[Math/Philosophy]: Does ELO hell exist?

• RockHopper
In summary, ELO hell refers to a belief that a player can get stuck in a low ranking due to consistently being matched with bad teammates. However, this is not necessarily the case as the ELO system is designed to increase a player's ranking if they consistently perform better than the average player. Over time, this should result in an increase in ELO. The belief in ELO hell is often fueled by the "swingy" nature of the game and the Dunning-Kruger effect, where players overestimate their abilities. ELO hell is a broad term and may not actually exist.
RockHopper
First let me explain what ELO hell is.

ELO is a popular rating system. It's mostly used for chess and competitive events.
In this context ELO is being used in League of Legends (A popular MOBA).

In League of Legends their are 2 teams of 5 people each.

When people first begin to play Ranked games they start at 1200 elo.
Since every player starts at 1200 elo their is a great gap in skill.

Let's say that when you begin at 1200 elo that you lose a few matched because of team mates being bad. So your elo begins to deteriorate and you continue to get matched with (and against) worse and worse people.

The Hypothesis is that if your elo goes down low enough that it would take an unbelievable amount to get out of elo hell.

This has been a topic of great controversy and I was wondering if you guys had any thoughts about it.

Lol, I have to take this argument from the LoL general discussion to this general discussion.

I'd say no, it does not exist. If you are better than the average player out of the 10 players in the game, you will win more often than chance. Period. Therefore, your Elo will increase. Even if you're way down in the 200 elo range, and you're paired with 4 terrible players, the enemy team will likely have 5 terrible players on the team.

The problem is, just because your team is better as a whole doesn't mean you're guaranteed to win. You still might only win 55%-60% of the time. However, over time, this will bring your elo up. Let's assume that you win 57.5% of the time, which is not unreasonable if you're far stronger than your elo suggests. Let's also assume that the average elo gain or loss per game is 12 (which is about right, it ranges from 11-13 usually). Then, the expected value for elo gained per game is .575(12)+.425(-12)=6.9-5.1= 1.8 elo. So, if you play 300 games and maintain that constant win rate, your elo should increase by 540 points. (A constant win rate is reasonable, because although the opponents get tougher as you move up in rank, you're presumably getting stronger through more playing experience as well).

Anecdotally, this has held true for me. Over about 500 ranked games, I've gone from 780 elo to over 1300 elo.

There are two reasons people believe in "elo hell." One is that the game can be very "swingy." It's not rare to go on long losing streaks or long winning streaks. During the long winning streaks, people start to believe they are much better than they are. During the long losing streaks, people assume that their teammates are much worse than they are, and thus conclude that the game is unwinnable. This is particularly true when they're near their true elo. The other reason is is called the Dunning-Kruger effect, where people generally overestimate their abilities (except for the very best, which tend to underestimate their abilities.) These reasons are why people believe that EVERY Elo level is "elo hell." I even hear Silver level players (meaning they're in the top 10% of all players) claim that they're in elo hell. If there is something called elo hell, it is so broad as to include 90+% of all players that it is a meaningless term.

For what it's worth, you can see my elo here to verify my personal anecdote: Link

1. What is ELO hell in relation to math/philosophy?

ELO hell is a concept in the gaming community that refers to a supposed state in which a player is trapped in a low ELO (a measurement of skill level in competitive games) despite their actual skill being higher. In relation to math/philosophy, ELO hell is often used to question the existence of mathematical or philosophical systems that seem to prevent individuals from reaching their true potential.

2. Is there any evidence to support the existence of ELO hell?

There is no scientific or mathematical evidence to support the existence of ELO hell. In fact, many studies have shown that a player's ELO is determined by their own skill and performance, rather than external factors such as teammates or game mechanics.

3. Can ELO hell be explained by the "confirmation bias" phenomenon?

Yes, ELO hell can be explained by the confirmation bias phenomenon. This is when individuals tend to only remember and focus on events that confirm their beliefs, while disregarding or downplaying contradictory evidence. In the case of ELO hell, players may only remember and blame their losses on their perceived low ELO, while ignoring their wins or individual mistakes that may have contributed to their losses.

4. How does the concept of ELO hell relate to the "Dunning-Kruger effect"?

The Dunning-Kruger effect is a cognitive bias in which individuals with low ability tend to overestimate their own skill level, while those with high ability tend to underestimate it. In the context of ELO hell, players may believe they are stuck in a low ELO due to factors out of their control, when in reality, their own skill level may be the determining factor.

5. Does the belief in ELO hell have any negative consequences?

The belief in ELO hell can have negative consequences on a player's mindset and motivation. It can lead to a defeatist attitude and discourage players from actively trying to improve their skills. It can also perpetuate toxicity in the gaming community, as players may blame their teammates or the game itself for their perceived "low ELO hell".

• General Discussion
Replies
28
Views
12K
• General Discussion
Replies
14
Views
3K
• General Discussion
Replies
1
Views
1K
• Biology and Medical
Replies
9
Views
14K
• General Discussion
Replies
18
Views
7K
• General Discussion
Replies
200
Views
17K
• General Discussion
Replies
5
Views
2K
• General Discussion
Replies
3
Views
3K
• Set Theory, Logic, Probability, Statistics
Replies
8
Views
1K
• General Discussion
Replies
9
Views
3K