Log in

No account? Create an account
The Game Theory of LinkedIn - Suit up, son! You're going to Mars! [entries|archive|friends|userinfo]
Suit up, son! You're going to Mars!

[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

The Game Theory of LinkedIn [Apr. 16th, 2013|04:07 pm]
Suit up, son! You're going to Mars!
I've been on LinkedIn a lot lately.

I've been wondering about the game theory of connection requests. I tend to mentally classify connection requests in terms of "people who I can help" vs. "people who can help me." I find myself eager to accept requests in the latter category, and reluctant to accept requests in the former category. I concede that this may be totally unfair. And it leaves me wondering what my optimal strategy ought to be.

So, assume you are optimizing for career success, which means you want to hire good people when you have a job, and find good jobs when you don't.

Given that, is a big network necessarily better? Are there "bad" connections? Are there "dilution" effects?

If bigger networks are better, then connection requests are just ultimatum games; someone else is asking for permission to raise your utility and theirs at the same time, and you have veto power.

But if larger networks aren't better, then what does the payoff matrix look like? Who should you try to link to? Whose connections should you ignore?

From: (Anonymous)
2013-04-17 02:43 pm (UTC)
I'd say that being connected on LinkedIn means basically zero in and of itself in terms of how much someone is then willing to do for you (and presumably vice versa). I think links are a sign of pre-existing goodwill rather than an undertaking to increase efforts to assist (or ask for assistance) in the future. So it's a signalling issue, and there's some dilution to having a bigger network (or linking to someone with a bigger network) because having 1000 connections means (by default) that you're less likely to do something helpful for any of them than having 5 connections. Not because of the fact that you have so many connections, but because of what it says about your view of LinkedIn if you're willing to have so many.

If that makes sense...I guess in this conception of the site you pretty much just want to link to as many people as possible, since you're information gathering ("Will this person respond positively and imply that there's some relationship capital I can tap?") with very little downside (the risk that by having "too many" connections you turn off someone who would otherwise have been flattered to be asked to link, but I have serious doubts whether such a person would be willing to provide meaningful assistance anyway).
(Reply) (Thread)
From: bakedweasels
2013-04-17 02:44 pm (UTC)
Whoops, commented anonymously. Which was probably for the best given it was banal.
(Reply) (Thread)
[User Picture]From: algorithmancy
2013-04-18 07:00 am (UTC)
So, I don't think that it's just a matter of who will do favors for you in the future. Linking to someone rubs off some of your "street cred" on them; people form opinions of strangers based on the quality of their links. People also use connections as a pretext to contact other people in your network.

I get a fair number of requests from near-strangers, some of which might want to spam, stalk or otherwise annoy me. Worse, they might want to use my connection as a pretext to connect to the actual person they want to spam, stalk or annoy. So on further thought I do think that there are "bad connections" in that sense.

I also think there is the occasional "shady character" whose connection can sully your rep. For example, a hedge fund manager would probably not want to link to Bernie Madoff or Jeff Skilling.
(Reply) (Parent) (Thread)
From: bakedweasels
2013-04-18 02:43 pm (UTC)
I think there are probably different models for using the site. A lot of people basically use it to say, "I know you!" E.G., I just linked to a bunch of guys I coached at MIT a decade ago. And I get linked to all the time by people who aren't really in my professional circle but are friends.

Now, that's probably not the raison d'etre of the site and your use/view is probably closer to the company's view. But I'd suggest that they got caught out by people's desire to "connect"/look popular and that this new "endorsement/skill" model is a way to ACTUALLY impart "street cred" to people. I'm sure over time that will also get devalued and then they'll try to find other ways to signify real recommendations.
(Reply) (Parent) (Thread)
[User Picture]From: countertorque
2013-04-18 01:38 am (UTC)
I think you're trying to make a network if smart people. Networks of people you want to work with are better.
(Reply) (Thread)
[User Picture]From: kelkyag
2013-04-18 05:26 am (UTC)
A possible benefit to you in networking with "people you can help" is that a good chunk of that help may be in connecting them with other people you know who are looking for employees. Those employers may well appreciate your help in finding good employees and be more likely to help you or value your input in the future.

(Eventually, people who appreciate your help on the job-hunting front might include your offspring. I suspect you won't want to optimize them out of your network.)
(Reply) (Thread)
[User Picture]From: algorithmancy
2013-04-18 07:05 am (UTC)
Most of the "people I can help" are near-strangers, such as students who took a class I taught. Or random recruiters. Just connecting to me gives them helpful exposure, and possibly access to more of my network.
(Reply) (Parent) (Thread)
[User Picture]From: kelkyag
2013-04-20 03:41 am (UTC)
And you're not mediating the connection on LinkedIn. My impulse would be against random recruiters but in favor of students who took your class (who might grow into interesting people in or near your field), unless they were memorably problematic, but I don't claim to be backing that with any sort of data.
(Reply) (Parent) (Thread)