Главная
Study mode:
on
1
Intro
2
Motivations
3
Esports, e-sports, eSports
4
Overwatch according to its website
5
Anatomy of an Overwatch game
6
The Overwatch League (2)
7
Following the action
8
Process outline
9
The Architectural Diagram Slide
10
Picking packages
11
Video splicing
12
Area cropping
13
Player name parsing using Azure
14
Data storage using TinyDB
15
Data plotting & stats
16
Chart example
17
Retrospective
18
Dataset
19
Pre-game standings (wins-losses)
20
Pre-game win percentage
21
Bias & hypotheses refresher
22
Results - roles
23
Results - teams
24
Never enough data (2)
25
Takeaways
26
Links - Tools (2)
Description:
Explore the world of eSports broadcasting through a data-driven analysis of professional Overwatch games in this PyCon US talk. Delve into the challenges faced by in-game video producers as they balance accessibility for casual viewers with engaging content for regular players. Learn how to extract data from video footage, process player information, and store results using various Python packages and cloud services. Examine the implicit biases and narratives created by broadcasters' choices in character class representation and team focus. Discover the process of validating hypotheses about viewer preferences in eSports content through data visualization and statistical analysis. Gain insights into the intersection of competitive gaming, data science, and media production while considering the limitations of available datasets and the potential for further exploration in this rapidly evolving field.

Decoding Bias and Narrative in Competitive Video Games

PyCon US
Add to list
0:00 / 0:00