UFC 215: Amanda Nunes vs. Valentina Shevchenko alternative stats and win probabilities

It’s that glorious time of year again when MMA’s pound-for-pound king throws down in the Octagon while very few people seem to care… and while this piece was being written, news of Ray Borg’s illness broke.

Dammit.

So Demetrious “Mighty Mouse” Johnson’s attempt to break his and Anderson Silva’s 10-fight record for UFC title defenses will have to wait. On the plus side, he and Borg will get to move to a jurisdiction not regulated by the Edmonton Combative Sports Commission where god knows what might happen. Let’s just hope UFC 215 will still be a fun night with medically-fit fighters that don’t take any unnecessary damage.

In the immediate aftermath of the disgusting white nationalist rally in Charlottesville, VA and their associated violence, I wanted to purge the name of these statistics to remove any possible association with the alt-right movement. But after reflection, it seemed the Alt Stats abbreviation was generating the primary connection (at least in my head). So the alternative stats name will be retained, but never shortened.

Remember, what you’re about to read are not official UFC statistics. They’re alternative stats designed to (1) give more weight to the recent present than the distant past and (2) not let one huge or horrible performance dominate the data.

Since this piece is being published on the afternoon of UFC 215, win probabilities are listed along with each fight and possible bets are discussed at the end. See the notes at the bottom for definitions of particular statistics employed.

Amanda Nunes (69.1%) over Valentina Shevchenko (30.9%)

Nunes’ alternative stats have certainly benefited from beating the snot out of former champions Ronda Rousey and Miesha Tate in her last two outings. She spends almost three of every five minutes at distance to Shevchenko’s 1:33 with the fight sometimes going to the ground from lower body takedowns for both ladies (and Nunes’ knockdowns in 18.4% of rounds). While almost six minutes of their first meeting at UFC 196 was spent on the ground, it was mostly a standup affair but for a Nunes takedown in the 2^nd round off a Shevchenko kick she probably wishes she could take back and a Shevchenko counter-trip in the 3^rd.

At distance, “The Lioness” Nunes mixes things up a bit better landing 5.3 and 5.5 power strikes per 5 minutes (P5M) to the body and legs, respectively, to Shevchenko’s 4.6 and 0.7. Alternative stat differences to the head are enormous. Nunes lands 38.2 power shots to the head P5M and 27.8 more than her opponents while Shevchenko’s more methodical, counterstriking style leads to 10.1 head shots and 6.7 more than her opponents. One their paces will probably win out and it could very easily end up a more Shevchenko-type pace similar to the last time, but either way, both fighters have been exceptional at out-power-striking their opponents at distance (and Nunes as well for head jabs).

Neither fighter is a distance takedown shooter, with both preferring to do that work from the clinch. Shevchenko’s various clinch trips have led to an above average alternative stats lower body takedown success rate of 65%. Nunes comes in at 37% but has much better defense (especially when she’s not tired) of 82% defended (vs. 56% for an average women’s bantamweight)

If the fight heads to the ground, we’ve already seen what Nunes can do – demolish opponents and possibly gas herself out if she doesn’t get the finish. She rains power shots down on opponents with 154% more volume than average, landing with 120% more frequency, and finishes 51% of her submissions with a 71% higher sub attempt rate than average. When Shevchenko’s on bottom, she’ almost exactly average at standing up and sweeping but is 273% worse than average at keeping opponents from standing up when she has top position.

Once the cage doors have been opened and closed a few times, this fight could very well come down to something not presently included in the model: The gas tank. I’d love to include a metric for cardio but it’s hard to find one that seems reliable while not also creating a bunch of missing observations waiting for fighters to get enough time in certain positions. As an example, Nunes has had nine fully-documented bouts and still hasn’t yet spent a grand total of two minutes at distance in the 2^nd round. But she has decent data for the 3^rd round.

Nunes’ gas tank looked questionable in the later rounds of her first fight against Shevchenko and her UFC 178 loss to Cat Zingano. By the numbers, Nunes’ distance power strike attempts drop off 51.6% between the 1^st and 3^rd rounds. Shevchenko’s, on the other hand, increase 51.8%. In fact, Shevchenko gets stronger in the 3^rd, 4^th, and 5^th rounds on every distance power metric I examine.

The last two times Nunes gassed it was from working over an opponent from top control. Has she learned that she either needs to finish quickly or conserve energy, especially in a five-round fight? We’ll find out soon.

The pick is Nunes, cardio correction not included.

Neil Magny (51.1%) over Rafael Dos Anjos (48.9%)

If it’s a Neil Magny fight, there’s a good chance it’s going to be somewhat close – that is, when he’s not 10-7ing your face into oblivion down under. On the contrary, RDA’s bout closeness score – win or lose – is 30% lower than average.

Magny’s recent takedown defense has been pretty awful. His lifetime takedown defense from distance shots is 54% but in alternative stats world it’s a pathetic 11%. True, Johny Hendricks and Demian Maia were some recent opponents, but the good news is RDA shoots for takedowns at only an average rate and has been worse than average completing them (21% landed vs. 30% welterweight average)

Magny kills opponents with head jabs at distance to the tune of 10.0 more landed P5M. Both RDA and Magny out-land their opponents to the head and body with power, but Magny’s been getting killed to the legs recently by Silva, Gastelum, and Larkin, and RDA tags the legs with power 82% more than average

The knockdown game isn’t even close. Magny gets clubbed to the ground in just over ¼ of his alternative stats rounds and his rate of getting knocked down is 197% worse than average. RDA is about an average knockdown artist (or slightly subpar) but has only been tagged to the canvas once in his documented career.

If Magny can get to the clinch, he’s 13% better than average landing takedowns and RDA’s 16% worse defending. Magny’s been a beast from top control, punishing opponents with 82.3 power strikes landed P5M, a whopping 477% more than average and 388% more than RDA.

It would be easy to just not pick close fights when the predicted winner’s in the 50-53.9% range (similar to a poker race), but where’s the fun in that?

This is basically a coin toss with a very slight edge going to the pick of Magny.

Henry Cejudo (66.3%) over Wilson Reis (33.7%)

We think of Cejudo as a power-punching guy, but he’s not yet won a single documented fight by KO/TKO, and neither has Reis.

Reis’ standup is not good. In alternative stats at distance, he gets out-landed by opponents in jabs to the head and power to the head and body. We’re talking 3.1 fewer head jabs P5M and 4.8 fewer overall power shots (including legs where Reis has tended to out-land).

Reis loves the takedowns, shooting in from distance 137% more than average and with a 90% higher success rate. From the clinch, his success rate isn’t special but his volume’s 124% more than a typical men’s flyweight. On the ground, he has the top 78% of the time and finishes 34% of his submission attempts (twice the average rate of 17%).

Reis’ problem is that he’s going against a former Olympic gold medal wrestler, Cejudo, who’s never been taken down and has legit striking chops.

Cejudo’s not much of a jabber but he puts out slightly more than average power volume at distance and lands 3.4 more power strikes than opponents P5M. He gets knockdowns in roughly 1/6 of rounds and has a knockdown percentage that’s 53% better than average (and infinitely better than Reis’ 0.0%). And, to repeat, Cejudo’s never been taken down in five documented attempts.

Cejudo will likely want to keep the fight standing to avoid Reis’ submission game and play towards his relative advantage, but should he choose to go for a takedown, he’s not very good at distance and roughly average in the clinch. If the fight ends up on the ground, Cejudo’s been on top 99% of the time and is a reasonably active power ground and pounder.

The pick is Cejudo, possibly even to get his first KO/TKO finish. Reis has been knocked down four times, each by a different one of his nine documented opponents.

Gilbert Melendez (61.0%) over Jeremy Stephens (39.0%)

It’s a battle of losing streaks! Stephens has lost two-in-a-row and five of his last seven fights while Melendez has come out on the ass end of three-in-a-row and four of his last five.

Neither of these guys really likes the takedown (except sometimes if they get clinched up) and both are solid at not getting knocked down. If the fight stays standing at distance, Melendez’s head defense is about 26% better than Stephens and he connects with power to the dome 3.3 strikes more than his opponents P5M while Stephens lands 0.2 times fewer.

Stephens certainly has knockdown and knockout power in his pocket, but the pick is Melendez.

Alex White (59.7%) over Mitch Clarke (40.3%)

Current Bankroll: $9,850.50

The live model is two bets in and has lost $50 on Demian Maia vs. Tyron Woodley at UFC 214 and $99.50 on Bryan Barberena vs. Leon Edwards – always a fun start.

When backtesting the model from Jan. 2014 – Jun. 2017, the $10,000 starting bankroll got down to $9,415.72 after two months and 14 bets of action. While it would certainly be nice to get off to a better start this time, edges in MMA – with a few picks each event and even fewer bets – play out over years.

The model has two bets on this card with $98.51 on Nunes to win at +110 and $24.63 on Magny to win at +170 as of this writing.

Bring on the glorious fights!

This is an experiment and entertainment. Do not bet on the fights using these numbers. Taking sensible gambles with an edge over time is similar to investing. But if I’ve made a mistake somewhere, those same gambles without an edge become sucker bets. Let me be the possible sucker. You've been warned.

Notes: Strike attempts are for an entire five minute round in each position (P5M) and are categorized as jab or power. A jab is just a non-power strike. Strikes are documented based on where they land or are targeted (head, body, legs), not the type that is thrown (punch, elbow, kick, knee). Visible damage rate is per five minutes the fighter is not on his back. It’s hard to bust up someone’s face while lying on your back. Damage percentage is per power head strike and distance head jab landed. Knockdown rate is per five minutes at distance or in the clinch off the cage. Knockdown percentage is per power head strike landed while standing. It’s really hard to knock someone down if they’re already on the ground. Clinch control is having the opponent pressed against the cage. Ground control is having top position. Submission attempts are per five minutes of ground control minus time spent in the opponent’s guard plus time spent with the opponent in guard. A bout closeness measure towards zero means a fighter is in blowouts (win or lose) and towards 100 means he is in very close fights.

Paul covers MMA analytics and business for Bloody Elbow. Follow him @MMAanalytics. Fight data provided by FightMetric.