#include <bits/stdc++.h>

double naive_solve(long long n, long long l, long long r) {
  l = std::min(6 * n, std::max(n, l));
  r = std::min(6 * n, r);
  std::vector<std::vector<double>> dp(2, std::vector<double>(r + 6 + 1));
  dp[0][0] = 1;
  for (int i = 0; i < n; ++i) {
    std::fill(dp[~i & 1].begin(), dp[~i & 1].end(), 0);
    for (int j = i; j <= 6 * i; ++j) {
      for (int d = 1; d <= 6; ++d) {
        dp[~i & 1][j + d] += dp[i & 1][j] / 6;
      }
    }
  }
  return std::accumulate(dp[n & 1].begin() + l, dp[n & 1].begin() + r + 1, 0.0);
}

double extreme_solve(long long n, long long l, long long r) {
  double avg = n * 3.5;
  double var = n * 35 / 12;
  return (std::erf((r + 0.5 - avg) / std::sqrt(2 * var)) - std::erf((l - 0.5 - avg) / std::sqrt(2 * var))) / 2;
}

signed main() {
  long long N, L, R;
  std::cin >> N >> L >> R;
  if (N < 8000) std::cout << std::fixed << std::setprecision(6) << naive_solve(N, L, R) << std::endl;
  else std::cout << std::fixed << std::setprecision(6) << extreme_solve(N, L, R) << std::endl;
  return 0;
}