You know, this is an interesting idea... do an actual study. I mean, I've seen a lot of MMA fights over the years, at all levels from amateur to elite.
So, let's see... factors to consider if we wanted to do an apples to apples comparison.
Street fights: we'd need a sample that is large enough to account for variations in context. Some contextual things I can think of that I'd want to account for:
- Skill level of the fighters
- Number of fighters (or even better, limit the study to one vs one fights)
- Weapons (same as above, limit to unarmed fights)
- Context of the fight (drug/alcohol? schoolyard fight? gang related? road rage?)
MMA fights: Some things to account for:
- Skill level of fighters (amateur, regional promotion, elite promotion, etc?)
- scheduled length of the fight
I think if you can study a pool of sufficient size for each and account for some of these variables, it would be really interesting to see the results.
Some predictions I'd make:
- Fights between two drunk dudes will be a lot shorter than fights between two sober dudes.
- Drugs or alcohol would lead to a higher KO rate, as well, I'm guessing.
- Conversely, I'm curious if a street fight between two sober dudes for some other reason is typically over as quickly. I'm guessing not... but would be interested in finding out.
- MMA fights will get longer the higher up the MMA ladder you go because it's much more likely to see a huge disparity of skill in lower level promotions, even though they try to avoid that. You're also much more likely to see what amounts to a pure grappler vs pure striker or something along those lines, as many people come into the sport with a "base" of some kind.
Anyway, a 22 year old thread, but I thought this was an interesting thought.
I think a study like this would be almost impossible to get right, the way you're describing. Mostly in the street fight areas. Ignore the formatting, copy/pasting and adding my own comments made it a bit weird.
To take all your factors into account:
- Skill level of the fighters
How would we know this? I'd assume most would be found via youtube videos and/or police reports. Can we make the judgment of how skilled someone is. If someone's been training BJJ for years, but tries standup first for whatever reason and does badly, are they not skilled? Would we know about their skills otherwise? What if they're legitimately really good, but get suckerpunched in just the wrong way.
- Number of fighters (or even better, limit the study to one vs one fights)
This one's easy to control for, which is nice. Only issue would be the potential for extra fighters being added in.
- Weapons (same as above, limit to unarmed fights)
Same comment as above.
- Context of the fight (drug/alcohol?
How will we know this? If someone calls someone else drunk in the video, do we take that at face value? If they're outside a bar, do we assume they had already started drinking? Can anyone (sans lab report) confidently state through a video of a fight, likely from a weird angle, if someone is high on coke, pcp, weed, or something else?
If we see someone swaying, are they drunk, or did they get rocked before the fight started?
- schoolyard fight? gang related? road rage?)
These mostly seem easy to tell, except for the gang related.
To add to this- how do we know when the fight started? And what do we consider the start to be? Is it when they start jostling each other, or yelling at each other, or throw the first punch? If we're doing a comparison to MMA, are we using the same measurements (ie: if it's when the first punch is thrown for the streetfight, do we not count any circling that happens in an MMA fight as time)?
Ideally if this was to be a study, we'd have rows for each of these with a yes/no, or numerical value indicating severity, but if the researchers are just independent observers guesstimating those values, it's ultimately useless. The results/discussion sections of the study would tell us more about the researchers and/or observers (who are hopefully different than the researchers) biases, then it would give us actual information.