Friendship paradox

The friendship paradox is the phenomenon first

observed by the sociologist Scott L. Feld in 1991 that on
average, an individual's friends have more friends than
that individual.[1] It can be explained as a form of
sampling bias in which people with more friends are
more likely to be in one's own friend group. In other
words, one is less likely to be friends with someone who
has very few friends. In contradiction to this, most
people believe that they have more friends than their
friends have.[2][3][4][5]

The same observation can be applied more generally to

social networks defined by other relations than
friendship: for instance, most people's sexual partners
have had (on the average) a greater number of sexual
partners than they have.[6][7]
Diagram of a social network of 7-8-year-old
The friendship paradox is an example of how network children, mapped by asking each child to
structure can significantly distort an individual's local indicate two others they would like to sit next to
observations.[8][9] in class. The majority of children have fewer
connections than the average of those they are
connected to.
Mathematical explanation
In spite of its apparently paradoxical nature, the phenomenon is real, and can be explained as a
consequence of the general mathematical properties of social networks. The mathematics behind this are
directly related to the arithmetic-geometric mean inequality and the Cauchy–Schwarz inequality.[10]

Formally, Feld assumes that a social network is represented by an undirected graph G = (V, E), where the
set V of vertices corresponds to the people in the social network, and the set E of edges corresponds to the
friendship relation between pairs of people. That is, he assumes that friendship is a symmetric relation: if x
is a friend of y, then y is a friend of x. The friendship between x and y is therefore modeled by the edge
{x, y}, and the number of friends an individual has corresponds to a vertex's degree. The average number
of friends of a person in the social network is therefore given by the average of the degrees of the vertices
in the graph. That is, if vertex v has d(v) edges touching it (representing a person who has d(v) friends),
then the average number μ of friends of a random person in the graph is

The average number of friends that a typical friend has can be modeled by choosing a random person (who
has at least one friend), and then calculating how many friends their friends have on average. This amounts
to choosing, uniformly at random, an edge of the graph (representing a pair of friends) and an endpoint of
that edge (one of the friends), and again calculating the degree of the selected endpoint. The probability of a
certain vertex to be chosen is
The first factor corresponds to how likely it is that the chosen edge contains the vertex, which increases
when the vertex has more friends. The halving factor simply comes from the fact that each edge has two
vertices. So the expected value of the number of friends of a (randomly chosen) friend is

We know from the definition of variance that

where is the variance of the degrees in the graph. This allows us to compute the desired expected value
as

For a graph that has vertices of varying degrees (as is typical for social networks), is strictly positive,
which implies that the average degree of a friend is strictly greater than the average degree of a random
node.

Another way of understanding how the first term came is as follows. For each friendship (u, v), a node u
mentions that v is a friend and v has d(v) friends. There are d(v) such friends who mention this. Hence the
square of d(v) term. We add this for all such friendships in the network from both the u 's and v's
perspective, which gives the numerator. The denominator is the number of total such friendships, which is
twice the total edges in the network (one from the u 's perspective and the other from the v's).

After this analysis, Feld goes on to make some more qualitative assumptions about the statistical correlation
between the number of friends that two friends have, based on theories of social networks such as
assortative mixing, and he analyzes what these assumptions imply about the number of people whose
friends have more friends than they do. Based on this analysis, he concludes that in real social networks,
most people are likely to have fewer friends than the average of their friends' numbers of friends. However,
this conclusion is not a mathematical certainty; there exist undirected graphs (such as the graph formed by
removing a single edge from a large complete graph) that are unlikely to arise as social networks but in
which most vertices have higher degree than the average of their neighbors' degrees.

The Friendship Paradox may be restated in graph theory terms as “the average degree of a randomly
selected node in a network is less than the average degree of neighbors of a randomly selected node”, but
this leaves unspecified the exact mechanism of averaging (i.e., macro vs micro averaging). Let
be an undirected graph with and , having no isolated nodes. Let the set of
neighbors of node be denoted . The average degree is then .

Let the number of "friends of friends" of node be denoted . Note that this

can count 2-hop neighbors multiple times, but so does Feld's analysis. We have .
Feld considered the following "micro average" quantity.
However, there is also the (equally legitimate) "macro average" quantity, given by

The computation of MacroAvg can be expressed as the following pseudocode.

Algorithm MacroAvg

1. for each node

1. initialize
2. for each edge

1.

2.

3. return

"←" denotes assignment. For instance, "largest ← item" means that the value of largest changes to
the value of item.
"return" terminates the algorithm and outputs the following value.

Each edge contributes to MacroAvg the quantity , because

. We thus get

Thus, we have both and , but no inequality holds between them.[11]

Applications
The analysis of the friendship paradox implies that the friends of randomly selected individuals are likely to
have higher than average centrality. This observation has been used as a way to forecast and slow the
course of epidemics, by using this random selection process to choose individuals to immunize or monitor
for infection while avoiding the need for a complex computation of the centrality of all nodes in the
network.[12][13][14] In a similar manner, in polling and election forecasting, friendship paradox has been
exploited in order to reach and query well-connected individuals who may have knowledge about how
numerous other individuals are going to vote.[15] However, when utilized in such contexts, the friendship
paradox inevitably introduces bias by over-representing individuals with many friends, potentially skewing
resulting estimates.[16][17]

A study in 2010 by Christakis and Fowler showed that flu outbreaks can be detected almost two weeks
before traditional surveillance measures would do so by using the friendship paradox in monitoring the
infection in a social network.[18] They found that using the friendship paradox to analyze the health of
central friends is "an ideal way to predict outbreaks, but detailed information doesn't exist for most groups,
and to produce it would be time-consuming and costly."[19] This extends to the spread of ideas as well,
with evidence that the friendship paradox can be used to track and predict the spread of ideas and
misinformation through networks.[12][20] This observation has been explained with the argument that
individuals with more social connections may be the driving forces behind the spread of these ideas and
beliefs, and as such can be used as early-warning signals.[17]

Friendship paradox based sampling (i.e., sampling random friends) has been theoretically and empirically
shown to outperform classical uniform sampling for the purpose of estimating the power-law degree
distributions of scale-free networks.[21][22] This is because sampling the network uniformly will not collect
enough samples from the characteristic heavy tail part of the power-law degree distribution to properly
estimate it. However, sampling random friends incorporates more nodes from the tail of the degree
distribution (i.e., more high degree nodes) into the sample. Hence, friendship paradox based sampling
captures the characteristic heavy tail of a power-law degree distribution more accurately and reduces the
bias and variance of the estimation.[22]

The "generalized friendship paradox" states that the friendship paradox applies to other characteristics as
well. For example, one's co-authors are on average likely to be more prominent, with more publications,
more citations and more collaborators,[23][24][25] or one's followers on Twitter have more followers.[26]
The same effect has also been demonstrated for Subjective Well-Being by Bollen et al. (2017),[27] who
used a large-scale Twitter network and longitudinal data on subjective well-being for each individual in the
network to demonstrate that both a Friendship and a "happiness" paradox can occur in online social
networks.

See also
Second neighborhood problem
Self-evaluation maintenance theory
List of paradoxes

References
1. Feld, Scott L. (1991), "Why your friends have more friends than you do", American Journal of
Sociology, 96 (6): 1464–1477, doi:10.1086/229693 (https://doi.org/10.1086%2F229693),
JSTOR 2781907 (https://www.jstor.org/stable/2781907), S2CID 56043992 (https://api.seman
ticscholar.org/CorpusID:56043992).
2. Zuckerman, Ezra W.; Jost, John T. (2001), "What makes you think you're so popular? Self
evaluation maintenance and the subjective side of the "friendship paradox" " (http://www.psy
ch.nyu.edu/jost/Zuckerman%20&%20Jost%20(2001)%20What%20Makes%20You%20Thin
k%20You%27re%20So%20Popular1.pdf) (PDF), Social Psychology Quarterly, 64 (3): 207–
223, doi:10.2307/3090112 (https://doi.org/10.2307%2F3090112), JSTOR 3090112 (https://w
ww.jstor.org/stable/3090112).
 3. McRaney, David (2012), You are Not So Smart (https://books.google.com/books?id=9Oc_hd
vqk50C&pg=PA160), Oneworld Publications, p. 160, ISBN 978-1-78074-104-8
4. Felmlee, Diane; Faris, Robert (2013), "Interaction in social networks", in DeLamater, John;
Ward, Amanda (eds.), Handbook of Social Psychology (2nd ed.), Springer, pp. 439–464,
ISBN 978-9400767720. See in particular "Friendship ties", p. 452 (https://books.google.com/
books?id=hXY8AAAAQBAJ&pg=PA452).
5. Lau, J. Y. F. (2011), An Introduction to Critical Thinking and Creativity: Think More, Think
Better (https://books.google.com/books?id=KCOeRQW2s0cC&pg=PA191), John Wiley &
Sons, p. 191, ISBN 978-1-118-03343-2
6. Kanazawa, Satoshi (2009), "The Scientific Fundamentalist: A Look at the Hard Truths About
Human Nature – Why your friends have more friends than you do" (https://web.archive.org/w
eb/20091107211753/http://www.psychologytoday.com/print/34391), Psychology Today,
archived from the original (http://www.psychologytoday.com/blog/the-scientific-fundamentalis
t/200911/why-your-friends-have-more-friends-you-do) on 2009-11-07.
7. Burkeman, Oliver (30 January 2010), "This column will change your life: Ever wondered why
your friends seem so much more popular than you are? There's a reason for that" (https://ww
w.theguardian.com/lifeandstyle/2010/jan/30/change-your-life-friends-popular), The
Guardian.
8. Lerman, Kristina; Yan, Xiaoran; Wu, Xin-Zeng (2016-02-17). "The "Majority Illusion" in
Social Networks" (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4757419). PLOS ONE. 11
(2): e0147617. arXiv:1506.03022 (https://arxiv.org/abs/1506.03022).
Bibcode:2016PLoSO..1147617L (https://ui.adsabs.harvard.edu/abs/2016PLoSO..1147617
L). doi:10.1371/journal.pone.0147617 (https://doi.org/10.1371%2Fjournal.pone.0147617).
ISSN 1932-6203 (https://www.worldcat.org/issn/1932-6203). PMC 4757419 (https://www.ncb
i.nlm.nih.gov/pmc/articles/PMC4757419). PMID 26886112 (https://pubmed.ncbi.nlm.nih.gov/
26886112).
9. Alipourfard, Nazanin; Nettasinghe, Buddhika; Abeliuk, Andrés; Krishnamurthy, Vikram;
Lerman, Kristina (2020-02-05). "Friendship paradox biases perceptions in directed
networks" (https://dx.doi.org/10.1038/s41467-020-14394-x). Nature Communications. 11 (1):
707. arXiv:1905.05286 (https://arxiv.org/abs/1905.05286). Bibcode:2020NatCo..11..707A (htt
ps://ui.adsabs.harvard.edu/abs/2020NatCo..11..707A). doi:10.1038/s41467-020-14394-x (htt
ps://doi.org/10.1038%2Fs41467-020-14394-x). ISSN 2041-1723 (https://www.worldcat.org/is
sn/2041-1723). PMC 7002371 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7002371).
PMID 32024843 (https://pubmed.ncbi.nlm.nih.gov/32024843).
10. Ben Sliman, Malek; Kohli, Rajeev (2019), "The extended directed friendship paradox" (http
s://ssrn.com/abstract=3395317), SSRN, doi:10.2139/ssrn.3395317 (https://doi.org/10.2139%
2Fssrn.3395317), S2CID 219376223 (https://api.semanticscholar.org/CorpusID:219376223)
11. Gupta, Yash; Chakrabarti, Soumen (2021), Friends of friends (https://www.cse.iitb.ac.in/~sou
men/doc/2021_Friends/FriendsOfFriends.pdf) (PDF)
12. Cohen, Reuven; Havlin, Shlomo; ben-Avraham, Daniel (2003), "Efficient immunization
strategies for computer networks and populations", Phys. Rev. Lett., 91 (24), 247901,
arXiv:cond-mat/0207387 (https://arxiv.org/abs/cond-mat/0207387),
Bibcode:2003PhRvL..91x7901C (https://ui.adsabs.harvard.edu/abs/2003PhRvL..91x7901
C), doi:10.1103/PhysRevLett.91.247901 (https://doi.org/10.1103%2FPhysRevLett.91.24790
1), PMID 14683159 (https://pubmed.ncbi.nlm.nih.gov/14683159), S2CID 919625 (https://api.
semanticscholar.org/CorpusID:919625).
13. Christakis, N. A.; Fowler, J. H. (2010), "Social network sensors for early detection of
contagious outbreaks", PLOS ONE, 5 (9), e12948, arXiv:1004.4792 (https://arxiv.org/abs/100
4.4792), Bibcode:2010PLoSO...512948C (https://ui.adsabs.harvard.edu/abs/2010PLoSO...5
12948C), doi:10.1371/journal.pone.0012948 (https://doi.org/10.1371%2Fjournal.pone.00129
48), PMC 2939797 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2939797),
PMID 20856792 (https://pubmed.ncbi.nlm.nih.gov/20856792).
14. Wilson, Mark (November 2010), "Using the friendship paradox to sample a social network",
Physics Today, 63 (11): 15–16, Bibcode:2010PhT....63k..15W (https://ui.adsabs.harvard.edu/
abs/2010PhT....63k..15W), doi:10.1063/1.3518199 (https://doi.org/10.1063%2F1.3518199).
15. Nettasinghe, Buddhika; Krishnamurthy, Vikram (2019). " "What Do Your Friends Think?":
Efficient Polling Methods for Networks Using Friendship Paradox" (https://dx.doi.org/10.110
9/tkde.2019.2940914). IEEE Transactions on Knowledge and Data Engineering: 1.
arXiv:1802.06505 (https://arxiv.org/abs/1802.06505). doi:10.1109/tkde.2019.2940914 (http
s://doi.org/10.1109%2Ftkde.2019.2940914). ISSN 1041-4347 (https://www.worldcat.org/issn/
1041-4347). S2CID 3335133 (https://api.semanticscholar.org/CorpusID:3335133).
16. Feld, Scott L.; McGail, Alec (September 2020). "Egonets as systematically biased windows
on society" (https://www.cambridge.org/core/product/identifier/S2050124220000053/type/jou
rnal_article). Network Science. 8 (3): 399–417. doi:10.1017/nws.2020.5 (https://doi.org/10.10
17%2Fnws.2020.5). ISSN 2050-1242 (https://www.worldcat.org/issn/2050-1242).
S2CID 216301650 (https://api.semanticscholar.org/CorpusID:216301650).
17. Galesic, Mirta; Bruine de Bruin, Wändi; Dalege, Jonas; Feld, Scott L.; Kreuter, Frauke;
Olsson, Henrik; Prelec, Drazen; Stein, Daniel L.; van der Does, Tamara (July 2021). "Human
social sensing is an untapped resource for computational social science" (https://www.natur
e.com/articles/s41586-021-03649-2). Nature. 595 (7866): 214–222.
Bibcode:2021Natur.595..214G (https://ui.adsabs.harvard.edu/abs/2021Natur.595..214G).
doi:10.1038/s41586-021-03649-2 (https://doi.org/10.1038%2Fs41586-021-03649-2).
ISSN 1476-4687 (https://www.worldcat.org/issn/1476-4687). PMID 34194037 (https://pubme
d.ncbi.nlm.nih.gov/34194037). S2CID 235697772 (https://api.semanticscholar.org/CorpusID:
235697772).
18. Christakis, Nicholas A.; Fowler, James H. (September 15, 2010). "Social Network Sensors
for Early Detection of Contagious Outbreaks" (https://www.ncbi.nlm.nih.gov/pmc/articles/PM
C2939797). PLOS ONE. 5 (9): e12948. arXiv:1004.4792 (https://arxiv.org/abs/1004.4792).
Bibcode:2010PLoSO...512948C (https://ui.adsabs.harvard.edu/abs/2010PLoSO...512948C).
doi:10.1371/journal.pone.0012948 (https://doi.org/10.1371%2Fjournal.pone.0012948).
PMC 2939797 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2939797). PMID 20856792
(https://pubmed.ncbi.nlm.nih.gov/20856792).
19. Schnirring, Lisa (Sep 16, 2010). "Study: Friend 'sentinels' provide early flu warning" (https://
web.archive.org/web/20130506173925/http://www.cidrap.umn.edu/cidrap/content/influenza/
general/news/sep1610friends.html). CIDRAP News. Archived from the original (http://www.ci
drap.umn.edu/cidrap/content/influenza/general/news/sep1610friends.html) on May 6, 2013.
Retrieved August 14, 2012.
20. Kumar, Vineet; Krackhardt, David; Feld, Scott (2021-05-18). "Interventions with Inversity in
Unknown Networks Can Help Regulate Contagion". arXiv:2105.08758 (https://arxiv.org/abs/
2105.08758) [cs.SI (https://arxiv.org/archive/cs.SI)].
21. Eom, Young-Ho; Jo, Hang-Hyun (2015-05-11). "Tail-scope: Using friends to estimate heavy
tails of degree distributions in large-scale complex networks" (https://dx.doi.org/10.1038/srep
09752). Scientific Reports. 5 (1): 9752. arXiv:1411.6871 (https://arxiv.org/abs/1411.6871).
Bibcode:2015NatSR...5E9752E (https://ui.adsabs.harvard.edu/abs/2015NatSR...5E9752E).
doi:10.1038/srep09752 (https://doi.org/10.1038%2Fsrep09752). ISSN 2045-2322 (https://ww
w.worldcat.org/issn/2045-2322). PMC 4426729 (https://www.ncbi.nlm.nih.gov/pmc/articles/P
MC4426729). PMID 25959097 (https://pubmed.ncbi.nlm.nih.gov/25959097).
22. Nettasinghe, Buddhika; Krishnamurthy, Vikram (2021-05-19). "Maximum Likelihood
Estimation of Power-law Degree Distributions via Friendship Paradox-based Sampling" (htt
ps://dx.doi.org/10.1145/3451166). ACM Transactions on Knowledge Discovery from Data.
15 (6): 1–28. arXiv:1908.00310 (https://arxiv.org/abs/1908.00310). doi:10.1145/3451166 (htt
ps://doi.org/10.1145%2F3451166). ISSN 1556-4681 (https://www.worldcat.org/issn/1556-46
81). S2CID 199064540 (https://api.semanticscholar.org/CorpusID:199064540).
23. Eom, Young-Ho; Jo, Hang-Hyun (2014), "Generalized friendship paradox in complex
networks: The case of scientific collaboration", Scientific Reports, 4, 4603, arXiv:1401.1458
(https://arxiv.org/abs/1401.1458), Bibcode:2014NatSR...4E4603E (https://ui.adsabs.harvard.
edu/abs/2014NatSR...4E4603E), doi:10.1038/srep04603 (https://doi.org/10.1038%2Fsrep04
603), PMC 3980335 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3980335),
PMID 24714092 (https://pubmed.ncbi.nlm.nih.gov/24714092)
24. Grund, Thomas U. (2014), "Why Your Friends Are More Important And Special Than You
Think" (http://www.sociologicalscience.com/download/volume%201/april/why-your-friends-a
re-more-important-and-special-than-you-think.pdf) (PDF), Sociological Science, 1: 128–140,
doi:10.15195/v1.a10 (https://doi.org/10.15195%2Fv1.a10)
25. Dickerson, Kelly (16 January 2014). "Why Your Friends Are Probably More Popular, Richer,
and Happier Than You" (http://www.slate.com/blogs/business_insider/2014/01/16/friendship
_paradox_why_are_my_friends_better_off_than_me.html). Slate Magazine. The Slate
Group. Retrieved 17 January 2014.
26. Hodas, Nathan; Kooti, Farshad; Lerman, Kristina (May 2013). "Friendship Paradox Redux:
Your Friends are More Interesting than You". arXiv:1304.3480 (https://arxiv.org/abs/1304.348
0) [cs.SI (https://arxiv.org/archive/cs.SI)].
27. Bollen, Johan; Goncalves, Bruno; Van de Leemput, Ingrid; Guanchen, Ruan (2017), "The
happiness paradox: your friends are happier than you", EPJ Data Science, 6,
arXiv:1602.02665 (https://arxiv.org/abs/1602.02665), doi:10.1140/epjds/s13688-017-0100-1
(https://doi.org/10.1140%2Fepjds%2Fs13688-017-0100-1), S2CID 2044182 (https://api.sem
anticscholar.org/CorpusID:2044182)

External links
Strogatz, Steven (September 17, 2012). "Friends You Can Count On" (http://opinionator.blog
s.nytimes.com/2012/09/17/friends-you-can-count-on/). New York Times. Retrieved
17 January 2013.

Retrieved from "https://en.wikipedia.org/w/index.php?title=Friendship_paradox&oldid=1165049646"