What is Elo in Gaming?

Elo is a numerical rating system originally created for chess by physicist Arpad Elo in the 1960s to calculate the relative skill levels of players. Though initially designed for chess, Elo and similar rating systems are now used in many competitive online games as a way to match players of similar skill levels and provide rankings. But what exactly do Elo ratings mean, and how do they work? As a fellow gaming enthusiast and data analyst, let me provide an in-depth look at Elo and how it applies to games like Overwatch, League of Legends, and more.

What Does Elo Measure in Games?

An Elo rating is a numerical representation of a player or team‘s skill level. The higher the Elo rating, the more skilled a player is assumed to be. As you win games against higher rated opponents, your Elo rating will increase. Similarly, losing games against lower rated opponents will cause your Elo to drop. The exact amount your rating changes depends on the relative difference between your rating and your opponent‘s. Defeating a much higher rated team will lead to a bigger boost than beating an evenly matched team.

Elo ratings are not a direct measurement of personal skill, but rather your ability to help your team win games against other teams of varying skill levels. In team games like Overwatch, the Elo rating reflects the overall skill of the team rather than individual players. However, your individual performance does factor into how much Elo rating you gain or lose after each match. Both personal skill and match results contribute to your ongoing Elo rating.

Elo Rating Distributions and Skill Levels

The average Elo rating varies between different games, but is usually centered somewhere between 1200-1500. Players with an Elo below 1200 are generally considered beginners, while 1500-1800 represents decent or above average skill. Highly skilled players typically have Elo ratings over 2000, with professional players and top leaderboard competitors reaching ratings of 2500+.

Let‘s look at the Elo distributions in a few popular competitive games:

As you can see, the averages fluctuate between games based on factors like randomness, team sizes, player base, etc. However, the general pattern of where skill levels lie remains similar.

Here are some rough Elo guidelines for different skill tiers across competitive games:

• 800-1100: Beginner
• 1200-1400: Novices with some experience
• 1500-1700: Average skilled casual players
• 1800-2000: Advanced players
• 2000-2300: Highly skilled and committed players approaching professional levels
• 2300+: Top professionals and elite players

However, the exact distribution and skill ranges depend on the specific game. The averages tend to fluctuate as new players join and others improve over time.

How Elo Gains and Losses Work

The amount of Elo rating exchanged after a match depends on the difference between the two teams‘ ratings. If you defeat a team rated much higher than yours, you‘ll gain more Elo points than from beating a similarly rated team. The exchange aims to bring the two teams‘ ratings closer together to accurately reflect their relative skill levels against one another.

More specifically, the Elo system uses a K-factor and expects results to calculate points won or lost after a game. The K-factor represents how much rating is updated each game – typically between 15-25 in most games. If you perform as expected against your opponent‘s rating, you gain or lose a few points. But exceeding or failing expectations substantially will adjust your rating more drastically.

In team games, your personal performance also impacts your Elo change after a match, above and beyond the team outcome. Excellent play against higher rated opponents will boost your Elo gains. Likewise, poor play may mitigate Elo losses if your lower rated team defeats you. This performance component allows your individual skill to eventually outpace teammates of lower skill within the same team Elo.

The Origins and History of Elo Systems

The Elo system was created by Arpad Elo, a physics professor and avid chess player. Elo devised the rating system to calculate relative skill in chess and track players‘ improvements over time. He derived the math underlying Elo from statistical models used in chess and other zero-sum games. The World Chess Federation adopted Elo‘s rating system in 1970, helping popularize it.

Fun fact: Elo himself was rated an Expert at 2200 Elo points at his peak!

Since then, Elo and similar rating systems have been implemented into many head-to-head competitive games with ranked skill-based matchmaking, including League of Legends, Overwatch, Counter-Strike and more. For example, Riot Games adopted a modified Elo system called the League of Legends Ranked Rating for ranked play. Other games use proprietary systems inspired by Elo math as well.

The concepts extend beyond chess into any competitive environment where relative skill levels impact expected outcomes. Today "Elo" often generically refers to the rating system itself, rather than just Elo‘s specific calculations.

Examples of High Elo Players

Let‘s look at a few players with insanely high Elo ratings and how they climbed the competitive ranks:

• Magnus Carlsen – The current World Chess Champion boasts a peak Elo rating of 2882, the highest in history! He seized the top ranking in 2011 at just 19 years old through relentless training and study.

• Uzi – This Chinese League of Legends legend achieved over 1500 LP Grandmaster rank on the Chinese superserver, estimated to be over 2500 Elo. His mechanics are unmatched, leading China to numerous Worlds finals.

• Gael "CrozzOver" Leveling – A professional Overwatch DPS player for Paris Eternal, he has reached #1 on the ranked ladder with over 5000 SR, equivalent to 3000+ Elo. His tracking aim and flexibility enabled the climb.

These players demonstrate what‘s possible to achieve through determination and high-level play. Their skills allow them to consistently win against world class competition.

Limitations and Controversies

While Elo and similar systems provide reasonably accurate models of relative skill, there are some limitations and controversies surrounding them:

• Elo Hell – Some players feel "stuck" at a certain Elo range and claim it‘s impossible to advance due to factors out of their control, like unskilled teammates ("Elo hell"). As a player myself, I understand this frustration. But in the long run, self-improvement and consistency can overcome bad teammates.

• Intentional Losing – In some games, players intentionally lose matches to tank their rating and then dominate lower skilled opponents on the way back up. For example, in League of Legends, this tactic is called "smurfing." Preventative measures like requiring phone verification help, but don‘t eliminate the issue.

• Day-to-Day Variance – Players‘ performance fluctuates daily based on many factors. Elo may not capture short-term spikes and dips in skill. As an analyst, I‘d love to see additional statistics tracked like consistency, fatigue, and mentality factors.

• Limited Variables – Elo systems generally only account for win/loss results. But as a data scientist, I know performance depends on many other factors not involved in the calculations, like mechanics, adaptability, and communication. More data could enhance accuracy.

• Team vs Individual – In team games, Elo measures team skill, yet queueing alone means you can face opponents of very different individual skill. Weighting individual performance helps compensate but doesn‘t fully address this mismatch.

So while Elo and similar rating systems have their limitations, they still provide reasonably fair matchmaking and skill measurement in most cases. The ratings strongly correlate with expected performance and results at each level of competitive play. For these reasons, Elo and similar ranking systems remain prevalent in most skill-based competitive video games today, even 60+ years after its invention!

Let me know if you have any other questions! I‘m always happy to dig deeper into gaming data and statistics.