The free rider problem as illustrated by a Japanese fencing video

Jason Kottke blogs:

A Japanese TV show took three expert fencers and pitted them against 50 amateurs.

I honestly didn’t think this would be that interesting and expected the Musketeers to easily get taken out right away or, if they survived more than 30 seconds, to handily finish off the rest of the crowd…nothing in between. But it’s fascinating what happens. The crowd, being a crowd, does not initially do what it should, which is rush the experts and take them out right away with little regard for individual survival. But pretty much every person fights for themselves. And instead of getting easier for the Musketeers near the end, it gets more difficult. The few remaining crowd members start working together more effectively. The survival of the fittest effect kicks in. The remaining experts get sloppy, tired, and perhaps a little overconfident. The ending was a genuine shock.


It looks like the free riders (or *superior strategists*) win. How is that a problem?

Because they tend to be bolsheviks.

Winning in this context, isn't landing the winning blow, it is still having an un-popped balloon at the end. So it doesn't matter if I engage the experts at all. I can just play defense as other amateurs engage. Eventually I might have to fight to win rather than fight to survive, but my odds overall are better just letting either the experts tire out or someone else to get lucky.

Right... so what's the problem?

It's not a problem if you win, but thanks to free riding, 96% of amateur fencers didn't.

Interesting point, but I assume that without the free riders the 96% would still have lost (besides, I can't tell if those at the end really were free riders or not).

Also this: TC calls it "free riding," I call it "haste makes waste" (or "don't just do something -- stand there"). Usually a good plan, regardless of all that "man of action" propaganda.

Don't have a problem. I was just saying that winning wasn't about besting one of the experts, it was about waiting around until they were all eliminated.

I think most kids would rather have their dodgeball team lose rather than win and be one of the first kids out. That's just humiliating.

That's a pretty good effort. They're fencing saber, which is what I used to fence. To 20 points without being scored on, especially in a gang setting like that, well, that's pretty good. With a little more luck, the final guy would have survived.

Just by random chance, there would probably some really good amateurs among the 50 who knew they were above average. So they waited until the end on purpose.

I saw very few amateurs playing with a deliberate strategy to just wear out the pro, but maybe there were and I just couldn't tell.

Well, I respect saber most of all the modern forms, but its still not a realistic way to fight with a blade of almost any kind. The big problems with this kind of contest are:

1) No grappling
2) No backstabs
3) Glancing hits count as much as "fatal" strikes.

These differences lead to very strange behaviour; look at all that wrist work! IRL, in a real "gang" setting, backstabs and rush tactics would be decisive.

Glancing hits won't pop that balloon.

Sabre, like foil, has the concept of "right of way". The attacker has the right of attack and the defender must defend before he can counter-attack. There is some logic in this, as it presumes that the individual's survival is more important than death or defeat of the opponent.

By contrast, epee has no right of way. First touch scores. In a sense, that's also realistic. In real life, there are no rules. So these represent different philosophies of combat, both with their own logic.

One of the interesting effects of fencing epee is that it destroys your footwork. In sabre and foil, right of way creates a very specific rhythm to the footwork, and that goes to hell in epee. If you fence sabre, you'll fence foil from time to time, but if you fence epee for fun, it will take the better part of a practice to get your footwork back in order.

They are not fencing sabre, they are fencing foil. The pros are even wearing their lamé jackets.

Ooops! My bad.

They are indeed fencing foil. (But they are fencing like sabre fencers!)

Two of the three pros are Sugura Awaji and Ryo Miyake. I am guessing the third is either Kenta Chika or Yuki Ota. That would make them Japan's Olympic foil fencing team, which won silver in 2012. So it's not just three pros, it's the second best collection of three pros in the world, if the 2012 Olympics is the standard.

So, this wasn't a samurai style event? - which, depending on your preference of Kurosawa or Weber as a source, pretty much involved a single fatal stroke? (till downloading from youtube - but as a comment mentions saber, it appears to be something more stylistic than actual sword fighting - which is fine, of course.

It's fencing...and modern fencing is about as close to real sword fighting or blade combat as all that Samurai "single fatal stroke" nonsense.

In a real fight, the experts would be stabbed in the back about 5 secs in when they get surronded.

In a real fight, the experts would have used to stairs to prevent that - as one can see at the beginning of the video. Though that strategy is quickly discarded - likely because this is entertainment, even if the Olympic fencers' first reaction was pragmatically defensive.

As for that first stroke - there is a difference between combat and dueling.

An interesting account here - 'In accordance with custom, the combatants exchanged names and swords were unsheathed, the three samurai on one side facing their solitary opponent, with whom the sympathies of the onlookers evidently lay. The keen blades of the duelists glittered in the sun. The ronin, seemingly as calm as though engaged merely in a friendly fencing bout, advanced steadily with the point of his weapon directed against the samurai in the centre of the trio, and apparently indifferent to an attack on either flank. The samurai in the middle gave ground inch by inch and the ronin as surely stepped forward. Then the right-hand samurai, who thought he saw an opening, rushed to the attack, but the ronin, who had clearly anticipated this move, parried and with lightning rapidity cut his enemy down with a mortal blow. The left-hand samurai came on in his turn, but was treated in similar fashion, a single stroke felling him' to the ground bathed in blood. All this took almost less time than it takes to tell. The samurai in the centre, seeing the fate of his comrades, thought better of his first intention and took to his heels. The victorious ronin wiped his blood-stained sword in the coolest manner imaginable and returned it to its sheath. His feat was loudly applauded by the other samurai who had witnessed it. The ronin then repaired to the neighbouring magistrate's office to report the occurrence, as the law required.'

I may be missing something, but I read it differently. The crowd initially rushed en masse. True, they did not swarm like lemmings, but since the balloon is only in the front you can't really do that.You have to engage the experts from the front , so only one or two people can attack at once. However, the fencers defeated this strategy by splitting up, which made it hard for the amateurs to coordinate their movement and timing. Thus the crowd scattered. They seemed to reform in smaller groups that could (implicitely) coordinate their activities. At that point it was simply a matter of hoping to get lucky or tiring the experts out.

Overall the crowd seemed to do the best it could, and the experts responded appropriately and the numbers favored the crowd.

I like this comment. To me, what was most interesting about that is that it looked like it was showing an emergent property of crowds. One that might actually have been very efficient in terms of satisfying the win condition. That crowd might actually have sorted itself out in a pretty efficient way.

Now, if the win condition changed to winning only if you're the one who tags a red guy, the 'free riders' might have turned into 'stampeders'.

The first expert eliminated had nothing to blame but himself. If he hadn't gone down then I bet the result would have been different. Foolishly charging forward too early is a frequent source of disaster in all combat

It wasn't a foolish charge; it was logical, and for the most part successful, behavior. Aggression is why the experts were winning.

He might have charged the wrong area - but optimal behavior in "one vs many" fighting is to periodically break out and force opponents to engage you in a one vs one manner. The experts did this correctly many times successfully.

My Japanese is a bit rusty, but I'm pretty sure the guy talking said something about "fake video to fool stupid lazy American economist."

Tyrone, is that you?

"Gojodan de syo?" (You gotta be kidding)

I've done some simulated sword combat ("boffer" weapons rather than fencing sabers, in my case) in a many-on-one scenario, and it's surprisingly difficult to effectively team up with other attackers in this kind of combat. If you were grappling, I imagine it would be relatively easy, but with swords, timing multiple attacks from different directions against a mobile opponent who does not simply hand you the initiative is a hard problem. Fencers -- even good ones -- don't regularly practice team-up scenarios, so I'm not surprised that the initial rush was underwhelming.

The mobility of the experts combined with the rules of this event made teaming up hard. If It was legal to chop the legs, back of the head, etc it would have been different, but since the only legitimate target was the balloon it played out different.

If they had worked together though, they could have cut the corners off trapping the experts. Or formed a ring.

Maybe. I never participated in a scenario which was as one-sided (16 people against one) as this, but when I did it, our rules were fairly permissive about target areas (no head shots, but legs and back were acceptable). The thing about forming a ring is that, first, it's hard to form a tight ring around a mobile person who's willing to attack elements of the ring while it's still being formed. You probably aren't much faster than your opponent, and some of you have to cover a ton more ground than he if you want to move into a ring formation.

Just getting two people to more or less the opposite sides of someone is really hard when that person has lots of space to move, can plausibly threaten any isolated opponent, and is willing to prioritize not being surrounded. I suspect that in a real fight, the big advantage of the group is that to avoid being overwhelmed, the lone expert has to run basically continuously, while also sword-fighting (which, for those who haven't fenced, is shockingly tiring, shockingly quickly). You're going to get tired really fast while evading the surround.

The ring doesn't need to trap the expert, especially when 50 on 3. They could have formed a ring like bison fending off wolves, for instance.

I also don't need to be that much faster individually. An inverted v will trap a person, if coordinated. In the video, people were repeatedly eliminated just because they weren't focused on the experts. I see this as an example of lack of training on teamwork more than lack of training on fencing.

Good for you - and I agree its harder to sync then one might suppose - but note the prohibition on rear attacks here too! If you're limited to a 120' arc then of course it's a lot harder...

But yes, in real life, "grappling" is not only allowed, but decisve in these kind of fights.

Everything in the video is to be expected.

The experts immediately look for the best location, which in this case are the steps giving a height advantage.
The crowd is unsure of itself without a leader and does not respond as quickly.
The crowd's mass means that the experts can consistently move from opponent to opponent without getting too involved with one person.
The experts understand the value of attacking on the move.
The experts stay aggressive. If they had chosen a defensive strategy they would have been overwhelmed.
The experts adjust to the event and are highly defensive in body posture while attacking. Only the balloon needs protection, so a low crouch with the lead side lower is used to protect against a blind side under arm strike.
So while they are on the offensive and are consistenly moving to the attack, they don't fail to keep their bodies aligned in a way to protect the balloon.
The experts consistently target the weak link, which in this event was an opponent not paying attention.
As the crowd thins, its mobility is somewhat restored increasing its effectiveness.
The crowd's numbers were useless when not because they weren't as effective with the sword, but because they lacked the teamwork. If they had just focused on trapping the experts and limiting movement, this would have been over much sooner. Instead they let the experts move at will.
Alternatively, they could have formed a ring facing out.

This was basic maneuver warfare. Agile, trained, and confident combatants almost won.

Out of idle curiosity, those of you surprised, have you ever been in a fight?

"Out of idle curiosity, those of you surprised, have you ever been in a fight?" - a knife fight?

I recall from something on Youtube that in a 'bum rush' of fifty people against a target, a kung-fu master, the kung-fu guy can take on a crowd because at most only three (3) people can fit around the target, so it's really three vs one not three vs a crowd of fifty.

I don't think the kung-fu master would have been as successful had he not been on Youtube. YMMV.

That's 16 three-on-one fights and 1 two-on-one fight, back to back. Kung fu master had better have a lot of endurance.

One thing in your response rings true - knife fighting is terrifying for the average person. Even in Ranger School they taught that if you are in a knife fight, you WILL get cut; the trick is to take the blade in a non-critical spot while killing or disabling your opponent. Clearly you can't do this against 50 people very well, but it helps if your opponents are rank amateurs.

The fencing masters show remarkable control of their saber tips - a slight flick, twist, and thrust and they're in the kill zone.

The rules certainly helped even the odds for the masters since they had only a small target to defend. But a large part of their ability to avoid being killed was their mobility. Even if their opponents were behind and couldn't strike, they could have walled them off to prevent their escape.

It would be interesting to see how the fencing behavior changes with the payoffs based on your team winning and you surviving.

The old saying is "Never insult seven men with a six shooter." Well, the reality is that you kill one of them immediately and dare the rest to be number two.

In fact, in infantry tactics, the support team suppresses the enemy with a high volume of fire so that the assault team can maneuver. The goal of the support team isn't necessarily to kill the enemy, but rather to keep them from concentrating fire on the assault team. In colloquial terms, the support team keeps the enemy's heads down. The individual motive to raise your head to fire is severely threatened. When trained in this battle drill, an infantry squad can usually destroy a much larger opponent which is not so well-trained and disciplined. Of course, a well-trained defender knows how disrupt an attack and regain fire superiority.

Great stuff Willitts; Are you service, military historian or military operations research?

Eight years of active duty in the JAG Corps including four years in the XVIII Airborne Corps.

Amateur military historian.

Sounds like a Youtube myth, Ray. I know enough people with martial arts, and my way around unarmed/melee tactics, and no-one should take those odds. Bum rush is a very effective way to conclude things.

Good analysis!

But the game is quite "rigged" to allow the experts to use their advantages. I'd like to see them allow rear attacks and grappling, and play again.

The 3 on 50 fencing video turned out better than the one using pistols.

Lol - ouch. Yeah.

I thought the experts did very well, and aptly demonstrated their expertise. Can't say I could tell if the free rider phenomenom was demonstrated at all - since none of the 50 were individually identifiable. Nor was their level of effort identifiable.

What was observable was that the experts put out a lot more energy than the attackers.

But that was pretty cool. Reduced the attacking odds from 50-3 down to 2-1. Pretty massive success, I'd say, so long as the prize wasn't the life of the crown prince.

You're bursting a lot of balloons.

Real pistol experts might have a similar performance.

Give them rifles with a lot of terrain between them, and those three could do very well against 50 untrained opponents also equipped with rifles.

However, victory would be much more likely with a whole squad of 9 against 150 rather than 3 against 50. Three riflemen simply can't lay down enough suppressive fire unless one has a machinegun, and three people can't mutually support each other for maneuver.

Good points. The tactical effects are NOT scale free; most people don't know that.

If you want to talk snipers, the 50 won't fire a shot. See Afghanistan.

No - pistols would not be such a disadvantage for the less skilled. That's the whole beauty of gunpowder weapons...they allow you to use levied mass much better, at least in the pre-modern era.

That is until the ranges got far enough that things like maneuver, unit coordination, and stealth became important. At that point professionals regained an advantage (which is why small South African merc groups tend to devastate larger untrained African forces).

Exactly. The average person thinks of firefights within rock throwing distance - an obvious artifact of too much television and too many drive-bys.

Well-trained infantrymen will use the entire maximum effective range of their weapons - in excess of 500 meters - and gain standoff if they can over shorter range weapons.

They will employ cover and concealment, fire and maneuver, night vision equipment, snipers, squad and platoon level tactics, deception, and special weapons such as machineguns and grenade launchers. A well-trained unit can easily defeat an untrained enemy force three times its size. Of course, the smaller unit bears the peril of bad luck.

This nicely illustrates what was wrong with the Mythbusters zombie hoard episode. They wanted to test the efficacy of various weapons against a zombie hoard, but had rules which severely limited the mobility of the human. So the human didn't do very well regardless of his weapon.

When you're better at combat than the other guy, every little bit of extra mobility is a huge advantage. You can choose to fight only when you are going to win with high probability.

The essence of military tactics is staging a series of unfair fights.

Maneuver is one of the Elements of Combat Power and also a Principle of War.

Maneuver is different than mobility; the former is using the latter to gain a tactical advantage. But I know that's what you meant.

All of the attackers were eliminated in one-on-one combat, and by the more skilled fencers in the crowd at that. I did not see any successful coordinated attacks, which, as pointed out above, are more difficult than they seem but not impossible: limiting mobility is key. Crowd fail.

As others have noted, it wasn't really possible for fencers to gang up. An amateur couldn't get a clear shot at the red balloon without pushing his compatriot out of the way. If you look closely you'll see this "cutting in" happening a lot (one fencer retires to make room for another). Also, the final olympian was felled because of a momentary lapse in attention. There was nothing inevitable about this outcome. The three champions could have won (though, I agree, they looked tired at the end, and that might have been the equalizer).

Good point - the restricted arc of attack eliminates most of the numbers advantage and grossly flatters the experts. I'd like to see this again with rear-facing or head-mounted balloons too.

Fatigue is a real factor though - melee combat is exhausting!

I actually used to do an activity very similar to this in fencing class. A handful of advanced students could fairly easily take on a much larger number of beginning students and prevail, even in single touch. This particular setup is harder for the rush-job than with regular rules, as there's a very small target area (our advanced suffered more casualties more quickly than these Olympians simply through errant lunges that landed, just not on whom they were targeted). If they actually played using sabre or epee rules, the massive rush would make much more sense.

And it was possible for the amateurs to gang up--they don't all have to be attacking. Form a wall facing away from the Olympians, reduce their range of motion, etc. Tyler's point is still well-made: even though the rush-job suggested is difficult in this particular game, there were cooperative possibilities, and many participants instead chose to free ride.

>many participants instead chose to free ride.

Sounds like you have never heard that term before, or know what it means.

Those that hung back were not calculatedly reaping the benefits of costs paid by others. It was a team event, and the entire team either won or lost.

That's not necessarily true when there is disutility to effort. The lazy are hoping to win without effort. This is precisely the free-rider problem.

There is a personal benefit to being among the remaining contestants--it's called "bragging rights". Note that at least two of the Olympians are taken out by amateurs who the Olympians dispatched simultaneously. Staying out of the way and hoping someone else wins the day, but gets killed in the process, is definitely reaping the benefits (being one of the last fencers standing) of costs paid by others.

And if you'd like to argue that the real glory belongs to those who took out the Olympians but got nixed in the process, then you really need to explain why so many of the amateurs were so eager to fall back. So thanks, but I have heard the term for this behavior. So has Tyler: it's called free riding.

Transport trucks for sale
We Buy & Sell Used Trucks!

We will work with you one on one to help you retail or wholesale your trucks. (We have many references). We only need you to send us the specifications as well as pictures of the interior & exterior of the truck(s) and let us go to work. All we ask is if there are any issues with the trucks please inform us. It never hurts having another salesperson working for you for free. The way we get paid is we add $1,000 to your bottom line price. When we receive a lead, we call and email you with the person’s information. You then will set up a time to show them the truck or trucks. If they buy from you, we will invoice you for $1,000. It’s that simple!

We can also locate used trucks at fair prices. We call on 7,000 companies throughout the USA every 2.5 months and do come across fair prices.

We want to thank you again for the opportunity to work with you and your company.
DSS Trucking Services Ltd
Darren Sheppard
Phone: (855) 701 6388
Cell: (416) 606 2372
Buy from here:>>!inventory/c16gl

Samurai truck seller willing to duel to the death - that would be amusing. This isn't.

I would lay a small wager that the linked page contains malware. I am not going to make the venture myself, as the potential impact is greater than I care to attempt to repair at the moment. But I would not be surprised if any of the readers who clicked through were joining a botnet.

There was really no advantage to ganging up on the experts. The target was a tiny sphere on the chest; having guys behind the expert was useless. A 2-on-1 frontal attack gave some advantage, but not much, because the expert could always shift to one side, negate it, and quickly kill a novice.

So the question here was really: Can 3 experts each defeat 17 novices in a row, mostly one-on-one? And the answer is yeah, pretty much.

A better question is: Why do the Japanese consistently put out more compelling television than the US does?

Having fencers behind may have proven decisive much earlier. Fatigue proved to be decisive, at any point the expert's only opportunity to rest was to run away and intimidate the amateurs into giving a breather. Often the expert ran backwards, but with another attacker behind he opens himself up to an attack from the side by an unseen foe.

This means his only options is to attack and break through, this is a much riskier strategy.

I think the experts were smart early on to counter this strategy by spreading out and covering eachother. The crowd had trouble encircling any expert because they were worried about an attack from another expert and a ton of the expert kills were actually instances of an amateur getting flanked by an expert as the amateur was busy with another expert!

The experts lost when they all gathered together at the end. This allowed the crowd to form a solid wall and not worry about flanking manoeuvres.

Nice video! Reminds me of Gordon Tullock and Geoffrey Brennan's article 'An Economic Theory of Military Tactics: Methodological Individualism at war':

You economists! The three experts are left-handed, and are likely well versed at protecting themselves against right handed fencers. Most of the novices are right handed and perish quickly. Most of those left standing at the end are lefties, although at least one of the guys who pops a balloon is a righty.

As a lefty who participated in a lot of weapons based combat sports that included both one on one and group combat, the lefty v righty thing makes a big difference against midlevel competitors. The best competitors can handle it because they see it all the time, the beginners don't know enough for it to make a difference.

Not shocked at the end, at least to an OMG extent.

It reminds of the standard crime situation where onlookers do nothing even though they would have easily stopped the crime together.

I'm truly enjoying the design and layout of your blog. It's
a very easy on the eyes which makes it much more pleasant for me to come here and visit more often.

Did you hire out a designer to create your theme? Exceptional work!

I played soccer (futsal, 5 on 5) with a player who was exceptional. I noticed that when facing a 1:1, he frequently slowed things up to attract a second defender. When I asked about this, he said it was because they tend to get lazy when double teaming, making it easier to get a breakaway on both of them. This looked similar.

surefire xt07 for sale

Comments for this post are closed