The full population simulation is really really cool - not something I've seen before. The reciprocators have awareness of each individual agent, right? It would be neat to add in a communication feature where word about defectors is spread through reciprocators and cooperators. Would also love to see how a generous tit for tat strategy does
Cool site. I've seen it before for some NBA vis. Impressive tool for visualizing the birthday paradox. The interface is clean and intuitive, making it easy to experiment with different parameters.
Some suggestions:
- Consider adding non-uniform birthday distributions for realism.
- A comparison between theoretical and simulated probabilities could be insightful.
XGBoost with LambdaMART ranking performed the best across all models(neural nets, log reg also tested).
Adding pairwise ranking with LambdaMART helped to solve overfitting with only some 360 data points. Instead of looking at binary win/loss MVP, I looked at the ranking of all players receiving votes in each of the past 38 years.
In general, log(x) does not refer to one base (although 10 is common). In this case, they specify that log(x) is referring to base 2 in the paragraph above, but yeah it could definitely be more spelled out.
