The latest forecasts of Harvard-based Oscar odds-calculator Ben Zauzmer, whose calculations I began paying attention to three years ago, appeared three days ago in the Boston Globe. Nothing startling — Birdman, Inarritu, Redmayne, Moore, Simmons, Arquette, Lubezski — but Zauzmer, who’s now a senior, has an above-average track record. Is it accurate to call him the Oscar race’s Nate Silver? I’m mulling that one over.

Oscar-predicting Harvard whiz kid Ben Zauzmer, who got 75% of his predictions correct last year, doesn’t give a hoot in hell for all this Emmanuelle Riva talk. He’s sticking with Jennifer Lawrence for Best Actress. He’s also predicting (a) Argo for Best Picture, Daniel Day Lewis for Best Actor, Tommy Lee Jones for Best Supporting Acto, and Anne Hathaway for Best Supporting Actress. Bens’ twitter handle is BensOscarMath.

I never rank very high in predicting Oscar winners because I’m psychologically unable to separate or compartmentalize my feelings about the contenders from what I’ve been told or otherwise led to expect will happen. Every year about two-thirds of my predictions are on the money and roughly a third are not. I could have done a little better than 63% (i.e., last night’s final score) if I’d listened to Ben Zauzmer and predicted Meryl Streep to win Best Actress, but I couldn’t push myself off the Viola Davis boat.

The Oscar prediction game is fundamentally naught but simple shit.

This aside, congratulations to Gold Derby reader “snuggle4” (whomever he or she may be) for scoring 88% correct. Ten other Gold Derby users scored 84%. But credit where due, Snuggle4 was way ahead of the top three Gold Derby “experts” who got 80% right — Deadline‘s Pete Hammond, Fox News’ Tariq Khan and Awards Daily‘s Sasha Stone.

Who is Snuggle4? I’m asking. Why doesn’t he/she have his/her own awards season column?

EW‘s “Safe Dave” Karger, Net Movie’s Kevin Polowy, TheWrap‘s Steve Pond, Gold Derby‘s Paul Sheehan and USA Today‘s Susan Wloszczyna got 75% correct last night.

The 71% Solution team: Vulture‘s Kyle Buchanan, Coming Soon‘s Edward Douglas, L.A. Times‘ Elena Howe, Gold Derby honcho Tom O’Neil, Rolling Stone‘s Peter Travers and Fandango’s Chuck Walton.

The 67% Crowd included were Hollywood Reporter‘s Scott Feinberg, WENN’s Kevin Lewin, In Contention‘s Guy Lodge and the Village Voice‘s Michael Musto.

At 63% were Yahoo Movies’ Matt McDaniel, Hollywood News’ Sean O’Connell, Moviefone‘s Chris Rosen, Indiewire‘s Anne Thompson and myself.

For what it’s worth, Harvard Oscar-odds cruncher Ben Zauzmer, whose predictions I briefly summarized on 2.22, got 75% of these predictions correct, which is pretty good. (75% of the 20 categories he made predictions on, that is — he abstained in four categories.) Among the top eight categories he batted 100%, obviously partly due to his somewhat surprising five-day-old prediction that Meryl Streep would beat Viola Davis in the Best Actress race.

Ben Zauzmer is a Harvard freshman interested in movies and math, and the creator of Oscarforecast, which presents Oscar predictions based solely on rigorous and dispassionate mathematical analysis. Ben’s calculations include “previous Oscar results, other awards shows, current nominations, critic scores, and guild awards,” he explains. “All of these numbers — over 5,000 data points! — were plugged into a bit of matrix algebra.”

Ben Zauzmer

And his system is predicting a Meryl Streep win for Best Actress. By a nosehair (0.7%), but still…Viola Davis gets the shaft? Everyone was sensing the closeness of this race, but I thought everything shifted in Davis’s favor two or three weeks ago. I’m not sure I buy it (or if anyone will), but Davis’s supporters now have a little something to fret about.

Ben ducked out of four categories (Best Makeup, Best Doc Short Subject, Best Animated Short, Best Live Action Short), because, he says, “there wasn’t enough data or indicators to create a reliable percentage score for each movie.”

In any event, here’s his rundown.