A Detailed Analysis of UFC Submission Rates

No idea and there aint no way in hell Im going back through all those results!! hehe

I just dont see the relevance of including tapping to strikes. I wanted to measure the rate of BJJ choke / joint manipulation finishes. If somebody rolls onto his stomatch and taps to avoid getting pummelled or pops his shoulder and taps after a takedown, its not relevant to me. There were a few of them scattered here and there but they wouldnt have changed anything significantly and the underlying trend (or lack thereof) would still remain.

You one upped me!! Great little graph. I probably could have just posted that and been done with it!!

Again, some clarification as to why only the numbered main cards. I got thinking about the lack of submissions due to those 3 main UFC cards not long ago that had 0 submissions. Already I was thinking solely about the main events so the extention of that was to just record the main event details.

Im glad Mike Fagan pointed out the UFNs though coz its interesting to see the steady submission rates they have and how close the average there is to Dream.

Focusing on the numbered events is informative. You can motivate it this way: if a fighter could choose where his bout would be, would he pick a UFN or a numbered UFC?

This is a decomposition exercise. We already have the analysis on the aggregated data. But aggregated data mask trends that may differ among categories.

I think of it this way, by analogy. Say we want to understand the state of education in the United States. For sake of argument, suppose SAT scores are a good measure of how well-educated kids are. (Again, this is just a simplification for illustrative purposes.) Suppose the national average SAT score for high school juniors is level for some time period. Would we stop there? Probably not. Looking at average SAT scores over time for different groups of people could be informative: rich and poor; black, white, hispanic, asian, etc.; males and females. Differential trends across groups suggest that there’s more than meets the eye in aggregate statistics.

Originally I didnt specifically exclude them. It just so happened that I only recorded the data from the main cards. Sadly, its that simple. The data then seemed to show declining rates which drew my attention. Then Mike Fagan pointed out the high number of submissions on UFN cards, so I gathered that info too. The fact that the UFN submission rates are more consistent and much higher is also interesting. What does it all mean?!?! Who knows. Thats why I posted it. :)

Re: (1). Mike, no matter how you cut it, your choice of words was poor.

This is a blog. This should be fun. You’re treating it like life-and-death.

To put it in perspective, this is not a university-level empirical seminar (or even a good SABR meeting). If it was, you’d get annihilated. Your “Power Law” post didn’t show what you thought it showed. Your claim about the regression to the mean for submission ratios rests on a dubious understanding of statistics. Would you have thought about an AR(1) process? Would you know the law of large numbers or the central limit theorem might apply? If you don’t, that’s fine. I personally don’t care as long as your tone is commensurate with your understanding (and everyone else’s).

Like I said, this is a blog. Relax.

In reponse to your points:

1) I have in no way been dishonest or attempted to mislead anybody. I dont have to declare a hypothesis when I present some figures. My comment on your post was purely as you said: “Whoa, not so fast, look at these numbers”. Thats all it was. My opinion on your post. The stats I had didnt agree with your post, so I thought I would contribute.

2) You are correct. My objective is muddled and ambiguos. I dont have a rock solid hypothesis or anything. I just gathered the data to see what it would tell us. When I first started gathering the data a few months back, I was also looking to compare the rate of decision finishes between the UFC and Dream. Also, I used the data I stored to look at the KO rates in the UFC compared to the MMA KO rates that the John Hopkins study found a few years back. I might post another topic on that some day as I have some more interesting info! Anyways, the point is, I just gathered the stats and looked at what turned up. In this case, its a decling submission rate in main UFC cards and we can all speculate as to why. I enjoy doing this and dont feel the need to formalize some crystal clear objective and declare any official hypothesis before I post anything.

3) Call it what you will. The percentages seem to show a consistent reduction in the number of submission finishes. I encourage you, like all the others on this forum, to draw your own conclusions as to why and share with us to discuss.

A random sample would give us less info. As it stands we currently have every single fight since UFC 50. We can go back prior to that but it wouldnt achieve anything. Whether the submission rate was climbing or falling rapidly or slowly before then has no real bearing as far as Im concerned. If anything it could reduce the validity of the data. Of what relevance is Royce Gracie submitting no names to the trend over the last 450 main cards?

I’m interested.

Can we just barricade ourselves in this litle corner of BE and discuss this rationally like adults without any risk of a smear campaign? Just the 3 of us……whaddya say gents?

I graphed the running average rate for fights decided by decision over a 10 main, numbered event window. Exactly the same methods for that I used for the submission rates. Check out the graph below. (its linked to a clearer version)

It says to me, that you are trying to discredit Mike’s findings and present your own as some authoritative piece. Why not just present them as is without the jab?

Im not trying to discredit Mike’s findings. His figures appear accurate when you look at all the events, but I question the validity of his conclusion. My main claim is that when looking at the main card submission rates, its hard to completely write off the suggestion that submissions are falling. At the very least, the jury is out, imo.

Also, there is no jab at Mike. I was hijacking his thread and as can be seen from the data I posted, its worthy of its own fanpost. The opening paragraph was just a background as to how this post came about. If I wanted to have a go at Mike I would have tried to change the data around and make a convincing case that I am right and he is wrong. I havent done that and have at numerous times in both the fanpost itself and the comments section given Mike credit for his work.

Also, when you do things like leaving out non numbered UFC events, it reeks of trying to manipulate results to fit your own goals. You shouldn’t be surprised when people assume as much. Why not only use numbered events with PPV buys of over 500k or 750k? You could assume the cards with the bigger names, therefore bigger buy rates, would produce even more pressure on the fighters or whatever the justifications are.

If using the biggest cards from the biggest org ‘reeks of manipulation’, it truly is a cynical world we live in. Using numbered events with PPV buys over 500k or 750k would be interesting. Why didnt I produce those stats? As Ive said all along, Im curious to see what people think when faced with all the info. 750k+ ppv buys would be a very small sample and not very helpful when trying to identify if a consistent trend exists.

I also have a problem with not including submission by strikes

Ive responded to this query previously in this post. Submission by strikes or injury isnt relevant. If you think it is silly you are obviously missing the point.

The proper way to figure out what may be causing the numbers to drop in main events is to use the numbers in conjunction with I don’t know, actually talking to the fighters and promotions

Thanks for that. Here I was trying to discuss these stats on an internet forum when all along the ‘proper’ way to get an answer was staring me in the face!! How could I have been so blind!! Ill call Dana and Chuck now. They’ll tell me why!!!

Is this at me? If so, you may want to actually try and say something in a response if you plan on sticking around here for a while. If you have an opinion on what I said, then say so. I don’t need your permission to post here or your stamp of approval for acceptance. Grow up.

Where have I said Mike is wrong?!?!? I have gone to GREAT lengths, in the face of his unfounded accusations of dishonesty, not to belittle his work. In my very first response to Mikes piece all I said was that its a little early for “I told you so’s” as the data is at best inconclusive. I have said multiple times that I dont disagree with Mikes numbers and I am at pains to differentiate that my submission rates are for main, numbered UFC events and that UFN/TUF events have a higher rate of submissions.

ANOTHER ONE!!!!!

How about looking at the CONTENT of what I am posting. Not running around trying to identify potential contradictions or insignificant errors in my post. Im surprised nobody has corrected my grammar yet!!!

If you didnt notice from your quote about, my frustration is not directed at Mike. Its at people like you who CONTINUE to nit pick and split hairs about trivial matters and ignore any of the substance……..which took me quite some time I might add……of the post.

THANKS RINEY!!!!!!!!!!! Thats one of the components of my theory explaining the falling rate line.

That would then be the promoters going out of the way to push for submissions, which is just as bad as paying Petruzelli to stand.

Because it’s warranted. I didn’t think GeeDub was getting a fair shake the first few times around. FWIW, my tone is in the ballpark of the comments of the original post, which played a big part in this new post.

And I’ve made plenty of statistical points. You can check all the threads.