Sending more participants gives more

medals - analysis of SportsStats (Olympic

Dataset – 120 years of data)
Data Analysis Project for SQL for Data Science Capstone Project Course
March 2021
Anna Witkowiak
Client
This presentation is planned to be seen by my colleagues from this
course. I believe that they will be interested in my approach to analyse
Olympic Games Data Set the same way I’m interested in their analysies.
Questions to Answer:
1. How many participants are sent by countries?
• Number of participants can change over time.
• Some countries can send more participants than the others

2. How many medals for country in a specific year?

• This will show if the countries with higher rank in number of

participants will have higher rank in number in medals
Initial Hypotheses
1. Number of participants will be larger every year.
• Traveling is less expensive and easier every year so countries can send
more participants.
2. The same teams will have highest ranks year after year. Those will be the
same countries that send more participants than the others.
• If you can afford to send more participants, you can afford to prepare them
in the proper way.
3. There will be more men than women participants but this will change over
time
• For many countries men’s medals are more important than women’s
Data Analysis Approach
1. I will add ranks for the countries by number of participants and by
number of medals.
• I will focus on descriptive stats of difference between ranks
Data Analysis Approach
2. I will build a table with number of medals and number of
participants. This table will also consist information about year, season,
olympic team (noc) and sex.
• I will check average and standard deviation for number of medals and
number of participants.
• I will divide my set into four groups: Summer-Men, Summer-Women,
Winter-Men, Winter-Women.
• I will check correlation between average number of participants and
average number of medals
Data Analysis Approach
3. I will check how percent of women participants will change over
time.
• I will check how many percent of women participants were in every
game (year and season)
• I will also check this information by country – how many percent of
women participants every team has
Entity-Relationship Diagram (ERD)
Number of participants in Olympic Games
Year season number_of_participants Year season number_of_participants
1,896 Summer 176 1,924 Winter 313
1,900 Summer 1,224 1,928 Winter 461
1,904 Summer 650 1,932 Winter 252
1,906 Summer 841 1,936 Winter 668
1,908 Summer 2,024 1,948 Winter 668
1,912 Summer 2,409 1,952 Winter 694
1,920 Summer 2,676 1,956 Winter 821
1,924 Summer 3,256 1,960 Winter 665
1,928 Summer 3,247 1,964 Winter 1,094
1,932 Summer 1,922 1,968 Winter 1,16
1,936 Summer 4,484 1,972 Winter 1,008
1,948 Summer 4,402 1,976 Winter 1,128
1,952 Summer 4,932 1,980 Winter 1,071
1,956 Summer 3,347 1,984 Winter 1,273
1,960 Summer 5,352 1,988 Winter 1,425
1,964 Summer 5,137 1,992 Winter 1,801
1,968 Summer 5,558 1,994 Winter 1,738
1,972 Summer 7,114 1,998 Winter 2,179
1,976 Summer 6,073 2,002 Winter 2,399
1,980 Summer 5,259 2,006 Winter 2,494
1,984 Summer 6,798 2,010 Winter 2,536
1,988 Summer 8,454 2,014 Winter 2,745
1,992 Summer 9,386
1,996 Summer 10,339
2,000 Summer 10,647
2,004 Summer 10,557
2,008 Summer 10,899
2,012 Summer 10,517
2,016 Summer 11,179
Number of participants in Olympic Games
Countries and medal ranks
• I built two metrics: rank by number Year season noc
number_participan medals_numb number_participants_ra game_ran participants_medals_d
ts er nk k iff

of participants and rank by number 1,896 Summer SWE 1 0 10 10 0

of medals for teams in every

1,896 Summer USA 14 19 3 2 1
1,896 Summer GRE 102 44 1 1 0

Olympic Games 1,896

1,896
Summer GER
Summer AUS
19
1
14
3
2
10
3
9
-1
1

• I checked the differences between

1,896 Summer AUT 3 5 7 8 -1
1,896 Summer HUN 7 6 6 6 0

those ranks 1,896

1,896
Summer GBR
Summer ITA
10
1
9
0
5
10
5
10
0
0

• Fun fact: In 1896 Summer Olympic

1,896 Summer FRA 12 11 4 4 0
1,896 Summer DEN 3 6 7 6 1

games there were teams that had 1,896

1,900
Summer
Summer
SUI
BEL
3
64
3
18
7
5
9
4
-2
1

more medals than participants. 1,900 Summer ESP 9 1 13 19 -6

Australia had only one participant

1,900 Summer IRI 1 0 22 22 0
1,900 Summer NZL 1 1 22 19 3

and he has got 3 medals. His name 1,900

1,900
Summer BRA
Summer NOR
1
7
0
5
22
14
22
11
0
3

was Edwin Harold "Teddy" Flack. 1,900

1,900
Summer ARG
Summer LUX
1
1
0
1
22
22
22
19
0
3
Descriptive statistics for participants_medals_diff -
Summer
Year season min max absolute_min absolute_max avg
1,896 Summer -2 1 0 2 -0.0833333333
1,900 Summer -6 12 0 12 0.5483870968
1,904 Summer -8 7 0 8 0.5333333333
1,906 Summer -8 5 0 8 -0.0476190476
1,908 Summer -11 5 0 11 0.3636363636
1,912 Summer -16 6 0 16 -1.7931034483
1,920 Summer -10 9 0 10 -0.4482758621
1,924 Summer -34 18 0 34 -1.3111111111
1,928 Summer -18 18 0 18 -0.1304347826
1,932 Summer -30 13 0 30 -1.170212766
1,936 Summer -30 19 0 30 -2.1428571429
1,948 Summer -30 32 0 32 -1.2542372881
1,952 Summer -34 36 0 36 -3.3333333333
1,956 Summer -51 37 0 51 -7.625
1,960 Summer -54 43 0 54 -8.4642857143
1,964 Summer -65 44 0 65 -10
1,968 Summer -93 46 0 93 -17.7321428571
1,972 Summer -86 67 0 86 -19.8760330579
1,976 Summer -64 26 0 64 -13.6847826087
1,980 Summer -57 37 0 57 -11.1875
1,984 Summer -114 56 0 114 -29.8214285714
1,988 Summer -128 118 0 128 -36.6981132075
1,992 Summer -141 89 0 141 -33.4733727811
1,996 Summer -145 120 0 145 -35.0659898477
2000 Summer -165 82 0 165 -38.105
2,004 Summer -154 80 0 154 -40.5323383085
2,008 Summer -140 85 0 140 -35.137254902
2,012 Summer -154 93 0 154 -30.2975609756
2,016 Summer -143 77 0 143 -35.8985507246
Descriptive statistics for
participants_medals_diff - Summer
Descriptive statistics for
participants_medals_diff - Winter
Year season min max absolute_min absolute_max avg
1,924 Winter -13 10 1 13 -0.0526315789
1,928 Winter -15 8 0 15 -2.4
1,932 Winter -10 9 0 10 -0.4117647059
1,936 Winter -23 13 0 23 -4.3214285714
1,948 Winter -18 7 0 18 -4.0357142857
1,952 Winter -18 11 0 18 -3.4666666667
1,956 Winter -22 21 0 22 -5.65625
1,960 Winter -23 7 0 23 -4.2666666667
1,964 Winter -32 14 0 32 -6.6111111111
1,968 Winter -28 18 0 28 -6.4324324324
1,972 Winter -17 14 0 17 -3.5714285714
1,976 Winter -29 18 0 29 -6.0540540541
1,980 Winter -22 23 0 23 -4.1351351351
1,984 Winter -28 14 0 28 -7.4081632653
1,988 Winter -38 20 0 38 -11.1052631579
1,992 Winter -46 43 0 46 -13.03125
1,994 Winter -46 17 0 46 -8.6865671642
1,998 Winter -43 42 0 43 -9.7638888889
2,002 Winter -52 16 0 52 -11.8961038961
2,006 Winter -45 12 0 45 -8.3670886076
2,010 Winter -43 14 0 43 -9.1341463415
2,014 Winter -42 19 0 42 -10.4943820225
Descriptive statistics for
participants_medals_diff - Winter
Descriptive statistics for
participants_medals_diff - conclusion
• I expected that all statistics will be close to zero. But they are not. In
winter 1924 min from absolute value was one. That means that none
of partcipants number and game ranks were the same.
• Fun fact: I noticed that in 1904 the standard deviation of number of
participants (and number of medals) for women in summer games
was zero. That could mean that all the women teams on that games
had exactly the same number of participants and got exactly the same
number of medals. I checked in the data that there was only one such
team, and it was USA.
Average and standard deviation for number of medals and
number of participants example for Summer-Men
noc season sex average_part_number stdev_part__number average_medal_number stdev_medals_number
URS Summer M 274.6666666667 37.6231311828 82.5555555556 27.7943839251
EUN Summer M 310 [NULL] 75 [NULL]
USA Summer M 264.9285714286 106.8390719802 67.5 35.8272604673
GDR Summer M 190 35.7281401699 42.4 18.0776104616
FRG Summer M 263.4 45.610305853 30.4 7.3348483284
GER Summer M 189.05 100.6366444717 28.35 17.7149208593
RUS Summer M 146.4 103.9136393571 26.2 22.0544780033
GBR Summer M 195.8965517241 120.6107507728 24.2413793103 23.5579152188
FRA Summer M 195.9655172414 125.8035551277 22.8275862069 17.844705028
ITA Summer M 157.8275862069 73.0890205579 18 9.1612538131
CHN Summer M 110.8461538462 82.0516566498 18 14.640127504
SWE Summer M 116.0357142857 80.5212105219 16.5357142857 16.5383552921
JPN Summer M 131.3181818182 66.9601063631 14.9090909091 9.0496466172
HUN Summer M 108.5925925926 53.7324277245 14.4814814815 7.4594821968
UKR Summer M 123.6666666667 20.5491281242 11.6666666667 2.7325202043
AUS Summer M 125.8888888889 101.5507956767 11.5555555556 8.919871219
FIN Summer M 75.3461538462 47.0938996539 11.5 10.8268185539
KOR Summer M 106.8235294118 71.8516138425 10.4705882353 8.6321900977
POL Summer M 120 60.2652865413 10.2272727273 7.7270817394
KAZ Summer M 69.5 10.89495296 8.5 1.8708286934
Linear regression: strong correlation between
average_part_number and average_medal_number

Summer-Men
corr r2 regr_intercept regr_slope cnt
0.8885850837 0.789583451 16.7894246363 4.5224768144 230

Summer-Women
corr r2 regr_intercept regr_slope cnt
0.8756011598 0.7666773911 7.1624204141 4.3315877441 222

Winter-Men
corr r2 regr_intercept regr_slope cnt
0.8226198447 0.6767034089 6.7220461281 6.1787707995 114

Winter-Women
corr r2 regr_intercept regr_slope cnt
0.8129183736 0.6608362822 4.2185210832 3.0377191215 90
Percent of women participants growing over
time
• At Olympic Games in 1896 there were no women Year season female_perc
participants and now they are almost half of all 1,896 Summer 0
participants. 1,900 Summer 1.8790849673
1,904 Summer 0.9230769231
1,906 Summer 0.7134363853
1,908 Summer 2.1739130435
1,912 Summer 2.200083022
1,920 Summer 2.9147982063
1,924 Summer 4.7911547912
1,924 Winter 4.1533546326
…
…
…
2,002 Winter 36.9320550229
2,004 Summer 40.7312683528
2,006 Winter 38.2919005613
2,008 Summer 42.2882833287
2,010 Winter 40.7334384858
2,012 Summer 44.2521631644
2,014 Winter 40.14571949
2,016 Summer 45.0308614366
Percent of women participants growing over time
– correlation between „Year” and female_perc
corr r2 regr_intercept regr_slope cnt
0.9520753909 0.90644755 1,915.8001528124 2.4633563774 51
Which of the team has the highest percent of
women participants
noc season female_perc
HKG Winter 80
TLS Summer 62.5
KOS Summer 62.5
PRK Winter 61.1940298507
CHN Winter 60.7594936709
CHN Summer 54.0057452921
BHU Summer 51.8518518519
UZB Winter 51.8518518519
MHL Summer 50
PLW Summer 50
CPV Summer 50
LCA Summer 50
ANG Summer 48.347107438
BLR Summer 46.9026548673
PRK Summer 46.6507177033
UKR Summer 46.1147421932
VIE Summer 45.5284552846
RUS Summer 43.908045977
UKR Winter 43.6666666667
DEN Winter 43.2432432432
Percent of women participants in teams
• There are many teams that never had any women participants, but there are also teams where most of the
participants were women.

Summer
avg min max
24.0928672197 0 62.5

Winter
avg min max
20.5519064359 0 80
Insights Discovered
• Analysis of Olympic Games dataset proves that number of
participants per country is highly correlated to rank in medals.
At first I build metrics with ranks and and number of participants and I
checked how they behave for olympic teams. I believed that looking at
those data would give me the answers for my main hyphotesis. I would
say that this approach led me to nowhere specific.
Second approach was much simpler: a checked corelation between
average number of medals and average number of participants per
team. They were highly correlated.
Insights Discovered
• I proved that percent of women participants is growing every year.
• 80% of participants of Hong Kong team in winter Olympic games were
women
• There are some teams that never had women participants in Olympic
Games
Recommendations and Actions
• In my analyses I focused on specific questions. I believe that this
dataset can give us much more. I didn’t even consider information
about age, height and weight of participants. I belived that there was
no place for them in this specific study but there is opportunity for
more analyses. Even with the data that I used you can find more
questions to answer.
• This data is becoming incomplete. It should be updated with new
data after every olympic games. It would be nice to know if the trends
that were observed would continue. Maybe they will slow down and
we would observe some kind of stagnation or maybe they would even
reverse. The future will tell.