Variance is a statistical term that is the measure of the amount of spread shown in a set of data. It is used in many financial models and when evaluating investment and scientific models when looking at consistency. It is a very powerful tool that can help decide whether you should believe in something – in other words, the risk of a decision. While I recently denounced the value of statistics in certain situations, variance is used to find the risks within the parameters of the statistics.
In life, variance is an interesting thing. In the financial world, you will see it in investments. Without getting too much into the intricacies, it is paired with an expected value (EV) that is then compared against several options to find the “safest” outcome. It is also used to regulate and put perspective on scientific conclusions and data.
In video games, variance is another important measure, but it is far less scientific. Everyone has a friend who plays with you that has a really high variance. Some games they will do very well, some games they do very poorly. We all also know that person who is solid as a rock, but won’t break out and do particularly great.
Of the previous examples, I am the solid player. I typically have a very average game and don’t often do very well or very badly. This means that I don’t necessarily win games by myself, but I’m not often the reason we lose them. Conversely, a variant player can get “hot” and go on a dominating streak or cold and lose games over and over.
This comes into effect when composing a team of any sort of being measured on your performance. If you’re making a team for your senior project in university and fill it with highly variant people, you can either have a remarkable product or fail the course. It also affects the individual in both life and games when measuring your performance against others. If you have a job and your boss assigns you a task, how well you do and how often you do well impacts your job.
But boss, the LAST building I designed was good!
In the mathematics of investment, say you have an option to host a fund raising party either inside or outside. If it’s inside, you’re guaranteed $50 return, no questions asked. If the party is outside and it doesn’t rain, you can make $100, but if it rains you lose $10 because nobody comes. This is a bastardized example, however your decision relies on many factors and probabilities. Do you bet on either getting $100 or losing $10, or do you just guarantee yourself $50?
While a mathematical model can “solve” the above problem with more input values, the question of whether you take the $50 (have a solid player), or risk going outside and either (having a variant player) is a valid question for your team’s performance. Do you stack your league team full of players that are inconsistent and ride whatever win streak you might get, or do you acquire solid players? Both are important to a team and it comes down to taking a risk and minimizing variability.
The third option is, of course, a math party.
Now common sense and math both dictate that some variant players are better than consistent players as well. I’m mostly including this for completeness’ sake, and this to say that a player that varies between 1800 and 2200 ELO is stronger than a consistent 1500 player. The issue of variance only arises when you have the 1800 – 2200 player vs. a 2000 ELO player. Deciding whether to take a risk in adding a hot-or-cold player onto your team is a decision you have to make in both League of Legends and life.
Now you can wrap your head around variance in both life and video games. Being less variable is typically stronger when you have to be consistent, such as in a job or tournament situation. However, the merits of being highly variable are still there and can have their power if harnessed correctly. Next week I will delve into the specifics of variance and how to make your performance more consistent. Until then…