New TweetPsych Feature: Top Lists

Now that there are over 80,000 users indexed by TweetPsych, I’ve added a new feature that ranks users by specific characteristics. For the select traits listed below, you can see the 20 users who scored the highest.

Please remember that this is for entertainment purposes only and that the codes are linguistic terms from LIWC and RID that may not be similar to their normal, English language meanings.

Read the original post for more information about TweetPsych, or read this post for more information about its matching functionality.

Read More

The URL Shorteners that will get you the Most (or Least) ReTweets

We know that most ReTweets contain a link, but there are hundreds of different URL shortening services available to help you save space with that link. I analyzed my database of over 30 million ReTweets and compared them to over 2 million random Tweets to find which shorteners are the most (and least) ReTweetable.

I calculated how much more or less often each URL shortening service appeared in ReTweets than it did in normal Tweets and presented this value as a percentage. For instance, in my data 9.28% more ReTweets than random Tweets used I took into account the fact that ReTweets tend to contain more links than average Tweets and normalized the occurrence values.

The short, post-Twitter shorteners,,, and were all more ReTweetable than the older, longer, tinyurl.

Read More

The 20 Least ReTweetable Words

I’ve looked at the 20 words and phrases that tend to get the most ReTweets, but what about the flip-side of that coin? What about the words that are least likely to get your ReTweets?

I used my database of over 30 million ReTweets, and compared it to a sample of over 2 million random Tweets and found the common words that occurred far more often in non-ReTweets. The percentages below represent the relative Un-ReTweetability of the 20 least ReTweetable words.

Some Highlights from the List

There are a number of “-ing” verbs, including “going,” “watching” and “listening,”  which reinforces my understanding that answers to the “What are you doing?” question don’t get very many ReTweets.

The presence of “sleep,” “bed,” “night,” and “tired” indicate that people often Tweet “goodnight” style messages, but generally don’t ReTweet them.

The relatively informal nature of many of the words on the list including “lol,” “gonna,” and “hey,” show that simple or slang conversation is not ReTweetable.

The lesson learned here is that if you’re trying to get more ReTweets, don’t just engage in idle chit-chat or Tweet about mundane activities.

Occurrences Word or Phrase
0.88% game
0.84% going
0.72% haha
0.71% lol
0.7% but
0.66% watching
0.63% work
0.6% home
0.48% night
0.44% bed
0.43% well
0.38% sleep
0.38% gonna
0.37% hey
0.32% tomorrow
0.3% tired
0.29% some
0.29% back
0.29% bored
0.28% listening
Read More

The Punctuation of ReTweets

140 characters doesn’t leave much room for extraneous letters, numbers or symbols, so you might think that punctuation would be sparse in Tweets. But I compared a random sample of over 1 million “normal” Tweets to a sample of over 10 million ReTweets and found that 85.86% of Tweets contain some form of punctuation, and an overwhelming 97.55% of ReTweets do as well.

Of course, the prevailing ReTweet format includes a colon to better display the original Tweet, but even when ignoring this form of punctuation, ReTweets still contain more punctuation than non-ReTweets (93.42% to 83.78%).

I then analyzed the frequency of specific types of punctuation and found that hyphens, periods and colons are the most ReTweetable punctuation, occurring far more commonly in ReTweets than in regular Tweets, while the rarest mark, the semicolon, is the only unReTweetable punctuation mark.

Read More

The Science of ReTweets: Updated Presentation

I study ReTweets because I believe they offer an unprecedented window into how people spread ideas. And while Twitter may be threatening to mangle them, I think they’re still the most important innovation to come from microblogging yet.

I gave this presentation, or a version of it, at a few conferences this summer and since then I’ve done a bunch more analysis. So I added all my new data to the slideshow, included a video interview with me after Social Media Camp and uploaded it to SlideShare for your viewing pleasure.

If you like this presentation, vote for it on SlideShare and vote for my panel at SXSW.

Read More

Twitter Plans to Mangle ReTweets #SaveReTweets

If you’ve read this blog, you know that ReTweets are one of my favorite topics. For a ton of reasons I think that they’re not only one of the most important developments to come from Twitter, but from social media in general.

How ReTweets Work Now

As you probably know, ReTweets were designed by the community, for the community, and currently look like this:

RT: @username Really Awesome Tweet

Granted, the “RT @username” prefix takes up some space, but that minor annoyance is more than made up for by the benefit users get from a Tweet clearly labeled as being ReTweeted from @username originally. When you see a ReTweet in your timeline it has the avatar of the person who did the ReTweeting, so you know who spread it to you and from whom they got it.

ReTweeters could add their own commentary (and lend social proof with their name and avatar), Twitter client developers could add one-click ReTweet functions and analysts (like me and Microsoft) could gather ReTweets and study them.

How Twitter Aims to Break ReTweets

In a stunningly disappointing move, Twitter has threatened to completely eviscerate most of the value out of ReTweets by “formalizing” a feeble version of a format that was already well understood and functional for all users involved.

Twitter plans to add a button to the Twitter web client that says “Retweet” that will allow you to send the same exact Tweet, with no editing, to your followers. Your followers will see the original poster’s avatar and name, even if they’re not following them, and the only indication they’ll see that it is a ReTweet will be a small line of light gray text underneath it.

I follow people because I trust and enjoy their point of view, I don’t nessecarily trust the POV of people I don’t follow, so using the original poster’s picture and name in my timeline destroys any social proof the ReTweeter may have lent the Tweet.

Most active Twitter users use third party desktop and mobile clients to Tweet, and there is no way of telling how those developers will indicate ReTweets in this new format just yet. The Tweets will not contain the “RT @username” prefix. There will no longer be a commonly understood format. Scanning my friend’s timeline is how I use Twitter, and I suspect how many of you do too. The new ReTweet format will make that much harder.

If more than one of my followers ReTweet the same Tweet, the screenshots seem to indicate that the ReTweet won’t appear more than once in my timline, it will simply be updated to say “ReTweeted by @user1 and @user2…” The problem here is that if @user1 ReTweets at 1pm and @user2 does it at 2pm, that Tweet will have been buried in my timeline and I won’t see it again.

The new version of ReTweets will come with a few new API calls. They’re calling your friend’s timeline by a new name so they can deprecate the old one (which worked fine). They’re going to allow you to see ReTweets you’ve posted (not sure why), and ReTweets your followers have posted (which you could already do). The only kind of cool API call is the one that will allow you to see the 20 most recent updates that are ReTweets of your Tweets; problem is, you can only get yours. You can’t see the most recent ReTweets of other people’s content, and you can’t check for ReTweets of a specific Tweet.

If I didn’t know better, this would make me think the team who designed this didn’t really understand how ReTweeting works. (Update: it turns out the project lead, @Zhanna, has only retweeted and using a non-standard version of the less popular “via” syntax) The new format will make them harder to use, more confusing, less valuable and kill the ReTweet.

How ReTweets Should be Adopted

The idea of a button, next to the reply button, is great; that absolutely should be implemented. But clicking that button should do the same thing that TweetDeck does: copy the Tweet into the text area, add “RT @username” and let me edit before sending.

An API call should be added so that 3rd party clients could signal to Twitter that a Tweet is a ReTweet of a specific update. The new API calls are otherwise fine, but there should also be a call to get all ReTweets of a specific Tweet.

My advice? Use the HashTag: #SaveReTweets to start making some noise about this, and keep using the old “RT @username” format.

Read More

Gender Differences in ReTweeting

Not all viral content sharing on Twitter happens in ReTweets, so when I designed my viral Tweeting survey, I included three similar questions:

1) What types of content do you ReTweet?
2) What types of content do you Tweet about?
3) What types of content do you Tweet links to?

They’re all pretty similar, but there are obvious differences in each. For instance, people are OK Tweeting about their own opinions, but are unlikely to Tweet links to or ReTweet other peoples’ opinions.

But here’s where the data gets really interesting, the graph below shows the differences in answers by the gender of the respondent. 10.5% more women than men say they ReTweet entertainment content, while 32.2% more men Tweet about their opinions.

Read More

Win a Free Server for Life in the Name of Science

>Being a mad scientist means that you don’t just accept common wisdom; you challenge it in favor of finding actual proven facts. Hard data and experimentation are the only way to the truth and you know it.

When people tell you that you can’t measure something, or that human behavior is not predictable, your first inclination is to laugh at them. You know that the course of human progress has been driven by people who push the limits of science. To put it simply, guessing didn’t get us here.

I know there are lots of us out there and I want to start bringing us together–and not just social and viral mad scientists like me, but anyone who pushes the edges with their work.

To help you tell the world that you’re a mad scientist, and proud of it, I commissioned the incredible artist Pat Perry to draw a sweet cube grenade (hat tip to Hugh Macleod for the idea).

Science doesn’t happen in a vacuum, so I’m also launching a community Page on Facebook and a newsletter to start to build an army of mad scientists.

To sweeten the deal, I’m kicking it all off with a contest. To enter, simply print out the mad scientist cartoon and proudly display it in your “lab.” You know, wherever you do your mad scientist work– an office, cubicle, studio, nuclear power plant or whatever. Take a picture and post it to the wall on the Mad Scientists Facebook Page.

The scientist with the craziest workspace will win a lifetime of free hosting on a MediaTemple dedicated virtual server. Yeah, you read that right, a free server for the rest of your life. And these servers are seriously awesome; it’s what I host all my crazy apps on. The second place winner will get a free 4gb Mimobot USB drive of your choice.

I’ll announce the winners in the newsletter, so make sure you’re signed up for that as well.

Read More