Abstract:This article considers a one-parameter family of circles F_C, which has the interesting property that the null isocline of the family is the largest member of the family. This family of circles is bounded and we consider the problem of deriving an equation for the envelope of F_C. We provide one standard solution, and three nonstandard ones. The article concludes with a discussion of the relative merits and aesthetics of the different methods.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is related to the "illusion of control" phenomenon in the Parrondo Effect (PE). Specifically, the paper explores the phenomenon in Brownian motion games where, when a seemingly reasonable optimization rule is adopted, the net gain originally obtained by alternating between two fair or losing games is instead reduced or even becomes negative.
### Main problem description:
1. **Parrondo Effect (PE)**:
- The Parrondo Effect refers to the fact that by alternating between two games that are individually fair or losing, a winning result can ultimately be obtained. This effect was initially observed in Single - Player Parrondo Games (SPPG) and Multi - Player Parrondo Games (MPPG).
2. **Illusion of control**:
- The paper points out that when certain optimization rules are applied to these games, the original Parrondo Effect will be weakened or even reversed. This optimization rule attempts to maximize the gain by choosing the game that performed best in the past, but in fact leads to worse results.
- The author calls this phenomenon the "illusion of control", that is, the performance of a low - entropy strategy (a strategy using more information) is worse than that of a random strategy (a strategy with maximum entropy).
### Specific research content:
- **Single - Player Parrondo Games (SPPG)**:
- The research found that in single - player history - dependent Parrondo games, even with the most natural optimization rule, the "current reversal" phenomenon will occur (that is, the positive Parrondo Effect is reversed to negative). This provides a natural example of the illusion of control in Parrondo games.
- **Time - Horizon Parrondo Games (THPG)**:
- In multi - player capital - dependent Parrondo games, the optimization rule will cause the system to choose one game more often, and this single game is itself losing. Therefore, the overall result becomes a greater loss.
### Mathematical models and formulas:
- **Steady - state probability and conditional probability**:
- For each game \( i \), its steady - state probability vector is \( \mathbf{r}^{(i)} \), and the conditional winning probability vector is \( \mathbf{p}^{(i)} \).
- The steady - state winning probability of the combined game is:
\[
P_{\text{win}}=\sum_{i = 1}^N\lambda_i\mathbf{p}^{(i)}\cdot\mathbf{r}^{(i)}
\]
- When the Parrondo Effect occurs, the winning probability of the combined game is greater than the sum of the probabilities of each individual game:
\[
P_{\text{win}}(1, 2)>P_{\text{win}}^{(1)}+P_{\text{win}}^{(2)}
\]
- **Impact of optimization rules**:
- Under the optimization rule, choosing the game that performed best in the past, its steady - state winning probability is close to the probability of the individual game:
\[
P_{\text{best win}} = 0.496
\]
- While choosing the game that performed worst in the past can obtain a higher gain:
\[
P_{\text{worst win}}=0.507
\]
### Conclusion:
Through theoretical analysis and numerical simulation, the paper reveals the negative effects that may be caused by applying optimization rules in Parrondo games, and emphasizes the widespread existence of the "illusion of control" and its misleading impact on the management and prediction of complex systems. This phenomenon not only exists in Parrondo games, but also widely exists in many human attempts to manage and predict complex systems.