Home › content

Linear X-Stat Estimator

Thu, 2009-04-09 19:57 — creativename

This is a generalized x-stat estimator that assumes creativename's naive linear model, as opposed to the more complicated and computationally intensive non-linear conditional probability model proposed by King Morgoth. The basis is the assumption that strength ratios are retained.

The formula is:
\%_i = \frac{\frac{X_i}{1-X_i}}{\sum_{i}^{n}\frac{X_i}{1-X_i}}
Where X.i are the input values for n inputs, scaled to x-stat values with the strongest in the pack set at 50%.

Estimated Results

EntrantPercentage
Mario38.184%
Bowser26.204%
Sonic26.204%
Robotnik9.408%

Enter strength values separated by newline (example values have been pre-filled)
All values are assumed to be percentages. They must be between 0 and 100.
You can also optionally pass names, for example: Link vs. Cloud vs. Snake