Jump to accessibility statement Skip to content

Footy data everywhere... but little hits the back of the net

Home / Study / Discover Sunderland / Business, Law and Tourism / Footy data everywhere... but little hits the net

Published: 29 November 2019

At the beginning of every football season predictions abound, ranging from fans’ hopes and fears to computer-generated calculations. This time last year, I picked one of these algorithm-based efforts* pretty much at random – though it was one that used a lot of calculations.

Sadly, the computers that carried out those calculations performed worse than most of the players of the team I follow.

Here’s the predicted and actual final placing and points totals.

 

predicted

placing

actual

placing

team

predicted

points

actual

points

1

4

West Bromwich Albion

82

80

2

2

Sheffield United

79

89

3

7

Middlesbrough

77

73

4

3

Leeds United

77

83

5

16

Stoke City

73

45

6

11

Brentford

72

64

7

10

Swansea City

68

65

8

1

Norwich City

66

94

9

6

Derby County

64

74

10

5

Aston Villa

64

76

11

9

Nottingham Forest

63

66

12

8

Bristol City

63

70

13

12

Sheffield Wednesday

70

64

14

17

Birmingham City

63

52

15

15

Blackburn Rovers

57

60

16

18

Wigan Athletic

56

52

17

21

Millwall

56

44

18

19

Queens Park Rangers

56

51

19

14

Preston North End

55

61

20

23

Bolton Wanderers

51

32

21

13

Hull City

49

62

22

22

Rotherham United

47

40

23

20

Reading

47

47

24

24

Ipswich Town

46

31

 

As the bold sections show; of the 48 separate predictions, the super-clever software managed to get only four correct – around 8%.

So why the innate failure of the power of smart computing?

Well, how about: players performing better than they had previously; players performing worse than they had previously; players getting injuries; players having a good day; players having a bad day; owners that sack managers; managers falling out with players; players falling out with managers; players falling out with players. For ‘my’ team, we can also add in; cheating opposition players and abysmal refereeing decisions.

Footballers contesting the ball

Do you notice the one thing that unites all of these? They all involve people. And people tend to act like, well… people. Not machines.

And the relevance of all this to business and marketing?

It is people who buy things. Not machines.

* https://fivethirtyeight.com/methodology/how-our-club-soccer-predictions-work


Alan Charlesworth is a Senior Lecturer and has been involved in what is now referred to as digital marketing as a practitioner, consultant, writer, trainer and educator since 1996. Alan also writes about and consults on the phenomenon commonly known as the Digital Transformation, a subject that transcends marketing into all aspects of business and management. Read more about his teaching and research interests and studying Business and Management at the University of Sunderland.   

ReciteMe accessibility toolbar button