Full text of my pseudo-thesis, "Why Support players are worse at League"
Be warned, it's a long read and is largely scientific argumentation.
Feel free to give it a read though!
⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️⬇️
"Why Support Players tend to be worse at League"
League is a VERY difficult game to learn, as it has unrivalled complexity and depth.
It is also a game which the primary way or learning is not study, but experience and practice.
In the psychological literature, "Reinforcement learning" is well documented. (Sutton RS, Barto AG, 2018)
web.stanford.edu/class/psych…
Reinforcement Learning is the primary way people learn in video games, especially games like League of Legends where most players engage in very little study and almost exclusively learn through experience.
Reinforcement learning may be referred to under many names, such as "Carrot and Stick" learning, or Pavlovian training, but always has the same very simple structure to it.
Good action > Reward.
Bad action > Punish.
Good play (Kill opponent, CS well, Secure Objective) > Rewarded > Gold/XP/Buffs ect
Bad Play (Die needlessly, Miss CS, Lose Jungle Camps) > Punished > Denied Gold/XP/Buffs ect
This is an EXTREMELY effective way to teach people about good/bad decisions and make them improve at the game.
It is a reinforcement loop that keeps people engaged in the game while simultaneously teaching them what to do.
Support BREAKS this reinforcement loop
(PARTICULARLY the negative punishment of doing "Bad Plays")
Supports are LESS gold reliant.
Supports are LESS XP reliant.
Supports scaling is FAR more guaranteed
(Proccing Relic shield is trivial and pratically impossible to deny in a meaningful way, compared to Lane CS/Jungle Camps)
This lack of punishment in Support means the negative feedback of making mistakes is significantly more absent compared to other roles, and as such, these players learn far more slowly and make more mistakes.
To address the most obvious critique of this analysis, people will reply:
"Well if you play badly on Support, even if you are not directly punished, your teammates will be, making them weaker and causing you to eventually lose the game, is that not a punishment...?"
This is TRUE, however the reinforcement loop is still violated regardless.
It is well documented in the psychological literature that Punishment/Reward timing is crucial for reinforcement learning.
(Vassiliadis P, Lete A, et al 2025)
pmc.ncbi.nlm.nih.gov/article…
(Perez, O.D., Urcelay, G.P 2025)
nature.com/articles/s41539-0…
For Reinforcement Learning to occur efficiently, two things must be true:
The punishment/reward must be TIMELY (It must occur very shortly after the decision is made)
The punishment/reward must also be OBVIOUS as to the action that caused it.
The problem with this with regards to Support in League is that the punishments are often not timely and not obvious. If a Support player roams top at 4 minutes, dies, gives over a kill a bad wave state, it is somewhat obvious this is a bad play.
However, the PUNISHMENT for this play only occurs far later ~15 minutes later when the enemy Toplaner leaves lane and takes over the map.
The punishment is also sometimes NON-OBVIOUS, as the enemy laner might leave lane 10/0 , and it can be hard to directly attribute how much their play contributed to this, versus how much it is the allied toplaner is just bad.
I strongly believe this is why there has been a very strong belief among the League community that Support players tend to be lower skilled than other roles.
It's not even necessarily their fault, it is harder to learn from your mistakes when the game does not punish you significantly for them.
Based on this theory and logic, I would recommend Support players to learn in a more abstract way, rather than brute force experience.
Engaging in research, study, VoD review (All roles can do this of course but I believe it especially important for Support players based on this).
I welcome any feedback or critique in the comments!