And on that note, it's a quite interesting to see that nearly 800 of you would rather commit suicide than see another comic posted here.

It would have been higher, if not for all the people who killed themselves anyway.

"if you want to make an omelet, you have to kill a few people". There you go, trying to be funny again. Survey Taker #2883 is going to be angry.

CSV file -- nice. As we've all learned here, flat files are the ultimate way to store data.

According to the data, there are 66(0.7%) people who never read the website (read freq = 0) yet still felt compelled to submit a survey. Of those 66, 3 people (4.5%) got on to say that MFD should be killed (gave it a 6).

Further, there appears to be 15 people who like to submit surveys without answering a single question.

I had to import the data in to Excel, then set that the first row was a date/time and then I could sucessfully import it in to Access. For kicks you understand.

Anyway, I now know that the 2 countries that hate MFD the most on average is Monaco and Iran. (Assuming ISO country codes).

Monaco:			7

Iran:			7

Netherlands Antilles:	6

Vietnam:		5.5

Venezuela:		5

Completely useless data

according to the data:

almost identical numbers of people like/dislike codesod and error'd - when grouping all the data together the breakdown of people choosing 1,2,3,4,5 has a correlation of 99.8997%.

However, when analysed on a per person basis, the correlation between the answers drops to 19.7919%.

So in plain English, codesod is just as popular as error'd, but not liked by the same people

Data as a SQL file:

http://ralph.sleigh.googlepages.com/wtf.sql

Chart showing that people who dont like MFD also voted please no more


TRWTF is that those are the default excel colours and you can only change them to other ugly presets, no nice gradients.

What I found interesting is that in Canada has 1/10 the population as the US but we submitted 1/7 the amount of comments as the US. Therefore a higher percentage of Canadians read this site. However, Australians must really like it because they have 1/13 the population of the US and submitted 1/8 the number of comments!

In the 'serious' department: I was curious if frequent readers preferred other sections than less frequent readers. Turns out that is not the case, the statistics are surprisingly even over all reader frequency groups.




And just for fun: if you are of the opinion that it is logical to have the same opinion on whether you like MFD and whether you want to remove it --- in that case people from Mexico, Spain, and Japan vote in the most consistent way, whereas people from Chile, Italy, and Croatia are less consistent :).



Note: People with a 0 vote in a relevant category have been removed from the data in the plots.

I wasn't privileged with the knowledge of thedaily's existence when this survey took place, but I figured I'd toss in my two cents on it anyway. I'm viewing the analysis of this survey from the perspective of traffic retention and increase, basically keeping the users you have and getting them to view the site more often.

It's interesting to note that the largest percentage of users who took this survey (and answered question 1) said they visited "once a day" (52.12%), while those who stated they visited "several" (19.03%) and "a few" (23.10%) times a day, made up nearly 10% less of respondents in comparison. That could either suggest the junkies who visit the site all the time just don't participate in the activities as much, or that your only providing enough content and entertainment for people to come around once a day.

Since I'm looking at this from a traffic perspective, I grouped the users into their types based on the 9124 that answered question 1, the types are below.

1. every now and then -- 136 -- 1.49%
2. every few weeks -- 97 -- 1.06%
3. once a week -- 292 -- 3.20%
4. few times a day -- 2108 -- 23.10%
5. once a day -- 4755 -- 52.12%
6. several times a day -- 1736 -- 19.03%

Digging deeper. In nearly all cases, every user type primarily voted in the same direction as everyone grouped together. For example if as a whole, the largest percent of users voted "Yes" for a theoretical question, then each user group separated primarily voted "Yes" as well. Which of course would make sense. However, this wasn't always the case.

For question 2. "Opinions on Feature Articles", the group as a whole voted primarily for "Love 'em". However, user types 1 and 2 ("every now and then" and "every few weeks respectively") leaned more towards the Feature Articles being on par with everything else with 42.65% and 42.27% of those user types voting for that option. So it would seem that the users who visit this site least often, are not particularly swayed by the feature articles. This happened on two other questions as well.

For question 6. "Tales from the interview", the group as a whole voted for "Pretty good, wouldn't mind more." However, user type 6 (those who visit several times a day) actually swayed more towards "Must have more!" with 51.32% of those user types voting in that manner.

And again for question 7. "off-topic Friday", the average user suggested they were fairly indifferent to these articles with 41.45% of them voting as such. But yet again, the user group 6 (several times a day) actually had the majority of its users voting this a step higher at "Keep 'em up".

These last two would both suggest that even though the largest subset of users you have visit once a day, those that visit several times a day are drawn in by the Tales from the Interview and Off-Topic Fridays more than the average user.

-----

Now then, on the not so pleasant side we have some more presumably irrelevant tidbits. The users who visit every few weeks participated the least in this survey with only 97 of them (or 1.06%) answering question 1. And of course remember that these users voted on par with all other groups with the exception of the Feature Articles where they rank it a step below the rest. Although on a brighter note, the "Feature Article" Feature ranked only 2nd to Tales from the interview as the most positively voted on feature of the site which suggests that those users who only visit every few weeks just don't fit the profile of the average user.

Back to the negative realm. When users voted down a particular feature they generally swayed towards a "Don't really care for it" or "skip it" mind set which means it's not really bothering them, but they wouldn't mind seeing it gone. The one exception to this is the "Erro'd" feature in which the largest percentage of negative comments suggest it doesn't belong on WTF. The largest user group backing this vote was the "Every Few Weeks" crowd at 24.74% of those users voting this way. With every user group across the bored primarily voting this feature as a "mostly indifferent" and well over 60% of all voting for indifferent or dislike, this feature could probably be axed.

By far, the most negatively ranked question was the "Vote for a new season of MFD" one. All groups voted primarily as "Could care less", however some 34% voted for "Vote of island", "please kill it", or "I'll kill myself". The largest subset of users backing this type of negative vote was "Every Few Weeks" user with nearly 37% percent. Individually, the largest single negative question vote for the "Vote for a new season of MFD" question was by the "Few times a day" crowd, who had a percentage of 17.55% vote for "I vote it off the island."

Taking all this into consideration, I would say it's pretty easy to see that "Feature Articles", "Code SOD" and "Tales from the Interview" are easily your most liked features of the site. Even the users who rarely come around voted them highly which means they're your meal ticket to more traffic.

No one is raving about "Error'd" and everyone but your Several Times a Day visitors don't really care about "Off Topic Fridays".

And of course, while oddly enough most people say they thought MFD was on par with the rest of the site, they basically all stood up to boycott having another season.

Thanks to both the time and the country being present in the data, with a bit of timezone conversion we can estimate what the local time was when each survey was filled out. It's not perfectly accurate due to multiple timezones in the US and other big countries, but it's not bad.

For example, if we correlate the time of day with the percentage of people who say they will kill themselves: