General study in gaming and eSports, part 1.

If you just want to have a quick overview, look at the infographic. There is more content below the infographic for the people that want to read a bit about the topic. The drop-down content is mainly technical for people who either have a lot of time, commitment to the subject, or students.

Click on the picture to see a bigger version.

**Note that PUBG have very good values overall, probably because it is/was a new game.

Let’s get to know our audience.


Methodology: Academic research and respondents

The survey is based on existing academic research (namely “Why do you play? The development of the motives for online gaming questionnaire” by Demetrovics et al 2011, and is inspired by ESA’s annual report. The ESA reports have some flaws, and one of the main reasons people are criticizing them, is the fact that they say 97% of households in USA have a computer, which is 13% off other statistics (the American Community Survey), which means they are not showing a representative of the “real” audience. The reasons I write this is, even though we are inspired by them, we will not have this problem since we target gamers, specifically. Meaning that all our respondents already own computers.


Our audience (N =3664) is mainly Danish (75%), the rest (25%) is obtained by forums such as Reddit, Steam, and Facebook. Note that not all of 3664 answered all questions: the lowest amount of answers on a question is 3269. If we assume that 500 million people play computer worldwide, this survey will have a confidence interval of 99% and a margin of error around 2,2%. This is higher than academic standards.


70% of the respondents are between 15-29, and 13,4% is under 15. This is a bit lower than my other surveys conducted on Reddit, this time we also gathered responses elsewhere.

Question: What is your age?
Frequency Percent Valid Percent Cumulative Percent
Valid Under 15 years old 492 13,4 13,4 13,4
15-19 years old 1155 31,5 31,5 45,0
20-24 years old 886 24,2 24,2 69,2
25-29 years old 517 14,1 14,1 83,3
30-34 years old 216 5,9 5,9 89,2
35-39 years old 154 4,2 4,2 93,4
40 years old or older 243 6,6 6,6 100,0
Total 3663 100,0 100,0
Missing System 1 ,0
Total 3664 100,0

We see that 95,3% of our respondents are male, so we can’t really make any analysis of the difference between male and female players.  Another interesting fact is that the youngest players play CS (by far), and the oldest players play WoW.

It seems that gamers need their daily fix: 70,2% of the respondents play games daily, and 98% play a couple of times a week. – But there is a big difference in how much they play.

Play time for an average player?

Frequency Percent Valid Percent Cumulative Percent
Valid 1-5 hours/week 381 10,4 10,5 10,5
6-10 hours/week 653 17,8 18,0 28,5
11-15 hours/week 560 15,3 15,4 43,9
16-20 hours/week 532 14,5 14,7 58,6
21-25 hours/week 396 10,8 10,9 69,5
26-30 hours/week 360 9,8 9,9 79,4
31-35 hours/week 216 5,9 6,0 85,4
36-40 hours/week 151 4,1 4,2 89,5
More than 40 hours/week 380 10,4 10,5 100,0
Total 3629 99,0 100,0
Missing System 35 1,0
Total 3664 100,0

As we see the most common is 6-10 hours a week. 1-30 hours a week have around 10% each, the interesting thing here is that it dips down, from 31-40 hours, but then it goes up to 10% again at more than 40 hours. That raises a couple of questions. First one is: Does age have a correlation?

The short answer is yes, and it significant. However, it doesn’t seem to be the defining factor in how much time is spent playing computer.

The longer answer

Explaining the effect (R2 squared .020) The question is a text question, so in theory you make correlations based on hidden values. To give you an idea about what people answered, the formula is Y= 5,006 -0,215x.

Every cell in a survey is given a hidden value, imagine this:

  • Under 15
  • 15-19 years
  • 20-24 years

Everyone starts at value 5,006 in the “how much do you game each week” question. This is 21-25 hours weekly. So if you are between 20-24 years old, you will put 3 as “x”, and you will lose -0,645. It doesn’t seem as much but 1 point is the 5hours difference in average.

Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) 5,006 ,086 58,225 ,000
What is your age? -,215 ,025 -,143 -8,679 ,000
a. Dependent Variable: How many hours do you play computer in an average week?

Note: That this is only for guidance, it doesn’t take into consideration that +40 hours may be more than 40-44 hours. Each group is equally big.

Is there a correlation between playing with friends and playing computer?

As above there is a significant correlation between how much you play with friends, and how much you play computer. However, this isn’t the main drive either. We see that 84,9% of respondents play with friends more often than rarely (occasionally, frequently and always).

Other factors

2 other factors worth investigating are:

  1. If playing different games influence playtime in general, this will be covered in this post about how the different game types have an effect on: gaming time, motivations, reputation, and much more.
  2. The other one is if the motivation to play has an effect of play time, we will go into that after presenting the different motivations to play.

eSports survey results

  1. 78,3 % see eSport as a real sport.
  2. 22,9% don’t watch eSports, 48% watch a bit of eSports and 29 follow eSports.
  3. Where do people watch eSports? Twitch is far ahead with 73,8%, Youtube got 20,4% and 5,8% answered “others”.

Let’s dig deeper into the data, because there is something interesting in there.

Do you see eSport as a “”real”” sport? * Which sentence fits you best? Crosstabulation
Which sentence fits you best? Total
I dont watch esport I watch esport I watch abit of esport
Do you see eSport as a “”real”” sport? Yes Count 376 990 1474 2840
% within Do you see eSport as a “”real”” sport? 13,2% 34,9% 51,9% 100,0%
% within Which sentence fits you best? 45,2% 93,9% 84,6% 78,3%
No Count 456 64 269 789
% within Do you see eSport as a “”real”” sport? 57,8% 8,1% 34,1% 100,0%
% within Which sentence fits you best? 54,8% 6,1% 15,4% 21,7%
Total Count 832 1054 1743 3629
% within Do you see eSport as a “”real”” sport? 22,9% 29,0% 48,0% 100,0%
% within Which sentence fits you best? 100,0% 100,0% 100,0% 100,0%

54,8% of the people that “don’t usually watch eSports”, don’t see eSport as a real sport, this category is pulling the number down to 78%.  To the people saying they “watch eSport”, this number is 6,1%. So if we only asked people that watched eSports regularly, 93,9% would say eSport is a real sport. I think we see this in all categories, take for example golf: if you play it you want to defend it.

Another interesting factor is, if you play CS, League and PUBG you are more likely to see eSport as a real sport, around 86% of the players do. If you play Battlefield and GTA 5 it is only around 60% who see eSport as a real sport..


I was trying to measure 7 items: Coping, Skill development, Fantasy, Recreation, competition, Escape, and Social. So to do this I asked 24 questions. For example:

Competition got measured by the following (I play online games…): “because I enjoy competing with others”, “because I like to win”, “because it is good to feel that I am better than others” and “for the pleasure of defeating others”. The reason to ask these 4 questions is that if I only asked: “because I like to win?” it doesn’t really reflect completion alone.

Technical stuff

KMO anf Factor analysis

Long explanation: So, in a factor analysis we take these 4 questions and see if they correlate, the idea is, that if it is important for a respondent to win, the 3 other questions will most likely be pretty important. A factor analysis can confirm this.

I had to remove 1 item from the list “because it helps me get into a better mood.” it simply fit into everything I was trying to measure, and that is not a good sign in factor analysis.

KMO and Bartlett’s Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. ,888
Bartlett’s Test of Sphericity Approx. Chi-Square 32975,415
df 276
Sig. ,000


The stats looks awesome! And let’s take a quick look at the reliability. You can see a table with questions, factor loadings and Cronbach’s alpha.

Item Factor loadings Cronbachs alpha
Because gaming helps me escape reality 0,842 .849
Because it makes me forget real life 0,840
To forget about unpleasant thing or offenses 0,751
Because gaming helps me to forget about daily hassles 0,746
For the pleasure of defeating others 0,839 .825
Because I like to win 0,820
Because it is good to feel that I am better than others 0,803
Because I enjoy competing with others 0,685
Because gaming sharpens my sense 0,804 .827
Because it improves my concentration 0,801
Because it improves my coordination skills 0,760
Because it improves my skills 0,581
Because I can get to know new people 0,864 .812
Because I can meet many different people 0,853
Because it is a good social experience 0,670
For recreation 0,864 .650
Because it helps me get rid of stress 0,853
Because it reduce tension 0,670
Because I can do things that I am unable to do or I am not allowed to do in real life 0,804 .749
To feel as if I was somebody else 0,631
Because I can be in another world 0,613
Because I enjoy gaming 0,853 .744
Because it is entertaining 0,853

The respondents choices

When the respondents were asked the 24 questions, they could choose between these 5 answers:

1 ‘Strongly disagree’

2 ‘Disagree’

3 ‘Neither agree nor disagree’

4 ‘Agree’

5 ‘Strongly Agree’

To illustrate it in a simple way.

And here are the numbers:

Social Escape Competition Recreation Skill_improvement Fantasy Entertaining
Mean 3,41 3,00 3,56 3,91 3,52 3,12 4,66

As you see “escape” and “fantasy”, which scored the lowest, scored just over 3 “Neither agree nor disagree”.  “Entertaining” is by far the most important factor.

Correlations between factors.

Numbers, 2 tailed sig, pearson correlation
Social Escape Competition Recreation SkillDevelopment Fantasy Fun
Social Pearson Correlation 1 ,250** ,262** ,264** ,516** ,344** ,204**
Sig. (2-tailed) ,000 ,000 ,000 ,000 ,000 ,000
N 3265 3263 3264 3264 3265 3262 3264
Escape Pearson Correlation ,250** 1 ,117** ,336** ,257** ,622** ,074**
Sig. (2-tailed) ,000 ,000 ,000 ,000 ,000 ,000
N 3263 3266 3266 3265 3266 3265 3266
Competition Pearson Correlation ,262** ,117** 1 ,115** ,438** ,134** ,209**
Sig. (2-tailed) ,000 ,000 ,000 ,000 ,000 ,000
N 3264 3266 3268 3267 3267 3265 3267
Recreation Pearson Correlation ,264** ,336** ,115** 1 ,334** ,280** ,340**
Sig. (2-tailed) ,000 ,000 ,000 ,000 ,000 ,000
N 3264 3265 3267 3268 3267 3264 3267
SkillDevelopment Pearson Correlation ,516** ,257** ,438** ,334** 1 ,343** ,230**
Sig. (2-tailed) ,000 ,000 ,000 ,000 ,000 ,000
N 3265 3266 3267 3267 3269 3265 3268
Fantasy Pearson Correlation ,344** ,622** ,134** ,280** ,343** 1 ,109**
Sig. (2-tailed) ,000 ,000 ,000 ,000 ,000 ,000
N 3262 3265 3265 3264 3265 3265 3265
Fun Pearson Correlation ,204** ,074** ,209** ,340** ,230** ,109** 1
Sig. (2-tailed) ,000 ,000 ,000 ,000 ,000 ,000
N 3264 3266 3267 3267 3268 3265 3270
**. Correlation is significant at the 0.01 level (2-tailed).

Intern Correlations:

“Escape” and “Fantasy” are very strongly correlated (38% variance)

“Competition” and “Skill development” are strongly correlated (19% variance)

“Skill Development” and “Social” are strongly correlated (26,6% variance)

Correlations with playtime.

Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) 1,990 ,378 5,268 ,000
Social ,335 ,053 ,130 6,353 ,000
Escape ,254 ,050 ,113 5,044 ,000
Competition -,015 ,053 -,006 -,290 ,772
Recreation -,270 ,062 -,086 -4,324 ,000
SkillDevelopment ,147 ,065 ,051 2,285 ,022
Fantasy ,095 ,054 ,040 1,747 ,081
Fun ,164 ,080 ,038 2,045 ,041
a. Dependent Variable: How many hours do you play computer in an average week?

What we can see here is, if one of your main motivations to play is either to be social (meet new people) or to escape (from reality), you will most likely be playing more. However, if you play for recreation, you will be playing less. – The other items are not significant, and therefore not important to mention.

Similarities from a study 6 years ago:

What we can read from this, is that our respondents had a higher mean value in general, when we know this, we can see the following:

Skill development and competition have highly increased since then – This is most likely due to eSport growth. It is backed by the fact that the only motivation factor that has lost value is recreation (relaxing). You might also notice that fun/enjoyment is missing. – The reason for this is the way the factor analysis was conducted: we could only measure with 2 items out of 3, so it not fair to compare them.

Young people are the most competitive, and the older you get, the more you play for recreation.

Let’s have a quick discussion about this section because, while this measures which motivational factors gamers feel important when they play certain games, there are some pitfalls.

  1. I asked what games you play the most, if a respondent play 80% League of legends and 20% Hearthstone, then it is not really precise.
  2. Imagine that young people are competitive, and a lot of younger people play CS:GO. Which statement is right: that young people are more competitive, or that CS:GO players are more competitive? Both are true, but if I only tell you one of the facts, then you won’t get the full picture.
Like our content? - Share it with other gamers!Share on FacebookTweet about this on TwitterShare on Google+Share on LinkedIn

Leave a Reply

Your email address will not be published. Required fields are marked *