====================================================================================================================================================================
player DR % 1st Serves In % 1st Serve Points Won % 2nd Serve Points Won % Return Points Won % 1st Serve Return Points Won % 2nd Serve Return Points Won
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
1 Hazem Naw 1.09 57.98 68.36 45.56 42.82 36.12 53.09
2 Michael Agwi 1.31 67.67 69.92 51.84 45.15 37.88 56.2
3 Peter Buldorini 1.03 66.02 62.5 48.02 40.27 35.06 50.15
4 Taym Al Azmeh 0.99 63.39 63.97 44.63 41.59 33.93 52.94
--------------------------------------------------------------------------------------------------------------------------------------------------------------------Serve Battle Ireland Vs Syria Davis Cup 2026
1 Introduction
Tennis matches are comprised of multiple interrelated battles, including the rally battle, the battle between attack and defence, and the battle between serve and return. Among these, the serve–return battle is widely regarded as the most significant, supported by performance data indicating that a substantial number of points end within the first four shots.
While this idea may hold particular validity among the world’s top 100 players—where serve velocity and first-serve percentages are typically exceptionally high—it remains unclear whether the serve–return battle exerts a similar influence at lower tiers of professional tennis.
To examine this idea, performance data were collected from two Irish players and two Syrian players in advance of their Davis Cup meeting in February 2026. This report seeks to identify the key factors shaping the serve–return battle and, consequently, its impact on match outcomes.
1.1 Dominance Ratio
The Dominance Ratio (DR) in tennis is a number that is used to provide information on who the dominant player is in a match. It is essentially a number that determines who is winning the serve-return battle, calculated by dividing a players percentage of return points won by the percentage of points lost on their own serve. A dominance ratio of above 1.0 means that a player is winning more points than they are losing . Can this number be used to accurately predict the winner of a match? And if so, what variables influence it?
2 Data Used
Below is a summarised table of data that was used for this report. The data was gathered from tennisabstract.com, and includes the previous twenty matches for the four chosen players. The variables that I identified as being most important in the serve-return battle were:
- Dominance Ratio
- Percentage of First Serves In
- Percentage of First Serve Points Won
- Percentage of Second Serve Points Won
- Percentage of Return Points Won
- Percentage of First Serve Return Points Won
- Percentage of Second Serve Return Points Won
The table provides the mean value for each of the variables from the players’ last 20 matches.
3 Dominance Ratio
Firstly, my aim was to determine whether or not Dominance Ratio (DR) was an important factor in the successful outcome of a match for these players. From the table above, you can see that all four players had a high dominance ratio. Was this enough to guarantee more wins than losses?
As you can see from the line on the above scatterplot, as a players DR increased, their win success also improved. In previous matches, if a player reached a DR of 1.5, it would almost guarantee that they were victorious. Could this variable also be used to predict the outcome of a match? I ran a binary logistic regression model to determine if DR is significant in affecting the outcome of a match.
=============================================
Dependent variable:
---------------------------
outcome
---------------------------------------------
dr 11.497***
(2.746)
Constant -12.027***
(2.885)
---------------------------------------------
Observations 75
Log Likelihood -17.975
Akaike Inf. Crit. 39.950
=============================================
Note: *p<0.1; **p<0.05; ***p<0.01When looking at the above model, we can see that ‘dr’ has a P value of <0.01. This suggests that DR is a significant variable in predicting the outcome of a match for these 4 players.
This might come as no surprise, as DR is a number designed to determine who has won the majority of points in a match, however tennis is a game of ever-changing variables. So which of the serve and return variables have the strongest relationship with DR?
If we observe the above plots, we can see that % of 1st serve Points Won and % of Return Points Won look to have the strongest correlation with DR. I am going to discount Return Points Won in this scenario, as this number is used directly in the calculation of DR. Thus we can determine that % of 1st Serve Points Won looks to have the strongest relationship with DR.
To confirm this, I ran a correlation matrix.
====================================================================================================
dr first_in first_percent snd_percent return_won first_return_won snd_return_won
----------------------------------------------------------------------------------------------------
dr 1 0.140 0.689 0.583 0.775 0.661 0.499
first_in 0.140 1 -0.071 0.200 0.088 0.072 0.167
first_percent 0.689 -0.071 1 0.314 0.243 0.205 0.134
snd_percent 0.583 0.200 0.314 1 0.240 0.211 0.074
return_won 0.775 0.088 0.243 0.240 1 0.865 0.658
first_return_won 0.661 0.072 0.205 0.211 0.865 1 0.238
snd_return_won 0.499 0.167 0.134 0.074 0.658 0.238 1
----------------------------------------------------------------------------------------------------From this, we can see that % of 1st serve Points Won (first_percent) has a correlation of 0.69 with dr, which is a moderate to strong correlation. We can therefore define % of 1st Serve Points Won as a very important factor in influencing DR, as well as the outcome of a match.
4 First Serve Performance
So how did our players compare when it came to % of 1st Serve Points Won?
As we can see from the above charts, Michael Agwi and Hazem Naw are the two players that perform highest when it comes to first serve. Both players are close to the Top 50 in the world when it comes to % of 1st Serve Points Won, which can be observed in the barchart above. Both players also have the most symmetric distribution of data when it comes to % of 1st Serve Points Won. This can be seen on the density plot above.
5 Davis Cup Matchup
Excitingly, the two best servers statistically, Michael Agwi and Hazem Naw, were drawn to play each other in the first singles match of Ireland and Syria’s Davis Cup tie in February. Knowing the importance of DR and % of 1st Serve Points Won in determining the outcome of a match, this would truly be a battle of serve and return. Below is a table summarising the match statistics for both players, as well as analysis of their first serve performances.
=================================================================================================================================================================
player DR % 1st Serves In % 1st Serve Points Won % 2nd Serve Points Won % Return Points Won % 1st Serve Return Points Won % 2nd Serve Return Points Won
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
1 Hazem Naw 0.62 56 76 77 46.66 37.78 73.33
2 Michael Agwi 0.4 74 62 27 23.81 24 23.52
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
Using the Violin Plots above, I was interested to discover how the players performed in this match compared to their previous 20, looking specifically at % of 1st Serve Points Won. The diamond marker displayed on each violin plot represents the % of 1st serve points won by each player during the Davis Cup tie.
Hazem Naw greatly exceeded his prior serving performance, achieving a percentage that positioned him within the fourth quartile of his performance distribution. In contrast, Michael Agwi demonstrated a substantial decline relative to his prior performance, recording a percentage that placed him in the first quartile. Unsurprisingly, this had a a large affect on the outcome of the match, with Hazem Naw winning 6-1 6-3.
I was keen to explore the sub-par performance of Michael Agwi further, so I watched the match again, this time to identify possible causes of this low percentage of points won on first serve. Below is a graphic of Michael Agwi’s serve location when he made a successful first serve.
When analysing these serving statistics and serve locations, I observed that Michael Agwi was much more successful when he was serving on the deuce side of the court. This can be seen in the barchart below.
As we can see from the barchart, Michael Agwi did not have much success when serving on the ad side of the court, and this largely contributed to the low serving performance observed. This was especially evident when serving to the ‘T’, with Michael Agwi winning less than 50% of his points when he served there on the Ad side. See Lollipop Chart below.
6 Conclusion
It is clear that the battle of serve and return is important in tennis at all levels of the game. A players ability to hold service games is vitally important if they wish to be the dominant player. In order for a player to hold as many service games as possible, it is important that they have an effective first serve, one that is equally effective in all locations of the box. Using statistical analysis such as this can be important in identifying specific areas of improvement for individual players.