Can We Predict Nezha’s Box Office? A Data‑Driven Forecast Using Polynomial Fitting

Recent days have seen the Chinese animated film Nezha dominate social media, with its opening day surpassing 100 million yuan and reaching over 1.5 billion yuan within eight days. The author, without having seen the film, decides to apply a scientific approach to forecast its total box‑office.

Historical box‑office data for several comparable movies were collected from Maoyan, and their daily revenue curves were plotted. Some films show a post‑release surge followed by weekend spikes, while others peak on release day and then decline.

To obtain a more systematic prediction, the author fits a polynomial curve to the historical data using NumPy 's polyfit and poly1d functions, which compute the least‑squares polynomial coefficients. np.poly1d(np.polyfit(x, y, n)) The fitted curves for nine reference movies are shown (red line = fitted result).

Because the new film has too few data points for a direct fit, the author scales the curves of other movies to match Nezha’s early earnings. Using SciPy 's leastsq function, the curve of a reference film is treated as a base model and fitted to Nezha’s existing data.

def func(x, p):
    A, k = p
    return A * z(x)  # z is the model derived from another movie

def residuals(p, y, x):
    return y - func(x, p)

plsq = leastsq(residuals, p, args=(y, x))

Applying the curve of Da Sheng Gui Lai (scaled to a 60‑day run) yields a forecast of 41.3 billion yuan, but the required scaling factor of 4.4 suggests a large error. Using the curve of Wandering Earth gives a more plausible estimate of 27.6 billion yuan.

Incorporating a conservative weekend guess (2 billion, 2.5 billion, 2.5 billion) and refitting raises the projection to 29.9 billion yuan.

Considering Nezha’s family‑friendly theme and a longer summer window with less competition, the author concludes that a total gross between 30 billion and 40 billion yuan is realistic, estimating roughly 35 billion yuan.

All related code, data, and result images are available for download; interested readers can request them by replying with the keyword “Nezha”.