Overwatch League 2020 Analysis

This notebook provides a preliminary analysis of the data produced from the 2020 Overwatch League season.

The purpose of this analysis is to show that basic insights can be derived from such data, and that further data analysis is possible.

Information extracted:

• Relative player performance on a number of in-game metrics including:
  • Highest all damage done
  • Highest all damage taken
  • Highest death count
  • Highest final blows (the elimination of an enemy)
  • Highest objective kills (eliminations of the enemy on or very near the objective - the payload, capture point)
  • Highest healing done

All of these metrics are based on the "per 10 minutes" statistic recorded in-game. So, for example: highest healing done per 10 minutes.

• Top played Heroes in each role (time played, not pick rate):

  • Most played DPS Heroes
  • Most played Support Heroes
  • Most Played Tank Heroes

• Adjusted stats by Team (averaging statistics from all players on each team)

  • Highest damage done
  • Highest damage taken
  • Highest final blows (the elimination of an enemy)
  • Highest objective kills (eliminations of the enemy on or very near the objective - the payload, capture point)
  • Highest healing done

All of these metrics are based on the "per 10 minutes" statistic recorded in-game. So, for example: highest healing done per 10 minutes.

• Correlation data plots (to see if two stats are related to each other):
  • Deaths vs. Final Blows
  • Damage vs. Healing Done

All of these metrics are based on the "per 10 minutes" statistic recorded in-game. So, for example: highest healing done per 10 minutes.

• Pie charts of time played for each character class (based on time played, and not overall match pick rate):
  • Most played Damage Heroes by Percentage
  • Most played Support Heroes by Percentage
  • Most played Tank Heroes by Percentage

Further analysis can show:

• The correlation of a number of different statistics, for example:

  • Healing recieved vs. Damage taken
  • Map played vs. Match length
  • Damage done and death count vs. Healing done

• The relationship between high/top-performing players (as well as their role and heroes they play) and the current meta - are these players high-performing only becuase of the current meta, or are they truly great players?

• The effectiveness of certain team compositions (in the current meta in OWL)

• The effectiveness of certain heroes in the current meta (based on healing done, damage done, eliminations, etc.)

• The performance of teams on certain maps/map types (win/loss vs. map played)

• The performance of teams/players based on match length - are certain teams losing in matches that are longer than most

• Hero-based statistical analysis, for example:

  • Doomfist with a high final blows/death count ratio
  • Winston with a high elimination/damage taken ratio

And so much more! Thanks for checking out my analysis of professional Overwatch play!