PICK THE RIGHT TEAM AND MAKE A BLOCKBUSTER A SOCIAL ANALYSIS THROUGH MOVIE HISTORY

PICK THE RIGHT TEAM AND MAKE A BLOCKBUSTER A SOCIAL ANALYSIS THROUGH MOVIE HISTORY

THE CHALLENGE: TO UNDERSTAND HOW TEAMS CAN WORK BETTER SOCIAL NETWORK + MACHINE LEARNING TO THE RESCUE

Previous research: team success Teamwork selection as an optimisation problem Anagnostopoulos et al. [2012], Tseng et al. [2004], Studied team success without social parameters Kim et al. [2013], Elberse [2007], Did not study the team as a whole Nemoto et al. [2011], Singh et al. [2011],

Previous research: social features Studied social parameter of individuals Papagelis et al. [2011], Li et al. [2013], Studied single social features Chen and Guan [2010], Schilling and Phelps [2007], Performed on small datasets Ghiassi et al. [2015],Oghina et al. [2012], No predictive analisys Uzzi and Spiro [2005], [Burt, 2009],

RESEARCH QUESTION: IN PREDICTIVE ANALYSIS OF TEAM SUCCESS, DOES USING MANY TOPOLOGICAL FEATURES FROM TEAMS HELP?

Methodology Start with large set of collaboration data (IMDB) Form a social network Filter irrelevant data Extract social features from team Characterize this never-before-seen data Apply Machine Learning Techniques Assess how social features help predict team success

DATASET IMDB [INTERNET MOVIE DATABASE] WORLD S LARGEST MOVIE DATASET DATE 1808 2014 SIZE 12,250 MOVIES 31,698 PRODUCERS

Associate Producer Co-Producer Executive Producer Line Producer Producer MOVIE S TYPICAL PRODUCING TEAM PRODUCERS THAT WORK TOGETHER ARE LINKED IN A SOCIAL NETWORK

Forming a Social Network Movies Producers Producer s Social Network

Removing inactive nodes

Filtering: 238K 32K Movies Filtering out movies that are Not connected to giant component Not from cinema Just one producer Released before 1930 (used for bootstrapping) Not feature length (< 30 min.) Not relevant (< 1,000 votes)

MOVIE S SUCCESS PARAMETERS NUMBER OF RATINGS (POPULARITY), AVERAGE RATING (ACCEPTANCE), GROSS (FINANCIAL SUCCESS)

Characterization: Movie Success Distribution of movie success Historical evolution of success distribution Correlation between different success metrics

(a) (b) (c) Movies Movies Movies 600 450 300 150 0 10 3 10 2 10 1 10 0 10 1 10 2 10 3 Gross (Million USD) 700 460 230 0 10 3 10 4 10 5 10 6 Votes 10 4 10 3 10 2 10 1 10 0 G 1 G 2 G 3 1 2 3 4 5 6 7 8 9 10 Rating HISTOGRAM OF MOVIE S SUCCESS PARAMETERS G1: TOP 10% MOVIES, G2: TOP 10 50% MOVIES, G3: ALL OTHER MOVIES

Movies 10 3 10 2 10 1 EXPLOSION IN MOVIE PRODUCTION 10 0 10 3 Gross 10 1 10 1 10 3 MORE MOVIES WITH LOWER GROSS NOW 10 5 10 5 c) Votes 10 4 10 3 OLD MOVIES RECEIVE LESS VOTES ) Rating 8.6 7.8 7.1 6.4 5.7 5.0 1930 1940 1950 1960 1970 Decade 1980 1990 EVOLUTION OF MOVIE S SUCCESS PARAMETERS DISTRIBUTION 2000 2010 ON TOP, HISTOGRAM OF MOVIE PRODUCTIONS COLOR CODED BY SUCCESS GROUP BIASED HIGHER RATINGS FOR OLD MOVIES

(a) (b) (c) Gross (Million USD) Rating (normalized) Gross (Million USD) 10 3 10 2 10 1 10 0 10 1 10 2 10 3 10 3 10 4 Votes 10 5 10 6 9 8 7 6 5 4 10 3 10 4 Votes 10 5 10 6 10 3 10 2 10 1 10 0 10 1 10 2 10 3 4 5 6 7 8 9 Rating (Normalized) HEXAGONAL SCATTER PLOT BETWEEN SUCCESS PARAMETERS DARKER BLUE SHADES REPRESENT HIGHER CONCENTRATION OF MOVIES 40 35 30 25 20 15 10 5 0 2.4 2.1 1.8 1.5 1.2 0.9 0.6 0.3 0.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 movies log(movies) log(movies) POPULAR MOVIES HAVE HIGHER GROSS POPULAR MOVIES HAVE HIGHER RATINGS MOVIES WITH HIGHER GROSS DEVIATE MORE FROM THE AVERAGE RATING

RESEARCH QUESTION: (IN THE CONTEXT OF MOVIE PRODUCING TEAMS) GIVEN DIFFERENT TEAMS THAT COULD PRODUCE A MOVIE, WHICH IS MORE LIKELY TO ACHIEVE SUCCESS?

Movie Characteristics GENRES (21) RUNTIME PRODUCTION BUDGET (NORM.) CONTINENTS (6)

Movie team Parameters: Ego # OF PAST EXPERIENCES LEVEL OF PREVIOUS SUCCESS IN-DEGREE CLOSENESS CLUSTERING COEFFICIENT BETWEENNESS NETWORK CONSTRAINT SQUARE CLUSTERING COEFFICIENT

Movie team Parameters: Pairwise SHARED FRIENDS NEIGHBOUR OVERLAP SHARED EXPERIENCE

Movie team Parameters: Global GLOBAL CLUSTERING COEFFICIENT AVERAGE SHORTEST PATH SMALL-WORLD-COEFFICIENT

Problem: many numbers from a single parameter ARITHMETIC MEAN HARMONIC MEAN MEDIAN STANDARD DEVIATION MINIMUM VALUE MAXIMUM VALUE NODE CONTRACTION

NUMBER OF FEATURES FOR EACH MOVIE 70 EGO FEATURES 10 PARAMS. X 7 AGG. WAYS 27 MOVIE FEATURES 21 GENRES + 6 CONTINENTS 3 MOVIE PARAMS. RUNTIME, TEAM SIZE, BUDGET 3 GLOBAL METRICS Q, CLUSTERING, AVG. PATH LENGTH 121 TOTAL DISTINCT FEATURES

Characterization: Movie Teams Parameters Distribution of parameters Historical evolution of parameters Relation between success metrics and parameters Distribution of movies in pairs of characteristics

6.0 10.5 150 120 9.0 7.5 90 6.0 60 0.24 7.5 0.16 7.0 6.5 0.08 6.0 0.00 5.5 Runtime Mean: Previous (minutes) votes Team Mean: metrics: Gross (Billion Previous USD) ratings 40 S 70 S: MOVIES WERE LONGER EFFECT OF CHRONOLOGICALLY BIASED RATINGS 120 10.5 90 9.0 60 7.5 30 6.0 Team metrics: Previous votes experience Mean: Previous votes 1930 1940 1950 1960 1970 Decade 1980 EVOLUTION OF MOVIE S FEATURES DISTRIBUTION PREV. VOTES, PREV. RATINGS, MOVIE RUNTIME 1990 2000 2010 RECENT MOVIES: +TEAMS THAT PREVIOUSLY PRODUCED POPULAR MOVIES

Mean: Gross (Billion USD) 0.24 0.16 0.08 0.00 Team metrics: Previous experience 120 90 60 30 0 1930 1940 1950 1960 1970 1980 1990 2000 2010 Decade EVOLUTION OF MOVIE S FEATURES DISTRIBUTION MEAN OF PREVIOUS GROSS, TEAM S PREVIOUS EXPERIENCE

450 300 150 0 16 12 8 4 0.4 0.3 0.2 0.1 Team metrics: Degree Team metrics: Team size Median: Closeness Harmonic mean: Clustering 2000 S: EXPLOSION OF TEAM S DEGREE 2000 S: MUCH HIGHER # OF PRODUCERS PER TEAM CLOSENESS: AN EVOLVING CHARACTERISTIC 0.8 0.6 0.4 0.2 0.0 1930 1940 1950 1960 1970 1980 1990 2000 2010 CLUSTERING: FAIRLY STABLE DISTRIBUTION Decade DISTRIBUTION MOVIE S FEATURES

Closeness Runtime Closeness Net. constraint Budget Team Size Degree Prev. votes G 1 G 2 G 3 Team Size Prev. gross INTERACTION BETWEEN PAIRS OF FEATURES DARK CLUSTERS SHOW CONCENTRATION OF BLOCKBUSTERS

A SUCCESSFUL, FEATURE LENGTH MOVIE CAN T BE TOO SHORT 50 100 150 200 250 (a) Runtime (minutes) 4 5 6 7 8 (b) Team metrics: Previous ratings 7 8 9 10 11 12 10 8 (c) Team metrics: Previous votes TEAMS WITH MODERATE PREVIOUS RATINGS PERFORM BETTER (! ) TEAMS THAT HAVE PRODUCED MORE POPULAR MOVIES BEFORE PERFORM BETTER HISTOGRAM OF MOVIE S PARAMS, PER SUCCESS GROUP

7 8 9 10 11 12 10 8 (c) Team metrics: Previous votes 0.0 0.5 1.0 1.5 2.0 2.5 3.0 (d) Mean: Gross (Billion USD) 10 8 G 1 G 2 G 3 50 100 150 200 250 (e) Team metrics: Previous experience TEAMS THAT HAVE PRODUCED MORE MONEY BEFORE PERFORM BETTER TEAMS WITH SUMMED LOW EXPERIENCE PERFORM BADLY 0 200 400 600 800 1000 1200 1400 (a) Team metrics: Degree TEAMS WITH LOW DEGREE PERFORM BADLY 0.2 0.4 0.6 0.8 1.0 (b) Team metrics: Network Constraint SOCIALLY UNCONSTRAINED TEAMS PERFORM BETTER HISTOGRAM OF MOVIE S PARAMS, PER SUCCESS GROUP

BEST PERFORMING TEAMS ARE NEITHER SMALL NOR BIG 5 10 15 20 25 30 35 40 (c) Team metrics: Team size 0.10 0.15 0.20 0.25 0.30 (d) Median: Closeness G 1 G 2 G 3 0.2 0.4 0.6 0.8 1.0 (e) Harmonic mean: Clustering TEAMS WITH LOWER CLOSENESS PERFORM WORSE TEAMS WITH LOWER CLUSTERING (BY HARMONIC MEAN) PERFORM BETTER HISTOGRAM OF MOVIE S PARAMS, PER SUCCESS GROUP

Movie Success Forecast Movie Producing teams characteristics as features Movie success parameters as target variables Regressor: Bayesian Ridge (better to handle noise) Feature selection: eliminate features with less significance until model starts loosing accuracy

Feature selection Out of 121 features, 23 features were selected 19 Non-topological: Genres (9), Continent (3), Runtime(1), Budget(1), Previous success (4), Previous Experience (1) 4 Topological: Degree (1), Team Size (1), Closeness (1), Clustering (1)

Test R 2 : 0.694 Baseline R 2 : 0.399 Votes True value Baseline Test IMPROVEMENTS IN PREDICTION ACCURACY WITH SOCIAL FEATURES RED BARS REPRESENT ACCURACY GAINS IN THIS SAMPLE, RED BARS, LOSSES

Target Years Non Topol. Topologic All Gain Votes Gross Rating 2008 2013.529, ±.0008.310, ±.0006.556, ±.0008 5.10% 2000 2013.484, ±.0004.294, ±.0005.517, ±.0004 6.82% 1990 2013.437, ±.0003.246, ±.0004.464, ±.0003 6.18% 2008 2013.431, ±.0008.170, ±.0013.448, ±.0009 3.94% 2000 2013.419, ±.0004.175, ±.0005.447, ±.0004 6.68% 1990 2013.392, ±.0004.174, ±.0004.435, ±.0003 10.97% 2008 2013.271, ±.0011.033, ±.0009.281, ±.0012 3.69% 2000 2013.267, ±.0006.038, ±.0003.273, ±.0006 3.37% 1990 2013.258, ±.0004.031, ±.0003.262, ±.0005 1.55% OVERALL GAIN IN PREDICTIVE ACCURACY (R2), 95% C.I.

Contributions Improvement to the state-of-the-art in movie success forecasting In-depth characterization of social aspects of a large collaborative network Presented a new approach for extensive aggregation of social metrics from agents in teams

THIS IS ONLY A FIRST LOOK IN HOW NETWORK TOPOLOGY ANALYSIS CAN HELP EXPLAIN COMPLEX HUMAN BEHAVIOR.