Math Exploration

Introduction
The following assessment will explore if the quantity of hours devoted to train has a correlation
with the quantity of matches won/lost by tennis players. The idea of this exploration comes from
the fact that I played tennis for very long time when I was younger, but my results weren’t good. I
insisted on quitting because I got upset every time I lost but my parents insisted on me to continue
playing and devoting more hours per day so as I could improve and start winning more games than
I lost. I didn’t pay attention to them complaining that although I trained 8 hours per day, I would
still be a bad tennis player, and indeed I quitted. Since that time, I asked myself whether their
theory was correct, or wrong. Therefore, the aim of this exploration is to deduce if my parents
were right, or if I was right. This theme will explore if the effort is linked to the results, and
therefore find if there is a strong, moderate, weak or very weak correlation between the hours
devoted to train tennis per week and the matches won/lost by the player. The methods to obtain
information will be by making a survey to the tennis players in the country club where I live. I will
ask them how much hours do they devote for training tennis, and how many matches did the win
or lost in the last season. Then I will divide the matches won against the matches lost, and
therefore, make a ratio.

Procedure
• Conduct a survey asking how many hours per week do they devote for training tennis
and what were their last season results in terms of won/lost.
• Divide the matches won by the matches lost to make a ratio.
• Plot a table displaying the information collected.
• Graph the data collected in order to have a clearer representation.
• Analyse the data obtained.
• Calculate the line of best fit using the least squares regression formula.
• Trace a line of best fit to examine the correlation.
• Calculate the Pearson’s correlation coefficient.
• Conclude the exploration answering the question introduced.
Data Collection and Analysis
As mentioned before, a survey was conducted asking 24 different tennis players in the country
club where I live how many hours per week do they devote for training tennis and what were their
results in terms of won/lost the last season. The results were the following:

Tennis player Hours trained per week Matches won Matches lost Ratio
1 4 5 7 0,7
2 2 3 9 0,3
3 6 8 4 2
4 4 9 3 3
5 8 11 1 11
6 6 6 6 1
7 2 5 7 0,7
8 5 4 8 0,5
9 8 6 6 1
10 3 7 5 1,4
11 6 9 3 3
12 8 4 8 0,5
13 8 10 2 5
14 4 9 3 3
15 1 1 11 0,1
16 6 7 5 1,4
17 3 4 8 0,5
18 2 2 10 0,2
19 6 8 4 2
20 8 12 0 12
21 8 9 3 3
22 5 6 6 1
23 4 7 5 1,4
24 3 3 9 0,3
Table 1: Hours trained per week and matches won/lost

As it can be seen, the addition of the matches played of the 24 players is 12. This is because in the
country club where I live, the season consists of 12 matches each player, and the one who wins
the most quantity of matches is the champion.
 Hours trained per week Average ratio won/lost
1 0,1
2 0,4
3 0,7
4 2,0
5 0,8
6 1,9
8 5,4
Table 2: Hours trained per week vs average ratio per hours trained

As it is seen in the Table 2, the players were divided into 7 different groups depending on how
much hours did they devote for training per week, and an average from the ratios of those players
was made. Later, a bar graph was plotted in order to display and simplify the data above. The
graph obtained was the following:

Average ratio of matches won/lost depending on hours

trained per week
AVERAGE RATIO OF MATCHES WON/LOST

0
1 2 3 4 5 6 8
HOURS TRAINED PER WEEK

Figure 1: Average ratio of matches won/lost depending on hours trained per week

Once the graph was obtained, I decided to plot a scatter diagram in order to “investigate the
possible relationship between two variables”1, or correlation. The scatter graph ended up as the
following:

1
Buchanan, Laurie, Fensom, Jim, Kemp, Ed, La Rondie, Paul and Stevens, Jill, “Mathematics Standard Level”,
Oxford University Press, Oxford, 2012. Page 334. 9/11/16 11:15am
 Correlation between hours of training tennis per week
and ratio of matches won/lost.
14
Ratio of matches won/lost

12
10
8
6
4
2
0
0 1 2 3 4 5 6 7 8 9
Hours of training per week

Figure 2: Correlation between hours of training tennis per week and ratio of matches won/lost

After analysing the graphs, I could notice “A general upward trend in the pattern of dots”2 which
show positive correlation. This means that “the value of the dependent variable increases as the
value of the independent variable increases”3. The ratio of matches won/lost increases as the
players devote more time to train per week. However, “a correlation between two data sets does
not necessarily mean that one causes the other”4 In order to find a direction of an association
between the two variables, I calculated the line of best fit, or regression line.

The formula for finding the gradient of a regression line is:

( ')( )) ' +
𝑆𝑥𝑦 = 𝑥𝑦 − and 𝑆' + = 𝑥- −
* *

2
Buchanan, Laurie, Fensom, Jim, Kemp, Ed, La Rondie, Paul and Stevens, Jill, “Mathematics Standard Level”,
Oxford University Press, Oxford, 2012. Page 335.
3
Ibid
4
Buchanan, Laurie, Fensom, Jim, Kemp, Ed, La Rondie, Paul and Stevens, Jill, “Mathematics Standard Level”,
Oxford University Press, Oxford, 2012 Page 336
As for the values of my survey, it would be:

x y xy x²
1 0,1 0,1 1
2 0,4 0,8 4
3 0,7 0,21 9
4 2 8 16
5 0,8 4 25
6 1,9 11,4 36
8 5,4 43,2 64
29 11,3 67,71 155

Therefore, S(xy) would be:

S(xy)=67,71 – (29 x 11.3)/7

S(xy)=20.9

and S(x)2 would be:

S(x)2=155- (29)2/7

S(x)2=34.86

The equation for the regression line is:

𝑦 − ȳ = 𝑆(𝑥𝑦)/𝑆(𝑥 - ). (𝑥 − 𝑥̄)

Therefore, it would be:

20.9
𝑦 − 1.6 = (𝑥 − 4.1)
34.86
𝑦 − 1.6 = 0.59 (𝑥 − 4.1)

𝑦 = 1.6 + 0.59 (𝑥 − 4.1)

𝑦 = 1.6 + 0.59 𝑥 − 2.42

𝑦 = 0.59 𝑥 − 0.82
 Correlation between hours of training tennis per
week and ratio of matches won/lost.
14
12
Ratio of matches won/lost

10
8
6
4
2
0
0 1 2 3 4 5 6 7 8 9
-2
Hours of training per week

The line of best fit, or the regression line clearly shows the tendency. The ratio of won/lost
matches increases as the hours devoted for training per week increase. Now, being able to
recognize this trend, we can calculate whether if the correlation is weak, moderate or strong by
calculating the r-value. This value will be between -1 and 1. The values between -1 and 0 will
indicate that there is a negative correlation, and the values between 0 and 1 will indicate that
there is a strong correlation. As closest the value is to 0, the weaker the correlation will be. This
can be seen in:

r-value Correlation

0 ˂ │r│≤ 0.25 Very weak

0.25 ˂ │r│≤ 0.5 Weak

0.5 ˂ │r│≤ 0.75 Moderate

 0.75 ˂ │r│≤ 1 Strong

Table 3 r-values

The r-value can be calculated using the Pearson´s correlation coefficient. “The Pearson product-
moment correlation coefficient is a measure of the correlation between two variables X and Y,
giving a value between +1 and -1 inclusive…”5 The formula for this is the following:

𝑆𝑥𝑦
𝑟=
𝑆𝑥𝑆𝑦
The value of S(xy) is already known due to the previous equations. However, we still need to know
the values of Sx and Sy. Such values can be calculated with the following equations:

( ')^- ( ))^-
𝑆𝑥 = 𝑥^2 − and 𝑆𝑦 = 𝑦^2 −
* *

(-?)^- (BB.C)^-
Therefore, Sx= 155 – = 5.9, and Sy= 38.07 – = 4.45
@ @

Now that we have all the values necessary, we can calculate the Pearson´s correlation
coefficient:
r= 20.9/(5.9x4.45)
r=20.9/26.255
r=0.796
As it can be seen, the r value is very proximate to 1. Therefore, this shows that there is a
strong correlation between the hours per week a player trains tennis and its results in their
results in the season.

Conclusion
In conclusion, it can be said that my parents were correct on what they told me. There is a
strong correlation between effort and results in tennis. As more hours you train, your ratio
between matches won/lost will increase. The correlation is strong but does not reach the
absolute 1 because there might be some cases were playing tennis is a natural gift for a
man, and therefore, this man may achieve more wins than losses training much less than
other tennis players. However, in the most quantity of cases, as you train more, your results
will be better. I should have trained more instead of getting frustrated and quitting, and I
should also have heard my parents.

5
Buchanan, Laurie, Fensom, Jim, Kemp, Ed, La Rondie, Paul and Stevens, Jill, “Mathematics Standard Level”,
Oxford University Press, Oxford, 2012. Page 349.