Unless you’ve been living under a rock, you’ve heard about speedrunning. This is the practice of going through many runs in one fell swoop, validating many runs with a single run. You’ve probably also heard of Mario 64, a game that’s full of puzzles, but isn’t quite a typical Mario game.
No learning curve
One problem with using a learning curve in super comparison analysis is that the e values used in the comparisons are often not meaningful without the g. For example, a g = -0.5 is not meaningful if e = -0.5, but a fixed g = -0.5 is. It also prevents direct comparison, and makes it more difficult to draw quantitative conclusions.
To illustrate the idea of learning curves, consider the following example: we want to compare the efficiency of two products. We first calculate the average time that it takes to perform the two products side by side. Then, we use that data to build a learning curve. For example, we’ll use the data for Places365.
A learning curve is easily visualized graphically. There are many points of data on the graph, with one representing the cumulative amount of time required to perform a given task or unit. Using a learning curve, we can compare the time taken to generate a certain number of tasks and units.
The learning curve model is useful for monitoring the performance of an organization and identifying areas for improvement. However, it has its disadvantages as well. First, there are many variables that influence learning progress. This means that tracking just one can result in misleading data. Second, some progress is hard to measure and quantify.