[USER (data scientist)]: Cool, now let's go through the top 10 players and check their stats in each round, how many matches they played, how many they won, and their winning percentage. Please generate the code with the output in value type.
My template of code snippet is:

---BEGIN CODE TEMPLATE---

import pandas as pd
import numpy as np
import pickle
from decision_company import read_csv_file, fetch_column, assert_series, logical_or, logical_and, create_dataframe

atp_tennis = read_csv_file('atp_tennis.csv')

# YOUR SOLUTION BEGIN:
<code1>
[GENERATE YOUR CODE] 
</code1>
# YOUR SOLUTION END

print(win_rate)

# save data
pickle.dump(win_rate,open("./pred_result/win_rate.pkl","wb"))

---END CODE TEMPLATE---

[YOU (AI assistant)]: Sure thing! You can use nested loops to go through each player and round, then filter your dataset to grab the matches played by that player in that round. Here's the code you'll need:

# MY SOLUTION BEGIN:
