Longsword Takedowns: The Good, the Bad, and the Ugly

To say that takedowns are a contentious topic in HEMA would be an understatement. For some takedowns make the game too dangerous, for others you are watering down and ruining the sport if you don’t include them, and everything in between. In this article, I’m not going to dig deeply into that debate, or proffer a personal opinion on whether or not takedowns should be included in practice or in tournaments, but I will explore some aspects of takedowns as they relate to longsword fencing. The content of this article relates specifically to longsword since that is what I am most familiar with. As a disclaimer, this content is all my personal opinion based on direct observation and anecdote, as far as I know there is no available data on takedowns in longsword.

Before I begin, some quick definitions. I will use the words “takedown” and “throw” interchangeably to mean an attempt to bring the other person to the ground using force or leverage of some kind, be it a lift and drop, trip, reap, sweep, hip throw, etc. I find that people sometimes use the term “grapple” or “grappling” to mean throw or takedown, but I have been careful not to use this word. In my opinion, “grappling” includes any use of the arms to directly manipulate the opponent’s sword or body, which includes many actions that are not takedowns. Those actions are outside the scope of this article. 

My Background

I started practicing longsword in 2015, after doing kendo for about 7 years prior. I took a semester of judo in college in 2011, and I enjoyed it, but that is clearly not enough experience to have any major effect on the way I fence. The club that I started longsword with was fairly grapple heavy, we had some classes that focused only on grapples, and takedowns were tolerated during sparring. From my first tournaments in the beginning of 2016, I scored points with takedowns, sometimes multiple times per tournament. I started judo again in August 2018, and continued until the beginning of the pandemic in March 2020, reaching 4th kyu (green belt). 

When I first did judo in 2011, I sustained a shoulder injury that prevented me from lifting my right arm for the next couple months, which affected my training of things that I cared more about, specifically modern fencing and kendo. Around 2015 or 2016, I was injured while performing a takedown during longsword sparring, when my opponent fell on my foot in a weird way and forced my toes back. During my judo run from 2018-2020, I did not sustain any injuries, but almost everyone at my club was injured at some point, including a friend breaking his leg in 3 places from taking a weird step while I was going for a throw against him. My reason for saying this is not to build a case against takedowns in HEMA, but to provide information about my background and personal experience with throw based wrestling with and without swords involved.

Two Types of Takedowns in Longsword

I broadly categorize takedowns in longsword into two types: “pre-planned”, and “last resort”. These categories are not absolute, and a lot of space exists in between. 


A “pre-planned” throw, as the name implies, is when you go directly for the throw without passing go or collecting $200. The intention to throw exists as soon as you make your initial engagement, and you set up or react to the situation specifically to accommodate the throw. This is how I interpret the “body throws” (leib ringen) of the RDL sources, as soon as the bind happens you go directly for the throw. Another famous historical one in this category is the famous von Baumman double leg, where you throw the sword over your back and rush in. In general, these seem to be less common in the modern landscape, though they still happen. The (in)famous “HEMA throw” that makes the rounds once in a while and polarizes people on the topic is an example of a pre-planned throw. 

Last Resort

Again as the name implies, a “last resort” throw happens when a fencer has expended all other tactical options and feels that there is no other choice left than to go for a throw. This is usually preceded by some attempt at sword infighting and arm grappling. For me personally, the most common setup for this is when my opponent has control of my sword in a grapple, and I manage to gain control of theirs. In a right hander vs right hander situation, this usually ends with me overhooking their left arm and them overhooking my right arm. This creates a potentially dangerous situation, as I will talk about when I discuss grips. The situation may also arise if one fencer involuntarily drops their sword and is able to initiate a grapple, with a takedown now being their only way to score (most rulesets require separation in order for a disarm to score, so a disarmed fencer can continue the exchange by clinching). In general I think that last resort throws are more common in our modern game. If I had to guess why, it would be because scoring with the sword or with a standing grapple is usually more rewarding and less risky than going straight for a takedown. 


Grips in longsword wrestling I find to be an interesting topic, because they are unique from both jacket and no-jacket wrestling grips. Judo (jacket) grips rely heavily on gripping the opponent’s judo-gi, which you can’t do with the gear that we wear when fencing. Longsword wrestling also offers different grips than no-gi or modern wrestling, because of both the gear and the presence of the sword. The presence of the sword provides opportunities for hooking with the pommel, requires you to dedicate a hand to hold onto it (if you want to), and provides a threat that must be neutralized in order to proceed with whatever else you want to do. Gear makes certain things more viable and certain things less viable – I mentioned that gloves make it difficult or impossible to grip jacket, but they also make it easier to control arms with something like a double arm wrap. If you tightly wrap the arms of a person who is wearing heavy gloves, they have no chance of escaping unless they can slip their hands out of their gloves. This is a lot more difficult to do in judo and wrestling, which is probably why you don’t ever see it done (though double sleeve grips are very potent in judo, which is a similar concept which takes advantage of gripping the jacket, and very effective). 

The main reason I am bringing up grips is because of an oft-cited safety factor with wrestling with swords, and that is the potential of sustaining an injury by landing on a sword in a weird way when you are thrown. I have not heard of any debilitating injuries from it yet, but it has happened, and if it does happen no amount of padded floors or break fall practice is going to prevent it. Therefore I will talk about grips specifically in reference to where the sword ends up when you throw. All visual examples used for the following can be found in this video.

Over-under/over-over/under-under, with sword

This is the most common one that I see and have historically used. The setup is generally from a “last resort” scenario, where both fencers have wrapped each others’ arms in a way that neither of them can be used, and hang on to their swords. This is a good position to set up a trip, reap, or forward throw. The problem with this position is that the sword is necessarily behind the other person’s back, so if you successfully execute the throw, unless you give up the grip beforehand or drop your sword, the other person will land on it. If the orientation of the cross is parallel to their back then they get lucky and will probably be fine, but if it is not, then there can potentially be an issue. 

No sword

Swordless grips happen either when one person has dropped their sword on purpose in order to grapple with both hands, or when one person was disarmed and has no other choice but to close distance and grapple. This grip usually manifests itself either as a bear hug (front or back), or a double leg. This was the type of grip used in the infamous “HEMA throw” video. While the grip itself does not present any danger of swords getting involved, it does lend itself to harder throws, such as lat drops and suplexes, as we see in the HEMA throw video. That’s not to say that it isn’t possible to do a “gentler” throw from this position, but it does enable harder throws more than other grips might. Also, when a sword is dropped, there is now a sword on the ground that can potentially be fallen on. 

This throw was excellently controlled
This one was not

Front arm wraps

Front arm wraps establish control of the opponent’s sword without bringing the arms around their back. In this way, the sword is not in danger of being landed on, as it remains in front of the person being thrown the whole time. As mentioned before, these types of grips are enabled a bit by the fact that we are holding a sword, as well as the large gloves that we wear, both of which make something like a double arm wrap more difficult to escape than it would be in something like judo or wrestling. 

Loose grips/clotheslines/one handed grips

These are generally what happens when you go for a pre-planned throw while holding on to your sword. The benefit to these is that usually the thrower stays on their feet, so there is less chance of landing on a sword. The downside is that you are usually leaving them to free fall instead of guiding them down in a controlled way, and something like a clothesline can easily dump them onto the back of their head. 

Joint manipulation

Joint manipulation involves positions meant to stress joints so they will become injured if you continue to apply pressure. Examples include straight arm bars and shoulder locks such as standing kimuras and americanas. These are usually disallowed in tournaments, and for good reason. I don’t want to go into all of the reasons why they shouldn’t be allowed in this article. This is the only opinion I will express on what should be allowed in a tournament, and thankfully it is not a controversial one. 

“Amplitude” of throws and pain of landing

How uncomfortable a throw is to be on the receiving end of and how likely it is to injure is a difficult value to quantify. In general, the more of a straight fall you have, the less comfortable it is going to be. This is why the infamous “HEMA throw” was probably painful, he picked him up and slammed him down. But there are many types of throws, and not all of them involve picking up and slamming down. So what else makes a painful throw?

Feet flying in the air

A metric that I have heard before and know people have made rules based on is whether or not someone’s feet fly in the air, or stay on or close to the ground. While this seems like a good rule of thumb at first glance, I don’t think it is, and I’ll explain why. Basically, if we are talking about how far your torso and head fall before slamming into the ground, whether or not your feet go up into the air has little to do with that. In many forward throws, the person being thrown is essentially doing a forward roll, which means their feet will end up over their head, but the fall will not necessarily be hard or painful. What matters more in my opinion is direct fall distance, which I will discuss next.

Direct fall/slam distance

Rather than the movement of the legs, as mentioned before, I think one of the important factors for how much the impact of a throw will hurt is direct fall or slam distance. For “fall” here I mean going to the ground without being slowed down either by yourself or your opponent, and by “slam” I mean the opponent applying a downward force to actively increase the force of impact. Here is an example that illustrates this (I was searching for throws that end in a KO, and this one does not despite what the title implies, but I find it illustrates the point well).

The first throw is uchi mata, a forward throw that flips the opponent so their feet are facing away from you when you finish. Notice they are essentially doing a forward roll as mentioned above. The thrower adds a little bit of extra power to it, so the throw probably didn’t feel great to receive, but here is what the direct fall distance looks like: 

The second throw is o soto guruma, a variant of the outer trip where you trip both of the legs instead of one. You can see that instead of flipping over, the person being thrown is driven directly into the ground. This greatly increases the direct fall distance, as seen here:

I’m sure neither of these throws felt great, but I guarantee the second one was more painful. Those kinds of throws can whip your head backwards into the mat, and/or hurt your neck by trying to prevent your head from being whipped backwards. 

Landing on the other person

Another potential source of pain in a throw is the thrower landing with their body weight on the person being thrown. What I mean by this is the thrower either falling on top of the person being thrown, or driving downward harder than they have to. This is not the case with every throw where the thrower goes down with the person being thrown, often they go down in a more controlled way, ending in a dominant position such as side control. There is a grouping of throws in judo infamous for this called makikomi, meaning wraparound, here is an example. Even though they try to be stoic and serious for these demonstrations, you can still see the person being thrown wince as they are landed on. 

Break falls

One thing that I hear a lot when discussions of the viability of throws in tournaments is break falls. 

What is a break fall?

A break fall is an action that someone takes when landing in order to mitigate damage from the impact of a throw or fall. It may do this in several ways:

  • Spreading the force of the impact over a large area
  • Dispersing the force via slapping the ground with one or both hands
  • Sometimes mitigating the force of impact by rolling through and ending in a standing position
  • Preventing arms or other limbs from “posting” the ground during a throw and thereby risking injury
  • Preventing the head from slamming into the ground

Here are some examples of break falls as taught and used in judo. I think that break falls serve their purpose quite well, and anyone who wants to practice throw based wrestling should know how to do them. However, I don’t think they are the answer to making throws in longsword safe, and I will explain why.

Pre-tournament break fall seminar/qualifier 

This is a suggestion I have heard for tournaments that potentially allow throws. The idea is that the participants should take a short seminar on how to break fall before the tournament, in order to be able to prevent injury while getting thrown. We can extend this idea to practicing break falls during practice, while not also including throws in your everyday sparring. 

The problem with this is you are isolating the break fall action from its context, that of being thrown. In the pre-pandemic Bucks curriculum, we included break falls as part of our warm up, so everyone practiced them every single class, even though we did not allow throws during sparring. We did this because I wanted to prepare everyone for tournaments where throws are allowed. However, I had some members try judo, and while they could demonstrate fine isolated solo break falls, when they were thrown it was like they had never practiced them before. Now that I know about the ecological approach it is obvious why this was, but at the time it was a big eye opener, and I lost faith that any effort not including full throws would be useful for learning break falls.

Break falls reducing injury

Putting aside whether or not the participants know how to do them, we can ask ourselves if they actually do anything to reduce injury. Like other topics in HEMA, we have no actual data on this, but I suspect the answer is no, and I have a few reasons for thinking this. First, if we go back to the grips section of this article, very few of them result in a situation where someone has a free hand to break fall. Usually the grip locks both fencers together, and the person being thrown has no control over how they will land. Secondly, if we look at judo competition, we can notice that no one actually utilizes their break falling skills during a live competition (see if you can spot any break falls in this video). This is mainly because if they land on their back they instantly lose, so they try to prevent that as much as possible, and also you don’t always know when you’re about to be thrown, so you don’t always have a chance to break fall. The latter is more applicable to longsword; a throw usually comes as a surprise to a fencer, so even if they are great at break falling they might not have a chance to do it.

Another issue counting against break falls preventing injury is the fact that, similar to the issue with mats, almost every debilitating injury that has been or can be inflicted by a takedown would not be prevented by break falls. A break fall will not prevent your leg from getting twisted if it gets caught in place while you are thrown, and it will not prevent you from landing on an upward facing crossguard. Like mats, break falls deal solely with the issue of impact when you hit the ground. 

Break falls are still useful

That being said, break falls are a very useful tool, just not at preventing injury during competition. Break falls allow judo players to be able to take many falls in a row during practice without getting hurt. This is necessary, as it allows you and your training partners to get in reps of throws that you want to learn. They are also a useful skill on the streets, as they can potentially create good reflexes for when you slip and fall on the ground, or fall for any other reason. You just need to make sure you’re practicing them while being thrown or falling non-voluntarily, otherwise the skill will not transfer.

Marked takedowns

This is another potential solution I’ve seen offered to include takedowns without the negative safety consequences. The idea is that the judges identify when you lift your opponent off the ground, or otherwise destabilize them in some way without actually finishing the throw. This is an interesting idea for a fix, and it does work to an extent. Here I will list two issues with it.

Limited grips

While in theory a ruleset may allow for a variety of ways for a marked takedown to be called, in practice the only one that will get called is a fencer lifting their opponent off their feet. Because of this, only a few grips are viable, namely the double under, and front and back bear hug. Something like a fireman’s carry is also potentially possible, but I’ve never seen anyone get close to something like that (it would be impressive to see). Either way, this severely limits the amount of viable takedowns in an otherwise large pool.

It benefits larger people

While all grappling benefits larger people (taller and heavier), marked takedowns take it to the extreme. The main method of scoring a marked takedown as I mentioned above is a bear hug and a lift, and a larger person is going to have a much easier time doing that on a smaller person than vice versa. While difficult, it is definitely possible for a smaller person to perform a takedown against a larger person. In doing so, they will most likely have to take advantage of an off-balance of their opponent and use some kind of throw that involves leverage, like a hip throw or a trip. If you are doing a throw like that against someone much larger than you that you can’t lift, there is no way to mark it, they are either going over or they’re not. 


Another common justification, usually for disallowing takedowns, is lack of mats. This is a logical argument, because many other sports that include takedowns practice and compete on mats. While this is a fair point to make, there are a couple of points to consider regarding this topic. 

Effect of gear

Unlike judo, wrestling, or any other modern sport that involves takedowns, we are covered head to toe in protective gear when we compete at longsword. This gear covers many of the most vulnerable parts that would otherwise be in danger of damage from impact with a hard surface, such as elbows and back of head. Sean Franklin did a study that suggested that the impact of a fencer in gear on a hard surface was similar to the impact of someone with no protection on a matted surface (no link, sorry, you’ll have to take my word for it).

On the flip side, gear can also dig into your body or get caught on things in weird ways, causing unexpected and potentially dangerous things to happen during the fall or during impact. However, the presence or absence of mats would not do anything to mitigate this danger.

Source of injury

The other thing we can ask ourselves is whether or not injuries in HEMA associated with takedowns come as a result of the impact with the ground. While there are certainly examples of uncomfortable impacts (the infamous HEMA throw comes to mind, and that was on a matted floor), I personally don’t know of any debilitating injuries resulting from the fall. Injuries tend to happen because limbs get caught in a weird way during the fall, and a joint or bone gets twisted or forced in a direction that it’s not supposed to go. The takedown injury that I personally received was a foot injury, which I sustained as the thrower while still on my feet, because my opponent landed on my foot. There is also the issue of landing on the sword. Neither of these things are affected by the presence or absence of mats. It is possible that impact does play a role, but for that I would need data, and as of right now the anecdotes that I know of suggest that it historically does not.

Effect of takedowns on modern longsword fencing

The final aspect of this complex topic that I will explore is the actual impact takedowns have on how the game of longsword fencing is played. I said in the beginning that there is no hard data about takedowns, but that is not entirely true, there have been a couple of tournaments that have recorded their exchanges granularly enough to differentiate which exchanges were scored as takedowns. They are Socal Swordfight (2018-2022), and AG Open (2022). Thanks to Sean Franklin and HEMA Scorecard for providing this data.

Before I show the data, a couple of caveats with it: The ruling on AGO 2022 and Socal 2022 were the same regarding takedowns, the thrower must remain standing to score. Here is the exact wording used: 

A standing takedown is a wrestling action that deliberately and safely puts the opponent on the ground, while the throwing fighter remains standing. Standing requires one foot remain on the ground, dropping to a single knee is acceptable.

-Socal 2022, AGO 2022 rules

This restriction reduces the amount of viable throwing options, and therefore contributes to a lower total number. Only the rules for the 2022 editions were available on HEMA scorecard, so it’s possible that earlier installments of Socal had different rules. Additionally, tournaments often will have rules restricting what takedowns can be done in some way, such as control, amplitude, or safety. 

Here’s the relevant data, total exchanges scored compared with total exchanges scored by throw, and percentage of score exchanges scored by throw:

TournamentTotal exchanges scoredTotal Throws scoredPercent of exchanges scored as throws
Socal 201820831.4%
Socal 201929731%
Socal 2020166750.3%
Socal 2022205540.2%
AG Open 2022168910.06%

Here’s the total breakdown of all scoring exchanges over all of these tournaments:

*Slice data was only recorded for AGO 2022 and Socal 2022

Though this data is only from a very small subset of tournaments, it tells us what we already knew in our hearts: takedowns only make up a very small subset of scoring exchanges in a tournament. It is likely that the 4 exchanges out of 2055 at Socal 2022 would not have drastically changed the results. 

However, there is another aspect that must be considered, which is how it affects gameplay. An action does not necessarily have to happen often in order for it to drastically impact the way people play. For example, in Super Smash Brothers Melee, there used to be an allowed mechanic called “wobbling” with the character Ice Climbers. Wobbling is an infinite combo that the Ice Climbers character could initiate whenever they grabbed their opponent, leading to an almost guaranteed kill. While Ice Climbers players almost never won major tournaments, and they were very beatable in competition, this mechanic dominated how they were played and played against in the meta. The game became making sure you didn’t get grabbed, because that matters more than anything else. Because of this, other options were opened for the Ice Climbers player, and they could very well win a match without wobbling at all. However, even if they did not use it, wobbling still had a huge impact on the way the game was played. 

All that is to say, that even if very few exchanges end in throws, it’s possible that the threat of it happening can have an impact on how people fence. How big of an impact this is, and whether or not there is an impact at all, is difficult to say without doing some kind of study on the phenomenon, I just want to bring it up as a possibility to take into account when making ruleset decisions. 


As promised, I will not present a personal opinion on whether or not takedowns should be allowed in tournaments. Ultimately it is up to the tournament (and at some point in the future, the community as a whole) to decide what level of risk they are willing to tolerate. Personally the legality of takedowns in a tournament has no bearing on whether or not I am willing to attend. In reality, takedowns make up a very small percentage of exchanges. They are a thing that you have to be aware of, but they are never a major scoring method or game changer, even for people who are known for doing them a lot. One thing that I will say for sure is that they are a crowd pleaser, as long as they do not look unnecessarily violent. 

Stats for nerds

Here is the raw data that Sean gave me for types of exchanges scored:

TournamentCutThrustThrowPommelSliceTotal exchangesCutThrustThrowPommelSlice
Socal 201811774314020856.2535.576923081.4423076926.7307692310
Socal 2019163113318029754.8821548838.047138051.010101016.0606060610
Socal 202013342995290166780.023995217.936412720.2999400121.739652070
Socal 202216104144216205578.3454987820.14598540.19464720191.021897810.2919708029
AG Open 2022133033511211168978.7448194219.834221430.059206631140.71047957370.6512729426