Study on fights

[DWx]

MMA fights are not representative in any way of fights on the whole, though; they are completely unsuitable for use as a control for anything but MMA fights.

MMA Fights are the only way you can control for differences in weight, height, strength, experience, etc. For an academic study of this nature to be published, you would need to control numerous factors, which you cannot do if you use Youtube fights.

The ethics are questionable but a better control group would be to pull people off the street and then match attributes like height and weight and make them fight whilst you watch

Using MMA fighters introduces extra variables. When looking at how likely it is that a fight goes to the ground, you can't use MMA fights because a good proportion of the fighters want the fight to go to the ground. They're trained in BJJ (or something similar) and can attempt to win the fight via submission.

You can't control the fights themselves on YouTube but you can be selective about what you include in your sample. So do a sample where you've selected fights where the participants are roughly the same height and weight. Or control the variable like location so do a sample where its only fights on a hard surface such as concrete.

[MasterPain]

No ethics, untrained fighters?

It's called Bumfights.

[Liver Punch]

Ahhh, bumfights. Anyhow, I think that studying fights - and what kind of fights you study - depends on what you want to know. As a martial artist and fighter, I don't care what a drunk football player and a fat old man in a wheelchair do when the fight hits the ground at the bar. That's irrelevant to me. (albeit entertaining)

So, using the scientific method, we must first determine what it is we're trying to learn.

Also, if we're looking to apply a traditional martial art rule of thumb to this, we've already established that we cannot have two untrained fighters. The TMA rules only apply to those in TMA.

[ShoriKid]

Ahhh, bumfights. Anyhow, I think that studying fights - and what kind of fights you study - depends on what you want to know. As a martial artist and fighter, I don't care what a drunk football player and a fat old man in a wheelchair do when the fight hits the ground at the bar. That's irrelevant to me. (albeit entertaining)

So, using the scientific method, we must first determine what it is we're trying to learn.

Also, if we're looking to apply a traditional martial art rule of thumb to this, we've already established that we cannot have two untrained fighters. The TMA rules only apply to those in TMA.

If I want to examine the sorts of violent attack I'd most commonly face in a defense situation outside of pre-arranged mutual combat, the drunk football play and the fat old man are what I'd be more interested in. If I want to learn what I will face when confronted with another highly trained martial artist in a traditional art, you will want to watch particular types of competition.

If I'm examining something concerning TMA it does not dictate that I can only examine those fights consisting of two traditionally trained martial artists.

[Liver Punch]

If we're attempting to study the types of attacks and the techniques employed by our attackers in a street fight, then yes - i think the fat guy and the football player are a great cross-section worth looking at. But what happens to those guys when they fall down in the middle of a fight will look different with someone educated in martial arts in the mix.

I'm just not necessarily convinced that I care what happens in a fight between two people that have no idea what they're doing.

For example: If, indeed an overwhelming number of people who go to the ground in a fight first, lose that fight, then I don't want to be the guy who ends up on the ground first. But if that number begins to favor the guy who lands on the ground first if he's a martial artist, that statistic is far more useful.

What happens to someone who doesn't relate to me in a fight is borderline irrelevant. Most people are not martial artists. This simply means that most fight data involves people who aren't trained. Using the idea that you're more likely to lose if you hit the ground first as a hard and fast rule means that we would be advising Royce Gracie to attempt to strike with Mike Tyson. I feel that this would be bad advice.

I think the rules of a fight remain to be common sense. If you're a striker, try to strike. If you're a wrestler, look for a takedown. If you're a guard-reliant grappler, get the fight to the ground no matter what. If your martial art doesn't have anything to offer you on the ground, then yes, you're better off not going there.

[Wolfman08]

If there's no control, the study is not academically valid. The whole findings of the study could be the result of a series of dumb luck. Maybe the reason why the guy who went down first tended to lose is because his opponent had 50 lbs and a BJJ black belt on him.

[tallgeese]

I think we're missing the point to get lost in academic details. I don't think we need to look at this from a stand point of test validity and what journal we can lobby it into.

Take a look at the numbers, the commonalities in the sample and infer what you can for your situation. Utilize what happen and what was noted to analyze your own game. The fact is, ugly or good, some of the samples are examples of a higher level of conflict and potential stakes that most people studying martial arts have been in. There's value there, and in looking at patterns.

Take them with a grain of salt.

[Liver Punch]

The ethics are questionable but a better control group would be to pull people off the street and then match attributes like height and weight and make them fight whilst you watch

I vote for this.

[DWx]

The ethics are questionable but a better control group would be to pull people off the street and then match attributes like height and weight and make them fight whilst you watch

I vote for this.

heh just for the record, I neither endorse nor condone this.

Wouldn't want this to come back and haunt me when I'm mega-rich and famous. Don't want the papers printing that I suggested "Bumfights 2011".

[bushido_man96]

Most people are not martial artists. This simply means that most fight data involves people who aren't trained. Using the idea that you're more likely to lose if you hit the ground first as a hard and fast rule means that we would be advising Royce Gracie to attempt to strike with Mike Tyson. I feel that this would be bad advice.

The data presented is what it is. Since most of us have enough control to not go out and start fights with anyone, its going to be little more than conjecture as to how things would change when a trained MAist is attacked. I think Liver Punch makes good points here, too; if you are a trained ground fighter, don't let the stats from a fight study dictate whether or not you take someone to the ground in a fight or not.

I haven't read enough of the study here, but I wonder if these fights are "picked" fights, where the two combatants started off in a mouthing off contest, followed with posturing, and then came to blows, or if they were "jumped" by someone not paying attention; a surprise attack. This information could change the dynamics of the study completely.