Jackson mma's ACTUAL finishing rate in 2012

[ThunderStruck]

Uhm, no. I've been watching MMA for 10 years, and after an event I check which fights ended in finishes. Unless I like the fighters or there's some buzz it was a good scrap, I pass on the decisions. That's what the sport has to deal with. Blogs which posted that stat need to retract the article.
What a troll job though. Awesome.

Just bleed fan since 05 am i right?

 

[Saskatoonmike**]

Look at the threads about this chart. They are all about how Greg Jackson fighters are boring because of his low finishing rate. That's why including losses is subterfuge.

I am simply pointing out that if you do not include losses the stat is meaningless.

Fighter 1 : 10 fights. 8 wins by finish, two wins by decision.
Fighter 2 : 10 fights. 9 wins by finish, 1 win by decision, 1 loss via finish.
Fighter 3 : 10 fights. 1 win by finish, 9 losses by decision.
Fighter 4 : 10 fights. 1 win by finish, 9 losses by finish.

By not including all fights including losses Fighters 3 and 4 are the most exciting with the best Finish rate. Absurd.

You cannot run away from that flaw when you exclude losses.

If, as the TS says the goal is to predict who is more exciting. that is fighter 1 and 2 and not 3 and 4 which his method says.
Fighters 1 and 2 finish more and are more exciting but

 

[flanellograf]

logic or any math beyond highschool will tell you these statistics are useless

dont get me wrong its not completely useless but just for the most part. not nearly enough fights to get anywhere near accurate statistics. training different fighters is comparing apples to oranges. a guy like brain stann could ko 3 guys in 2011 then face chael and bisping in 2012. there are way too many factors that these stats dont reflect for the info to be valuable

at some point this data (or similar) might be useful though

First, the thread is not really about whether it's a valid way to judge anything, but merely refuting erroneous data.
Second, the data is obviously only good to give a you general idea or what the trend is. You can't really prove anything conclusively with it, no.
Fighters are obviously different, and it also depends on the general talent in the camp, but you can at least infer something about a camps way of approaching a fight and what they teach their fighters. All the fighters share trainers and overall philosophy.
If you wanna be statsy about it, you can't really say anything about fighters skills anyway. The sample size is usually abysmally small in mma, with few fighters having over 20 fights, and even fewer having over 30, and even less shared opponents (also fighters grow and change, so they can be quite different from one year to another, so cross-comparisons are also hard).

 

[Rolento]

Does Greg advise these dudes to only fight like once or twice a year? I never realized how little they all fight.

 

[flanellograf]

I am simply pointing out that if you do not include losses the stat is meaningless.

Fighter 1 : 10 fights. 8 wins by finish, two wins by decision.
Fighter 2 : 10 fights. 9 wins by finish, 1 win by decision, 1 loss via finish.
Fighter 3 : 10 fights. 1 win by finish, 9 losses by decision.
Fighter 4 : 10 fights. 1 win by finish, 9 losses by finish.

By not including all fights including losses Fighters 3 and 4 are the most exciting with the best Finish rate. Absurd.

You cannot run away from that flaw when you exclude losses.

If, as the TS says the goal is to predict who is more exciting. that is fighter 1 and 2 and not 3 and 4 which his method says.
Fighters 1 and 2 finish more and are more exciting but

These are statistical anomalies, and obviously should be excluded to not skew the data, should they exist. You could for instance have some condition for including data.
The question is how relevant these outliers are when you just do a general comparison, using fights from a lot of different fighters.

 

[theheater]

First, the thread is not really about whether it's a valid way to judge anything, but merely refuting erroneous data.
Second, the data is obviously only good to give a you general idea or what the trend is. You can't really prove anything conclusively with it, no.
Fighters are obviously different, and it also depends on the general talent in the camp, but you can at least infer something about a camps way of approaching a fight and what they teach their fighters. All the fighters share trainers and overall philosophy.
If you wanna be statsy about it, you can't really say anything about fighters skills anyway. The sample size is usually abysmally small in mma, with few fighters having over 20 fights, and even fewer having over 30, and even less shared opponents (also fighters grow and change, so they can be quite different from one year to another, so cross-comparisons are also hard).

the info doesnt tell us anything except that we now know that statistically we still have no evidence. which is something i guess

sometimes conducting a study only tells you that you need to conduct more studies, which is something. i think that kind of sample size could possibly give interesting data on which camp has more subs, ko's, and decisions. but also maybe not.

like i said that study didnt demonstrate anything so in a way i guess it showed people convinced one way or another that its really inconclusive still

 

[flanellograf]

Does Greg advise these dudes to only fight like once or twice a year? I never realized how little they all fight.

Lot's of injuries could be the answer. Think most of them fought more last year.

 

[Saskatoonmike**]

logic or any math beyond highschool will tell you these statistics are useless

dont get me wrong its not completely useless but just for the most part. not nearly enough fights to get anywhere near accurate statistics. training different fighters is comparing apples to oranges. a guy like brain stann could ko 3 guys in 2011 then face chael and bisping in 2012. there are way too many factors that these stats dont reflect for the info to be valuable

at some point this data (or similar) might be useful though

Yup. As a guy who majored in math with a statistics minor I can tell you that calculating the Finish Rate as a percent of wins only would receive a failing grade.

Doing that way says these guys are equally exciting and the exact same when it comes to finish percentage.

Fighter 1 : 20 fights. 16 wins via finish. 4 wins via decision. 80% Finish Rate.
Fighter 2 : 20 fights. 8 wins via finish. 2 wins via decision. 10 losses via dec. 80% Finish Rate.
Fighter 3 : 20 fights. 4 wins via finish. 1 win via decision. 15 losses via dec. 80% finish rate.

If you cannot see how wrong and skewed the stats get via this method you simply do not understand even the most base math.

If your goal is to determine the most exciting fighter that is not fighter 3 and yet the stat (finishes over wins only) says he is right up there.

 

[Ughugh]

I stopped reading at "Brian Stann" because he's not had his last two camps at Jackson's.

If you're gonna fact check and be a dickhead about it TS then at least make sure your own facts are in order you dummy...

 

[theheater]

Yup. As a guy who majored in math with a statistics minor I can tell you that calculating the Finish Rate as a percent of wins only would receive a failing grade.

Doing that way says these guys are equally exciting and the exact same when it comes to finish percentage.

Fighter 1 : 20 fights. 16 wins via finish. 4 wins via decision. 80% Finish Rate.
Fighter 2 : 20 fights. 8 wins via finish. 2 wins via decision. 10 losses via dec. 80% Finish Rate.
Fighter 3 : 20 fights. 4 wins via finish. 1 win via decision. 15 losses via dec. 80% finish rate.

If you cannot see how wrong and skewed the stats get via this method you simply do not understand even the most base math.

If your goal is to determine the most exciting fighter that is not fighter 3 and yet the stat (finishes over wins only) says he is right up there.

lol yeah his study was pretty useless and i didnt even look at the terrible methodology. good points

 

[Saskatoonmike**]

These are statistical anomalies, and obviously should be excluded to not skew the data, should they exist. You could for instance have some condition for including data.
The question is how relevant these outliers are when you just do a general comparison, using fights from a lot of different fighters.

lol. Take a statistics class. It is the anomolies you search for to see if the data is being presented reliably./

FACT is if you were to calculate your way and line up every Finisher into baskets of 100% down to 1% anyone and everyone below a guy above could be a better finisher and more exciting.

I could give you close examples as easy as extreme ones. Your method fails every single test.

Here is a general comparison of two otherwise very good fighters.

Fighter 1 : 20 fights. 18 wins. 15 via finish and 3 via decision. 2 losses.
Figher 2 : 20 fights. 14 wins. 12 via finish and 2 decision and 6 losses.

If your goal is to see an exciting finish. If your goal is to use the stat of FInishes to help predict that... then the way you calculate the stat is still useless as the worse finish is ranked above the guy who has proven to be a better finisher.

You can never escape that potential with your method. But if you use Finishes as a percent of all fights then it 100% gives you the right information.

 

[Saskatoonmike**]

Further proof of why flanellograf's, method is so flawed.

Fighter 1 : 20 Fights. 10 wins via finish, 10 wins via decision.
Fighter 2 : 20 fights. 5 wins via finish, 5 wins via decision. 10 losses via decision.

flanellograf's method say these guys are equal in Finishing Percent and equally exciting. But if Fighter 1 comes to flanellograf's and say's how can I become a better Finisher and raise my finish rate, the easiest answer is to tell him to start losing those decisions.

Yup to gain a better finish rate, via flanellograf's method you never need to win more fights via finish. Simply losing more fights via decision makes you a much better finisher. If fighter 1 had instead lost those 10 decisions like fighter 2 did, then Fighter 1 becomes a 100% finisher.

Imagine that. Losing by decision makes you a better finisher then winning by decision, and actually winning by finish is not necessary to improve, when you used flawed stats.

 

[Ughugh]

Further proof of why flanellograf's, method is so flawed.

Fighter 1 : 20 Fights. 10 wins via finish, 10 wins via decision.
Fighter 2 : 20 fights. 5 wins via finish, 5 wins via decision. 10 losses via decision.

flanellograf's method say these guys are equal in Finishing Percent and equally exciting. But if Fighter 1 comes to flanellograf's and say's how can I become a better Finisher and raise my finish rate, the easiest answer is to tell him to start losing those decisions.

Yup to gain a better finish rate, via flanellograf's method you never need to win more fights via finish. Simply losing more fights via decision makes you a much better finisher. If fighter 1 had instead lost those 10 decisions like fighter 2 did, then Fighter 1 becomes a 100% finisher.

Imagine that. Losing by decision makes you a better finisher then winning by decision, and actually winning by finish is not necessary to improve, when you used flawed stats.

Yeah he's a certified idiot.

How about

Fighter 1: 34 fights. 4 wins. 2 wins via finish and 2 via decision. 30 losses.
Fighter 2: 17 fights. 15 wins. 5 wins via finish and 10 via decision. 2 losses.

According to TS genius "fighter 1" would be the better finisher (50% vs 33%).

And if he's been finished in his 30 losses then he'd be off the charts...

 

[JudontknowMMA]

can TS go back and see if the author for some reason only used main card fights for statistics? would kinda suck if so.

 

[dc207]

1. I discredited the graph on the sole basis that it's a line graph and not a bar graph. If you don't know how to properly represent the data, there's a good chance you also fucked the numbers up.2. Why would you try to lead people astray by calculating finishes per win? That's closer to representing the need of a "lucky punch" than "finishing ability".

That's a good point, ha. I looked at it and thought, "Something does not look right."

 

[Jblizzle]

I couldn't care less about these numbers or that graph MMA isn't a team sport

 

[Lucifer Alpha]

I am simply pointing out that if you do not include losses the stat is meaningless.

Fighter 1 : 10 fights. 8 wins by finish, two wins by decision.
Fighter 2 : 10 fights. 9 wins by finish, 1 win by decision, 1 loss via finish.
Fighter 3 : 10 fights. 1 win by finish, 9 losses by decision.
Fighter 4 : 10 fights. 1 win by finish, 9 losses by finish.

By not including all fights including losses Fighters 3 and 4 are the most exciting with the best Finish rate. Absurd.

You cannot run away from that flaw when you exclude losses.

If, as the TS says the goal is to predict who is more exciting. that is fighter 1 and 2 and not 3 and 4 which his method says.
Fighters 1 and 2 finish more and are more exciting but

We would never be talking about a 1-9 fighter in the first place.

 

[JoeHogans**]

Alpha male! Dem midgets finishing shit like a man.

 

[TankAbbott4Eva]

I couldn't care less about these numbers or that graph MMA isn't a team sport

I reckon most fighters would disagree with that, they all need a good team so they have good sparring partners.

 

[Jblizzle]

I reckon most fighters would disagree with that, they all need a good team so they have good sparring partners.

In that aspect it is but once they're in that cage it's one on one. It's up to the individual to perform