Scoring probability model based on service landing location and ranking points in men’s professional tennis matches

How to use

The first slider specifies the ranking ratio between the server and the reciever.
Scoring probability in service areas (Deuce/Ad sides) are displayed.
- This probability does NOT mean whether each service is ace or not.
Four probabilites are depicted. Two sides (Deuce/Ad) and 1st/2nd services. Serviced ball fly from the bottom to upwards.

Ranking point ratio (Server/Reciever)

Scoring probability of 1st service

Scoring probability difference between server and reciever. 1st service

Scoring probability of 2nd service

Scoring probability difference between server and reciever. 2nd service

Description

Every play in tennis starts from service. In professional tennis, the scoring probability of servers is higher than 0.5 (0.66 in the 2019 ATP tour).
Several tournaments in the ATP tour disclose information on the landing location of services and their final score, e.g.,This match). The authors collected from 30 tournaments, and 897 matches in ATP Tour 2019. The total number of services is approximately 115 thousand. The authors constructed a mathematical model that expresses the scoring probability of services. In this model. the authors assume that the scoring probability can be explained by their landing location and (logarithm of) ranking point ratio between server and receiver.

Fig: Histogram of service landing location for each service side.

Further work

Model for difference surfaces
Include service speed as explanetory variable. The authors can not find the data with both landing location and speed.

License

This document is written by Eiji KONAKA a.k.a. konakalab. Japanese version:2022/2/28. English translation:2023/1/20. Coding for data collection and model construction has been done by Fumiya SHIMIZU.
Cite as: F. Shimizu and E. Konaka, "Scoring probability model based on service landing location and ranking points in men’s professional tennis matches," https://www-ie.meijo-u.ac.jp/~konaka/tennisServiceProb_Eng.html

"Scoring probability model based on service landing location and ranking points in men’s professional tennis matches" (Shimizu and Konaka) is licensed under CC BY 4.0.