|

Data Scientist Previews 50/50 UFC Fights

There are two title fights this weekend, and we have some pretty solid choices. Our data scientist, Brandon Bergstrom, previews the two championship fights and an extra fight that is the closest 50/50 fight for the totality of the card. We will touch on this fight first and then go through the championship fights.

Brown vs #10 BCDC Welterweight Williams

Power, power, and more power will be the key to this fight. Khaos Williams has ridiculous power. In the below graph, Williams averages 40 strikes to get a knockdown, whereas Brown is below average at over 120 strikes. Randy Brown is 4-3 against fighters with 70 or less for Sig. Str. Per Knockdown. This shows me that Brown struggles with powerful fighters… which again is the key to this fight.

Let’s take a look at another stat for these fighters. Khaos Williams is 4-0 against fighters with 5 or more for Total Strikes Landed per minute. This tells me that Khaos can put the power on someone and handle the kind of pace that Brown has in the octagon. Will Brown spend energy in trying to get a takedown? Below shows the general skills of these fighters.

See the comparison graph to help identify who will win this fight. This graph shows how each fighter does across multiple statistics, which might help you see who has an edge.

I do not see any glaring weakness other than that Khaos wants the fight standing due to his power. If Brown gets the takedown, maybe it’s a unanimous decision win for him. Our models are projecting this fight to be a ground fight and for it to go the distance. This clearly favors Brown to win. However, I will be going against the models and picking Khaos Williams to win by knockout or decision

#2 BCDC Lightweight Oliveira vs. Gaethje

This is another fight that is going to come down to power and will to win. This should be a great fight. It will really show how good Oliveira’s striking is these days, or it will show that Gaethje’s defense has gotten much better. This fight will end by knockout or submission. Our model projects this fight to be a striking fight and for it to go the distance! If you trust in our distance model, the price is about +330 right now for it to go the distance.

Here are some stats on these fighters. Justin Gaethje is 4-1 against fighters with 113 or less for Sig. Str. Per Knockdown. Charles Oliveira is 15-7 against fighters with 174 or less for Sig. Str. Per Knockdown. These fighters both have handled similar power fighters before.

Charles Oliveira quality control chart on strikes per knockdown compared to the usual fighters. Charles is showing more and more power throughout his 28 fights in the UFC.
Justin Gaethje quality control chart on strikes per knockdown compared to the UFC fighters. Justin is below average with power. Falling in the upper control limit and higher which tells us its not the same as it once was.

Justin Gaethje is 0-2 against fighters with 47 or more for Control Time. Gaethje struggles against the best fighters. He is good but not great. I hope he ruins that narrative, but he has lost to the top tier fighters in the UFC (Poirier and Khabib). Oliveira does have a suspect chin, and it will truly depend on his ability to handle the destruction.

Using Bayesian Simulation and not machine learning models. beforethecagedoorcloses is 64% sure that Charles Oliveira is the better fighter with a risk of being wrong at 10%. -195 odds so huge edge available here unless the bookies increase the price

Here are how these fighters stack up in historical odds in the UFC.

Justin is 4-2 as an underdog and Charles is 14-5 as a favorite.

I see Oliveira’s chin as hugely suspect, and you cannot keep getting this lucky with no one finishing you after getting clipped. Gaethje may have lost power, but he has gained IQ since his fight against Poirier. I will be picking Justin Gaethje to win this fight.

#1 BCDC Strawweight Namajunas vs. #10 BCDC Strawweight Esparza

See the comparison graph to help identify who will win this fight. This graph shows how each fighter does across multiple statistics, which might help you see who has an edge.

What do you see here? The only notable stat is that Namajunas has a bad takedown defense. Our model projects this fight to go to the ground but not go the distance. So, does Rose Namajunas get taken down in this fight and lose by submission? Or does Rose Namajunas get taken down a lot but gets up with another highlight reel knockout?

Both fighters took on Joanna Jedrzejczyk. Namajunas won by Unanimous Decision on April 07, 2018. Esparza lost by KO/TKO on March 14, 2015.

Namajunas handles adversity better than most in the division, and there is a reason she is #1 in the weight class. However, can she avenge the loss to Esparza and stuff the takedowns? Esparza’s reach is 63 inches. Namajunas is 6-2 against Womens Strawweights with a reach of 63″ or more. A fighter with similar odds of 195 to 155 and a reach of 63 inches is 3-7 in the last 10 fights.

Lastly, I’ll look at the Bayesian simulation model to provide more info for this fight.

Using Bayesian Simulation (not machine learning models), BCDC is 62% sure that Rose Namajunas is the better fighter with a risk of being wrong at 11%. Odds are -210, so no edge is available here. However, check back on Saturday for our machine learning picks. For now, I am going to go against the data here. Rose is not very good on her takedown defense and did better with Weili Zhang. I’m saying it will go to a decision win for Carla Esparza. Carla will win due to control time and successful takedowns. I pick Esparza at +175 to win the fight.

Share this…
Share on Facebook
Facebook
Tweet about this on Twitter
Twitter

Similar Posts

Leave a Reply

Your email address will not be published.