Joel Spolsky on managing large sites + dealing with Idiots + non-citable dogs

Nice things

We at thetrainline were lucky enough to have a visit from Joel Spolsky @spolsky.  This was a good thing and bad.  For instance, in ensuring that I got the spelling for his website correct, I have just spent 30 mins reading the damn thing.

…and it’s still more interesting that writing up his talk….damn another 15 minutes lost.

Focus Mr Shoop.

Anyway back to Mr Spolsky.  He talked to us about several things

The below is my take on what he was saying and what I found interesting rather than just a precis of his talk.  Therefore it contains my analysis of what I think we was meaning ..which of course may be wrong.  Joel if you are reading this and disagree – then please comment.  Also  – Joel if you are reading this….really?

A group is its own worst enemy

Worst enemy

You start small with an interest group and then after you grow to around 10,000 users the chances are pretty good that the group will now have attracted someone who is more interested in playing with the group that what the group was setup for.

This individual will  then start behaving like a teenager pushing the limits of what is permissible and will keep going until they provoke a reaction.  And they cry hard and loud that they are being oppressed.  Then the freedom loving members of the actual community by default come down on the side of the individual and the scene is set for a time-wasting, meaningless discussion on what should and shouldn’t be allowed.

Evidently Clay Shirkey has researched the issues relating to the growth of online communities.  (the article is a somewhat academic phrasing of a problem that Joel has addressed…in a way that Clay indicated that such problems need to be addressed)  A convergence of ease of action, instant gratification , an audience and anonymity unleashes the worst of behaviours in some people.

Thus something that was nice is now broken. So rules/processes have to be created in order to have nice things.  These are not optional if you still want the nice things.

 Diversion – Wikipedia – some things you probably didn’t know

Joel then meandered away from the subject to talk about Wikipedia.  Thinking about this, as I now write-up the talk, this was more or a storytellers approach (a Ronnie Corbett type story for those old enough to have watched the Two Ronnies in their youth) to meander off into a seemingly unrelated but entertaining topic and thus surprisingly it arcs back to the topic in hand. on entertaining meander…..

The wikipedia rule

Q:  What does Wikipedia publish?

A:  It seems that the answer isn’t  The Truth.  Or even those aspects of The Truth that can be verified.  But actually it’s – what other sources have reported about the truth. ie it’s a tertiary source.  This is why Philip Roth couldn’t directly change one of the articles about himself. His direct change to Wikipedia couldn’t be referenced to a source outside Wikipedia.  I guess if he had the change written up in the Guardian, then Wikipedia could have made the change and referenced the Guardian article.

For the same reason Joel couldn’t get an article about his dog Taco. There are no external articles about his dog that Wikipedia could refer to and so there can be no entry in Wikipedia – in their language Taco the dog is WP:N ie not notable.

Note that this blog has no such restriction


As an aside, which again was another spiral that we would come back to, Wikipedia is running short of admins (those who regularly edit).  This could be due to the conflicts around notability – eg “Wikipedia doesn’t want things that are correct or provable only what notable people say.”  This is a misunderstanding of their rules and is causing issues as people assume they know what notable means without actually knowing how Wikipedia uses the term (ie notable does not equal famous)

Anyway back to the mains story – why did Wikipedia use this rule?  Because if it didn’t then facts couldn’t be checked and so could be filled with incorrect info.

But linking back to the main topic that if you create something large then bad things will happen.  There will be people with a vested interest in having a page in Wikipedia read a certain way – and there will be others who will just get a kick out of having their edit in a page even if, or especially if their contribution is wrong.  Thus by insisting that whatever appears in Wikipedia must have appeared elsewhere then the worst of this is mitigated and by using citations at least everyone will know from where Wikipedia is getting the information it publishes and can form an opinion on its likely veracity accordingly.

As an example of what happens when a reasonableness test is removed – look at comments on YouTube … and it is just drivel.

But this rule for Wikipedia contribution has some bad effects eg Philip Roth couldn’t correct a statement about himself that he felt was wrong.

But remember the rule of idiots – Wikipedia has now scaled so that it is no longer dealing with just reasonable people.  The rule is to lock out the idiots and idiots who will try to use unreasonableness as a reason to let them ruin the site.  So I guess Wikipedia would rather than a restricted goal that can be defended rather than the nice thing to break.  The notability rule wasn’t there at the start of Wikipedia, it was only needed when it got large.  Timing is everything,  Mr Roth

The rule

I guess this is an example of trying for the best result possible rather than the best possible result.

But but but, in enforcing a rule you are going to lose friends (remember those declining Wikipedia admins) so rather than not enforcing it, and letting in the idiots, its a good idea to see if you can work on the presentation of the restriction eg Stack Overflow (of which Joel was a co-founder)  introduced a rule that said that questions should be closed if the questions were a matter of opinion.

This was important as he realised that the main beneficiaries of a question and answer session weren’t the original questioner and those that answered (typically 2 to 4) but those who came next.  For every question and answer session there are hundreds or thousands of readers of that session that came after looking for an answer to the same thing.  Thus if Stack Overflow allowed itself to be clogged with religious wars or question that could only be answered by opinion, then it wasted the time not only of those engaged in the futile discussion but everyone who wanted a genuine answer to the issue under discussion.  And this would only get worse and if it got so bad that most sessions were unproductive then people would stop using it as a resource and a nice thing would be broken.

So hence the not constructive rule (too opinion based).

They also stopped shopping basket questions (what are all the things that can do this? – the info gets out of date too quickly)

Also too localised – too specific and so not of interest to others (even tho it might be of interest to questioner)

..and it wasn’t for community/jokes

So to change perception of their rules they changed terminology to make it more aligned to the actual principle behind the exclusion

  • Closed -> On hold (it can be re-opened)
  • Too Localised -> off topic
  • Not constructive -> Primarily opinion based
  • Not a real question -> unclear what you are asking

The result of these changes saw an increase of questions that were edited after ‘closing’ to try to get them reconsidered from 5% to 10%.

changes  after closing

And the level of engagement with StackOverflow increased – those who post at least 5 times has increased over time ie Stack Overflow has introduced rules to protect itself which by definition will stop some contribution and will also almost inevitably have unintended/unfortunate consequences, but by putting effort into explaining their position, they have managed to take their core audience with them.

5 users


The conclusion seems to be – if you have something nice and enough people find it then you will attract those who want/result in breaking it.

In order to retain the benefit you have to remove things.  Either content of purpose. In order to remove things you must have rules.

Those rules are likely to be misunderstood by those who you value.  This can ruin your site as much as the barbarians you were trying to protect yourself from.

So have the rules and work on communications with those you care about to minimise the damage to/from them.







One thought on “Joel Spolsky on managing large sites + dealing with Idiots + non-citable dogs

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s