Upper Dublin High grad Ben Zauzmer lives in Los Angeles now, where he churns numbers for the Dodgers’ analytics group, but he still makes it home for the holidays and the Zauzmer family ritual of watching the year’s Oscar contenders.

“My family has a best-picture marathon every year. I come back to Philly and we get together and watch all the nominated movies,” said Zauzmer, adding that this year, the movie marathon went unfinished. By the time the family had gotten through with The Irishman, everyone was winded.

"We kept pausing so we could talk about it, and by the time we finished, we were like, ‘Look, it’s been six hours,’ ” Zauzmer recalls.

When he’s not making complex sabermetric calculations for the Dodgers, he’s using his math skills to analyze the Academy Awards, a habit he picked up at Harvard and parlayed into a second career.

He can gauge, for instance, whether The Irishman’s prodigious 210-minute running time will help it or hurt it on Oscar night. His model shows that best-picture nominees with a longer running time traditionally stand a better chance of winning, but he also knows that trend has reversed itself in the last 10 years — with shorter movies gaining favor among the Academy electorate.

You can find these and other facts in his new book, Oscarmetrics: The Math Behind the Biggest Night in Hollywood (BearManorMedia), which explains his methods and conclusions in lucid detail, and includes amusing riffs on the quirks of Oscar nominations and selections.

"I don’t really want it to be just a collection of data and numbers. So there’s a lot of stuff in there to make it fun for people to read,” Zauzmer said. (In the Oscar-winning song “High Hopes,” composer Sammy Cahn made sure that every time Frank Sinatra sings the word “high,” the note is higher than the note that follows, all 12 times).

By the way, he’s also able to predict which movies are going to be nominated, which is why his family is able to watch the “nominated” movies a month before the nominations are made.

Before we get to the book, though, a question: Why is avowed Sixers, Eagles, and Flyers fan Zauzmer not using his giant Harvard brain to help the Phillies?

“If you had told the 6-year-old me that the grown-up me would be working for anyone but the Phillies, he probably wouldn’t have believed it,” said Zauzmer, who joined the Dodgers in 2015 just as they were placing new emphasis on baseball math.

“It was just the right fit for me. They were really investing in analytics, and it was a chance for me to be part of something that was new and exciting and had a very strong commitment behind it." he said.

Also a chance to live at the epicenter of the movie business, which gave him a chance to indulge his Oscar habit — Zauzmer writes an annual prediction column for The Hollywood Reporter.

It’s all based on data he started collecting at Harvard — filling out spreadsheets with as much raw Oscar data as he could find (other awards, critics groups, Rottentomatoes.com aggregations, craft guild nominations, etc.), measuring the effectiveness of each element based on its predictive reliability, and reweighting the results every year as new winners come in.

The upshot: a model capable of predicting the outcome across categories 77% of the time, and one that picks up interesting trend lines. Zauzmer documents, for example, what actresses know in their bones to be true — the older you get, the harder it is to win.

"I wasn’t surprised by the fact that it’s true, but I was surprised by the degree to which it is true. Especially for best supporting actor and best supporting actress. Men actually have more of an opportunity to win as they get older,” he said.

On the other hand, a woman can slightly improve her odds of winning if she chooses to play a historical figure. Just not an old one.

Other findings: There really is a “career achievement” bias in Oscar voting. If you’ve never won, your odds improve with subsequent nominations — a 5.9% bump. And higher-billed performers in best supporting actor/actress categories have a definite advantage. Boffo box office used to be a factor that improved best picture chances, but that trend has also reversed itself in the last 20 years (Moonlight!) and is increasingly a nonfactor.

The book is not all about predictions. Zauzmer wrote a chapter that looked at the long-term, enduring (and measurable) critical and popular opinion of Oscar-nominated movies and used it to identify, with the benefit of hindsight, the movies that should have won best picture. According to this model, Citizen Kane should have beaten How Green Was My Valley, Saving Private Ryan should have beaten Shakespeare in Love, and, more surprisingly, Ferris Bueller’s Day Off should have beaten Platoon.

How does the race look this year?

“I won’t have final Oscar predictions until after all the data comes in. The final piece of information will be the BAFTAs on Feb. 2. The acting front-runners already seem secure, with Joaquin Phoenix (Joker), Renée Zellweger (Judy), Brad Pitt (Once Upon a Time … in Hollywood), and Laura Dern (Marriage Story) likely to be favored on Oscar Sunday. Best picture and best director are still up in the air, pending the results of the Directors Guild Awards and the BAFTAs. Right now, I would predict 1917 for both of the top two categories, but that could easily change over the next couple weeks,” he said.

Personal favorite?

“Coincidentally, 1917 was also my personal favorite film of the year. I was blown away by the technical expertise required to make such a gripping and realistic war film with the perception that there were no cuts in the action. But my own movie opinions have nothing to do with my mathematical predictions, which are purely based on the data,” he said.

Speaking of data, all of this “inside baseball” math could be made lucrative. As gambling on the Oscars becomes more widespread (it was legalized in New Jersey last year), the opportunity to apply Zauzmer Oscarmetrics to wagering exists.

He’s not tempted.

"I do look at the betting markets as a predictor of the awards, just like I use the other data sets. The craft guilds, the critics awards, and so forth. I use anything I can get my hands on that is numerical. But I don’t encourage people to bet, and I certainly don’t bet,” said Zauzmer. “It’s extremely frowned upon in the world of baseball.”