#include <algorithm>
#include <cmath>
#include <cstdint>
#include <iostream>
#include <iomanip>
#include <limits>
#include <vector>

int main() {
    int32_t r, c, p;
    std::cin >> r >> c >> p;

    // dp[1]
    std::vector<double> dp(c + 1, 0.0);

    for (int32_t j = 2; j <= c; ++j) {
        dp[j] = p * 0.25 * (j - 1);
    }

    double scale = 0.5 / p;

    for (int32_t i = 2; i <= r; ++i) {
        std::vector<double> ndp(c + 1);

        ndp[1] = (p / 4.0) * (i - 1);
        for (int32_t j = 2; j <= c; ++j) {
            double right = ndp[j - 1];
            double down = dp[j];

            double diff = std::abs(down - right);
            ndp[j] = (right + down - diff * diff * scale) * 0.5;
        }

        dp = std::move(ndp);
    }

    std::cout << std::setprecision(20) << std::fixed << dp.back() << '\n';

    return 0;
}