|

Data Scientist Analyzes Close Matchups at UFC Columbus

In last week’s Sunday UFC IcePack Analysis (UFC London Sunday IcePack Analysis – YouTube), we talked about how BCDC is 10-2 with over 7.2U on these analyses by our data scientist. This weekend’s fights are great to look at from a data perspective, as, both matchups are toss up fights and have exciting, well-known fighters.


Matt Brown vs Bryan Barberena

This is going to be a slugfest! Analysis done.

I am kidding, but I think we all will be shocked if this one doesn’t end in a knockout. Let’s start by looking at some trends.

Matt Brown is 1-5 against fighters with 6 or more for knockdown total. However, Brown has the better power values of 71 strikes compared to Barberena’s 158 strikes to get a knockdown. As the red corner fighter, Brown wins 79% (15-5) of his fights. Matt Brown is 6-3 as the favorite and red corner fighter. I know these types of trends might not seem to mean much, but when you have a fighter who has fought 28 times… any trend should be considered. The guy has only won one fight in the blue corner. I have no idea why red corner works, but it clearly works. Let’s run the famous and trusty Bayesian Simulation for this fight.

BCDC is 58% sure that Matt Brown is the better fighter with a risk of being wrong at 11%. It is -110 odds, so we have edge available here unless the bookies increase the price to over -130. My only concern is his chin, and Bryan Barberena having no bad trends against him… but he doesn’t have any good trends either. I think you should get Brown while you can and watch for the machine learning prediction, because it’s a gold mine right now. We predict Matt Brown to win by finish in the second round.

Ilir Latifi vs Aleksei Oleinik

The graph below shows Latifi compared to Oleinik on a control time, reversals, knockdown differential, submission attempts and UFC wins. We can easily see that, if this is a standup fight, it will go to Ilir Latifi. If it ends up being a grappling match, it will go to Aleksei just based on their career stats. Latifi has never been taken down in his career, but does he want to force the fight to the ground? Our AI models say no. The fight will go the distance and be a striking fight. Let’s look at their decision history and who wins decisions more often. Who carries more favor from the judges? Latifi is 3-3 when he goes to a decision, and Oleinik is 1-2 when he goes to a decision. Neither fighter seem to be particularly beloved by the judges, but one is slightly better than the other. I believe in the models saying it’s going to be striking fight that goes to a decision, because these fighters don’t have extreme cardio that will dominate the other fighter. The winner will be the fighter who has the more impactful shots. This fight is such a toss-up, but the biggest takeaway from these fighters right now is that Latifi has a 100% takedown defense… This gives Latifi o chance of submission on the ground. Latifi has above average power with only 54 strikes to get a knockdown on average. Neither fighter has above average cardio and both are not particularly successful in a decision. A strong, impactful shot may or may not be the key moment for the judges and the fight.

These fighters usually perform to what the sportsbook projects of them in the UFC. Oleinik is 6-6 as an underdog, and Latifi is 5-2 as a favorite. This shows a 10% edge for both fighters at their prices. Right now, Latifi’s odds are -170, which equates to ~62% and he wins 71% of his fights as a favorite. Oleinik has edge too, with 50% winnings as an underdog. His price is +150 which is 40%.

Latifi will get a knockdown and win a terrible split decision because he does better in the eyes of the judges, has the power, and will dictate where the fight goes by his defense/wrestling. The sportsbooks seem to be so on point providing a 10% edge for him as a favorite. We predict Latifi to win by decision.

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

Similar Posts

One Comment

Leave a Reply

Your email address will not be published.