Figure 2

Fluctuations of impact, luck and Q. (a) According to the classical test theory, the normal distribution of an observed variable (green in the example) can be decomposed as the sum of the distributions of the true score (blue) and the error term (red). (b) Distribution of p̂ and of Q̂ for two different, fictional fields. In Field A the distribution of p̂ has a low variance compared to Q̂, therefore randomness has a negligible role (\(R \to 0\)). Field B exhibits the opposite behavior, with a narrow Q̂ and broad p̂ distribution meaning that the individual’s luck dominates impact (\(R \to 1\)). (c) We show the studied 28 creative fields on the \((\sigma ^{2}_{\hat{Q}}, \sigma ^{2}_{\hat{p}} )\) plane, marking fields from different data sets with different colors. We denoted a fitted line by continuous black line and added the diagonal as a continuous grey line as a reference. The gradient-coloring of the background changes in a diagonal direction, illustrating that the points being on the same off-diagonal lines have the same \(\sigma ^{2}_{\hat{S}}\). (d) The table shows the values of the R randomness index for the different fields