique? A metric for fighting skill would be having them duel 1000 people, with a break for recovery after each fight. Could they win all the duels? There are a limited number of ways and directions to dodge an attack
As a martial artist, studying karate and hapkido, I have fought higher-grade martial artists on many occasions. I have also trained in western fencing.
Given my training in these armed and unarmed martial arts, I can state that the difference between a martial artist with low-level skills and one who has high-level skills is a matter of probability. As the skill gap increases, the probability becomes higher that the more skilled martial artist will hit the less skilled martial artist, and the probability becomes lower that the less skilled martial artist will hit the more skilled martial artist. As a case in point, as a beginner, I have occasionally scored hits against my teachers, who were greatly more skilled than myself, in both armed and unarmed martial arts.
When dealing with lethal weapons such as a sword, the probability becomes high that the more skilled martial artist will inflict a serious or greater injury upon their less skilled opponent before receiving an injury themselves.
However, when dealing with probability, unlikely events become more probable that they will occur at least once when the test is repeated many times. For example, if there is a 1% chance that the more skilled martial artist will be hit during any one of their thousand duels, there is a 99% chance of not being hit. To calculate the odds of not being hit in any of a number of bouts, we multiply the odds of not being hit in each bout. If we raise 0.99 to the 1000th power, that will give the probability of not being hit during the thousand duels, assuming that the odds are the same for each bout:
So, it is almost a certainty that the skilled martial artist will be injured during these thousand duels... perhaps seriously or even fatally, even though they may leave behind them a horde of defeated opponents.
So... the answer to this question is almost certainly No... this sword master is highly unlikely to be able to defeat all of their thousand opponents, no matter how likely they are to defeat each opponent individually, since the odds of their victory in any bout will never amount to a certainty. The laws of probability are heavily against them.
Here's my perspective as another martial artist (Hapkido, Kali, HEMA) and as a writer, to speak to a couple aspects of this question that haven't been adressed. @Monty Wild and @jamesqf raise good points about the numbers: Winning a thousand duels in a row is both unlikely in terms of probability and unfeasible in terms of human lifespan. For a swordsperson with a a 20 year career as a "master," that averages to about one duel a week, every week, no breaks, no time off for injuries. For 20 years straight, which would be an incredibly long period of time to be at the top of one's game, period.
The points you raise about a "limited number of ways to dodge an attack" indicates you're thinking in terms of movie fighting, which isn't wrong, as a storyteller, but it does fall into some common traps. A common adage* in martial arts is, "Fear not the fighter who has practiced a thousand kicks ten times, but the fighter who has practiced one kick ten-thousand times!" (*or some variation thereupon). Your swordsperson doesn't need to know a million fancy moves, and it doesn't matter if they have used a strategy to defeat a previous opponent. A good strategy works multiple times, and in a variety of situations. As an example, I'll share an anecdote from a Filipino swordfighting teacher. The Philippines has a rich history of blade combat, and one tradition in duels was for each fighter to step forward and show off, to scare the other fighter into surrendering. One famous master was challenged, and so his challenger stepped forth and whipped his sword around in a blaze of twisting angles and flourishes. The master then stepped up, but all he did was one simple, sharp downward slash. The challenger wasn't too impressed, so the fight began. The challenger lunged in, and immediately the master did his sharp downward slash and sliced the other fighter's thumb clean off. I don't know if you know this, but you can't hold anything without your thumb. The fighter couldn't keep hold of his sword, and with that one strike the fight was over. Now, I don't know how many fights this master won over his lifetime—it probably wasn't as high as a thousand—but he won not by knowing more things, but by having a deeper understanding of combat.
You also mention, "It's technically possible to dodge 4 sword stabs from different directions" and "how much can one human dance around"... While it's true that skilled fighters are good at predicting where strikes will come from and moving accordingly (and it's true that movie directors like to depict this with invisible-wire-harness acrobatics), a plan that consists only of dodging is no plan at all. Simply put: When you're defending, you have to be lucky every time. When your enemy is attacking, he only has to be lucky once. This is where strategy comes in. If you ask your expert swordsperson how they would defend against four incoming sword stabs simultaneously, their response should be, "Why the hell would I ever let four people with swords surround me?" A good (i.e. non-Hollywood) fighter never engages more than one opponent at a time, because fighting even two at once is exponentially more difficult. A good fighter will outmaneuver their opponents, forcing them to engage one a time, and even using one opponent as a shield between themself and the others.
The last main thing you want to consider with your expert swordsperson is their environment, or more generally, externalities that they cannot control. This is related to the last point, because a good fighter will use their terrain to their advantage. Silat was developed in Indonesia, where frequent monsoons meant that you might be fighting while up to your knees in floodwater. As a result, the style emphasizes many low, stable, crouching positions—if you tried to fight like a Thai boxer in those conditions, you'd lose your footing immediately! And when a Silat fighter does get his opponent to the ground, many of the holds are so low that their opponent is essentially being drowned as they fight.
So a good fighter can use their terrain, sure, but what about the other externalities? What happens when they get up for their 106th duel and their allergies are going crazy? What happens when they decapitate their 231st enemy and their sword breaks? What happens if sand gets kicked in their eyes, or they pop their ACL, or they get a kidney stone? What happens if their opponent is a foot and a half taller? (height and reach are major advantages in swordfighting). In all these cases, your swordsperson must be able to fall back on their most basic instincts to adapt to the situation and find some way to disengage without getting killed.
How skilled can one human swordsperson be in terms of technique? You can make them do every sword trick in the book, cut with the speed and precision to slice a hummingbird's wing off, and have the endurance to swing around a Montante sword while wearing full plate armor... But are they strategic? Are they efficient with their movement? Can they adapt to changing variables?
Like others have said, if you want to show that your character is a master, "won a thousand duels" isn't a great metric. Defeating a circle of enemies, while flashy, is an unrealistic cliché. I won't write your book for you, but I will suggest you consume other media and examine how they show "the master." Do they defeat a big tough enemy that had beaten many good guys before them? Does everyone treat them with great deference? In Star Wars we don't see Yoda fight until the 5th movie, in Avatar we only see hints of Iroh's power until Book 3, but in both cases, they give thoughtful and wise advice, and people listen to them (except the protagonists, usually, who realize later that they should have). If you're looking for a place to start, try Ip Man.
And remember, for the fighter, it pays to be patient: https://i.imgur.com/0NENJTz.gifv
As already outlined in @Monty Wilde's great answer, it is all about probabilities and thus a numbers game.
However, @ETL touches upon a great point that deserves to be elaborated: Sun Tzu points this out in his writings a lot: You think of skill as too narrow, don't.
A great martial artist or fighter in general, knows not just how to block a punch, dodge a kick or swing a sword. He also knows where to position himself, how not to be in the way of the kick in the first place, and which fights to pick and which ones to avoid. There's a whole martial art about summed up in the nicely japanese statement: "If you want to thrust your knife there, I will not be so disrespectful as to block its way with my body." - Aikido.
Every good fighter can not just move his body, but also does the fight in his mind. He will assess his opponent long before the first blow is exchanged. He will analyze his style and adapt. He will use the environment, but he will also know not to get into that situation in which the numbers work against him.
The limit of human ability is not just the immediate situation. That is a training or competition setting - but when it's really life or death, a good fighter would not be in a thousand fights, not in his entire life. The legendary warriors, both of old and of the World Wars, such as the elite snipers, all had considerably fewer fights than that.
So, all that out of the way, let's try to answer your question, from my own martial arts and LARP experience and the literature I've read:
A grandmaster-level fighter can reliably (see Monty's answer) take down any beginner trying his luck. Betting on the beginner would be a lottery-level move - you are highly unlikely to win, in a hundred, a thousand or ten-thousand fights. A large amount of martial arts training (and almost certainly swordfighting and other fighter training in other time periods) is to get those moves into muscle memory. Once you have trained a move often enough, it becomes automatic. You can trigger it without much thought. The way you think "push button" or "pick up glas" and the subconscious part of your brain does the fine motor control details. That subconscious part is fast. Much, much faster than your conscious thought.
So anyone who needs to think about his moves doesn't stand a chance against a grandmaster. He may land a one-in-thousand lucky hit, but even that is unlikely to end the fight. Adrenaline is a thing. One of the first things I was told in martial arts was to expect to get hurt when you get into a fight. That is another thing that sets the experienced apart from those who are in a serious fight the first time: The experienced guy will keep fighting after he got hurt, the loser will clamp up, focus on the pain, or try to run away. In fact, in many cases that is exactly how he loses.
Someone who trained every day for his entire life is far beyond anyone who didn't. For comparison, look at professional athletes, chess players, boxers or really professionals in any sphere and compare their abilities to regular people. There are YouTube videos about people doing their job at a speed and precision we can only marvel at.
So in theory, given time to rest and heal and ignoring the sheer requirement of time, I would give the greatest fighter on the planet a good chance to defeat a thousand opponents. But if death is on the line, he wouldn't be in that series of fights at all, that's a part of what makes him the greatest.
Who are the opponents? What is a duel?
Here a chess master takes on 70 opponents in a big game. There are a fair number of kids there. I think one could reasonably expect the master to beat all of these opponents. A chess match has rules and the master will not be hurt or damaged. After he rests up he could do it again, and again. He could beat 1000+ such opponents. 1000+ persons who were themselves skilled at chess is another matter but he could probably beat them all too, unless among them was a master of comparable ability to his own.
But what is a sword fighting duel? Is this fencing, which might as well be chess as regards rules and the prospect of many wins. Or are they the Musketeers, or are they in a gladiatorial arena fighting to the death? If the last then rules are less important or not important and it is more of a roll of the dice each time for the sword master. They will eventually be up against some large and durable novice who knows they are going to die but is determined to hurt this sword master on their way out, and they do die, and they do hurt the sword master in a way that does not recover by their next fight, which is their last fight. The novice was only number 204.
Technological Edge (cheat):
If you want a warrior to have a huge advantage in skill, such that he can fight and win battle after battle, then what you need is not merely skill - unless everyone else in the world is an idiot or the opponents are preselected to lose. You need an edge - probably more than one.
- First, a rigid set of codes around your fighting helps. Once you eliminate as many variables in a fight (the other guy using a weapon you are unfamiliar with, or pulling out a crossbow and shooting you), then the battle becomes more predictable. You can develop a skill set to deal with a narrow set of circumstances.
- If your hero has armor of an advanced design that is unrivalled by his opponents, then he is less likely to be hit, hurt, and outmaneuvered. Better armor means you may be both better protected, but also faster, more maneuverable, and less fatigued than an opponent.
- Give your hero a better weapon. If there are rigid rules for how to fight, a higher grade of steel means a blade that doesn't dull, a longer sword that won't break and is lighter, or something that kills opponents that can only at best hurt you. The fighter with a Macuahuitl battling opponents with clubs will decimate his rivals as he consistently slices them in half on the first blow.
- Or you need a hero with some purely physical advantage that overwhelms his opponents. Imagine Goliath against ordinary opponents in a fighting style completely dependent on strength and leverage. Or your hero might have a mythological-type advantage. Imagine a warrior with hysterical strength and lots of time to rest between bouts. It's biologically possible, and would give them a huge advantage in any kind of fight that can be resolved quickly and with brute force.
But if you don't have an edge that puts you significantly ahead of your rivals, then eventually someone will either get lucky, or be willing to sacrifice any chance of honor or victory purely to hurt you.
There is a fantasy story by Heywood Broun about a teenage boy in Knight School who is rather cruelly and tyrannically forced to become a dragon slayer. The boy practices his skills before going out to hunt his first real dragon. What happens next? Read the story and see.
And one way this story could be interpreted is through probability.
I guess that the probability of a master swordsman winning one thousand fights is approximately equal to the probability that Kirk would survive all 79 episodes of TOS, or that Will Robinson would survive all 83 episodes of Lost in Space, or that Buffy the Vampire slawyer would be killed only once in 144 episodes. Rather low.
Suppose that Master Swordsman (MS), has a 90 percent probability of winning each fight. The probability that MS will will two fights will be 0.9 times 0.9, or 0.81. The probability of winning three fights will be 0.729, four fights 0.6561, five fights 0.59049, six fights 0.531441, seven fights 0.4782969, and that is already less than 50 percent after only seven fights.
Suppose that MS has a 99 percent probability of winning each fight.
So that makes 2 fights 0.9801, 3 fights 0.970299, 4 fights 0.960595, 5 fights 0.95099, 6 fights 0.9414801, 7 fights 0.9320652, 8 fights 0.9227445, 9 fights 0.913517, and 10 fights 0.9043818.
So that means that MS has a 0.8179064 probability of winning 20 fights, 30 fights 0.7396996, 40 fights 0.6689708, 50 fights 0.605005, 60 fights 0.5471555. 70 fights 0.4948374, 80 fights 0.4475219, 90 fights 0.4047306, 100 fights 0.3660051.
200 fights 0.1339597, 300 fights 0.0490299, 400 fights 0.0179451, 500 fights 0.0065679, 600 fights 0.0024038, 700 fights 0.0008798, 800 fights 0.000322, 900 fights 0.0001178, and 1,000 fights 0.0000431.
So even if Ms has a probability of winning any indivudal fight as high as 99 percent, making each fight rather boring and predictable, their probability of winning 1,000 fights will still be as low as 0.0000431, less than one chance in 10,000.
If you keep on increasing the probability that MS will win one fight, you can eventually give them a greater than 50 percent chance of winning all 1,000 fights. If you keep on increasing the probability that MS will win one fight, you can increase their probability of winning all 1,000 fights to 75 percent, 90 percent, 95 percent, or some other high percentage. But that would require making the probability that MS will win one fight, very, very high.
And it is one thing to make MS win 1,000 fencing matches on points where nobody gets injured, and another thing to make MS travel around challenging all comers to a fight to the death and kill 1,000 opponents. If MS knows that he is such a deadly fighter, and yet continues to challenge people to duels to the death, he will seem like a cold blooded murderer. But if MS starts challenging people to duels to the death before he knows that he has such a high probability of winning and surviving, he will seem like a suicidal fool.
Tha tis why protagonists who are expert deadly fighters are usually depicted as being forced by circumstances to fight.
That is my problem with long running television shows where the protagonists survive dangers week after week after week. The tiny probability they would have of surviving until the end of the series. They would have to be incredibly foolish to continue with their dangerous lifestyle after a few episodes.
So if an adventure series is highly episodic, like TOS, instead of serialized like Star Trek: Discovery, I picture the episodes happening in various alternate universes. Every episode, except for the few episodes which are sequels to previous episodes, happens in a separate alternate universe of it own. So I can pretend that the creators of the show searched though millions of alternate universes for ones where the protagonists had exciting advantures, and then searched though those universes to find ones where the protagonists survived those exciting adventures, and which also had otherwise good story potential, discarding alternate universes where the problems were solved by long and boring negotiations, for example.
Thus in a typical episode's alternate universe, the protagonists might get in a very exciting and dangerous situation where they have only a ten percent chance of survival, and do survive, one and only one time in their entire careers as police officers, or firefighters, or starship crewmembers.