www.FightGrub.com
Now a big fad in medicine is Evidence Based Medicine (EBM). The idea being that the boundaries between Western "scientific" medicine vs Eastern medical traditions vs folkloristic medicine don't have to mean much, because all that matters is what works. EBM proponents like to point out that even most of Western medicine is mostly folk practice, in many cases procedures have not been scientifically validated. So the idea is to accept all, but only, those medical practices that have been empirically validated to work as advertised, regardless of source.
A while back I decided the self-defense and hand to hand combat game needs a similar viewpoint. Yeah, sure we have terms like "reality based self defense" and "mixed martial arts" and so on that purport to be beyond style, culture, and economics - just the facts. Only what works.
But I feel they don't have enough data. Sometimes you hear statistics about x% of fights end up on the ground, or y% of fights are initiated with a wild right haymaker and so on. But I don't think a real comprehensive dataset has ever been assembled.
A really good dataset would permit us to issue queries against it, in other words allow for data mining and exploration, just like any large 21st century data collection. It can be stored as a relational database, standard stuff. Then researchers could run queries against it (using SQL but I'll give the queries in quasi natural language for readability), such as:
How many fights in the DB have END_CONFIGURATION = GROUND ?
How many fights in the DB have STARTMOVE = "WILD_RIGHT_HAYMAKER" ?
And we could find out what works best to finish an attacker off:
SORT DESCENDING by TERMINATOR_MOVE
Whatever MOVE code came out on top of the above query was most frequent, and we should train that, be it eye gouge or whatever. We could learn all kinds of things like how often women were able to disable male attackers, what and how improvised weapons are used, whatever attribute of a fight can be given a standard code could be the basis for searching, sorting, querying, collating and counting statistics.
Of course, there remains the problem of assembling the raw fight data itself. Maybe police departments have some records that could be fed in, but I figure nowadays, with all the vids up on youtube and everywhere, we should be able to assemble such a db with just some routine data entry work, off the internet videotapes of street and schoolyard battles. Maybe we have to do some initial pre-filtering, on the basis of guidelines. For example, maybe any fight that appears to have been either genuinely or falsely staged for no other purpose than getting into the DB itself should be excluded. There would be some tricky calls, but no matter, every DB construction effort has issues with raw data sources. We shall overcome!
I'll show you what just one DB record might look like:
URL = http://www.youtube.com/watch?v=M-79K6wpbwY&NR=1
SITE = youtube
TITLE = "road rage knockout punch"
TYPE = 1-1
CONFIGURATION = vehicle
INITIAL_STATE = verbal faceoff
DECISIVE_MOVE = defender right to aggressor face
OUTCOME = knockdown(aggressor)
TAG = "road rage"
Get the idea? Of course there'd be lots more standard tags eventually, to facilitate statistical analysis by experts, and individuals could do their own custom searches as well. Then and only then can we construct evidentially validated martial arts training.
So, in anticipation of that day, I plunked out $9.99 for the multi-year lease on domain registration: www.fightgrub.com.
I figure that's what the website that eventually provides access to the whole dataset will have to be called.
That's "grub" as in Merriam Webster definition:
intransitive verb1 a: to dig in the ground especially for something that is difficult to find or extract b: to search about <grubbed in the countryside for food — Lamp>
What's that you say? Why don't I do it myself? Are you out of your fricking mind? Listen I'm only going to say this once more: I'm a cat. A feline! We don't dirty our paws with this kind of human ugliness. It's only the humans who love databases, love to feel they have reality roped and tied, cornered and corralled, the whole universe Guantanamo'ed in a bright orange jumpsuit. But we cats don't think like that. We are right-brain animals to the core. So I am decisively above the fray. Besides this idea is ahead of its time. I need to wait for the ponderous pendulum of human thought to catch up. But when it finally does, I'll be there, paw outstretched, just waiting to sell my ownership of the domain and hit the beach with the proceeds.