Is a class better than another one?

import random
import numpy as np
import matplotlib.pyplot as plt

x = [72, 68, 75, 84, 61, 78]
y = [64, 68, 78, 81, 59, 62, 60, 48, 82, 77]

real_diff = (np.mean(x1)-np.mean(y1))

trials = 100000 #a big number

diffs = []
for _ in range(trials):
    i = x+y
    random.shuffle(i)
    a = i[:6]
    b = i[6:]
    diffs.append(np.mean(a)-np.mean(b))

i = 0
for d in diffs:
    if d > real_diff or d < -real_diff:
        i += 1
print(i/trials)
print(i)

plt.hist(diffs)
plt.show()