Writers of the Future Contest 1Q29 Update: Statistical Torture

Back in December, I posted an article on the Writers of the Future Contest, and then provided an update earlier this month once contest participants started receiving rejections. As of today, I have both good news and bad news. The good news is that I estimate that the contest has already sent out between 44% and 53% of the contest’s form rejections. The bad news is that those who have survived this far (including me) still have an 80-83% chance of receiving a rejection notice before the contest is over. What follows is my analysis of the most recent Duotrope data.

Contest Status

Of the 410 Duotrope-reported submissions to the Writers of the Future Contest over the past year, only 1.95% of these submissions won the contest. About 90.5% reported rejections, and the remainder either reported their submission lost or returned, withdrew their submission, or failed to report anything at all. My analysis assumes that people do not report their results if they are non-winning finalists, semifinalists, or receive honorable mentions.

Source: Adapted from Duotrope

Based on the most recent data, I estimate that between 4% and 37% of the Writers of the Future Contest submissions are included in the Duotrope database. This estimate implies that total 1Q29 submissions range from 374 to 3,233. That said, it still likely fits into the more narrow range of 560 to 1,210 based on previous my prior analysis using historical contest data. My analysis is also based on the following Duotrope rejections over the past 30 days:

Source: Duotrope

So far, authors have reported 45 rejections on Duotrope, and there are currently 112 stories pending. That said, I estimate that only about 44 of these rejections are for the 1Q29 contest as I eliminated one entry that someone submitted 240 days ago.

Source: Adapted from Duotrope

Another problem with the data is that it is unclear which of the 112 stories pending are submissions for 4Q28, 1Q29, or 2Q29. However, based on the descriptive statistics that Duotrope provides for the pending population, we can rule out 4Q28. We can rule out this submission because the maximum number of days waiting in the data set is 130, implying a November 6, 2011 submission date — well after 4Q28’s September 30, 2011 deadline. Unfortunately, we can not rule out submissions for 2Q29 since the minimum days waiting of 1 implies a more recent March 14, 2012 submission date, making the submission ineligible for 1Q29. Therefore, some uncertainty remains around what percentage of the 112 pending participants submitted 1Q29 stories. The good news is that one can still do some analysis to estimate one’s probability of receiving a form rejection at this point.

The data set’s median value of 75 days implies that half of the 112 pending stories were submitted at or after 75 days. Unfortunately, since people tend to submit at the very last second, there are likely multiple 75-day submissions. This implies that while 75 days may be the average of the middle two numbers (since the data set has an even number of variables), there still may be a number of 75-day submissions above and below this median.

I simply assumed that 50% of the pending participants submitted before the 75-day deadline. The chart below shows how I arrive at my results of an 83% probability of rejection at this point in the contest, and assumes that the contest administrators have sent out 44% of rejections thus far.

Source: Adapted from Duotrope

Alternatively, I could estimate the 1Q29 pending population’s size using a Gaussian or normal distribution, and solve for the percentage of participants waiting longer than 75 days. This technique results in a slightly better 80% probability of rejection, and assumes that the contest administrators sent out 53% of rejections thus far.

Source: Adapted from Duotrope

Unfortunately, both techniques could likely underestimate the probability of receiving a rejection at this point, since many of the pending response participants are heavily skewed near the 75-day median. This would imply that over 50% of the pending response participants are waiting for 1Q29 results.

On the following table, I performed a sensitivity analysis that calculates the percentage of total rejections the contest administrators sent out thus far. It varies the percentage of the pending population waiting for 1Q29 results vs. the number of rejections reported thus far on Duotrope. The higher the percentage of pending participants in 1Q29, the lower the percentage of reported 1Q29 rejections and the higher the probability that remaining participants will receive a rejection notice.

Source: Adapted from Duotrope

The bottom line is that at worst, the contest has sent as little as 30% of its rejection notices thus far. However, it is more likely in the 40-55% range.


About Sean Patrick Hazlett

Conservative clean energy crusader, national security hawk, financial analyst, engineer, and former military officer.
This entry was posted in Science Fiction, Technology, Writing and tagged , , , , , , , . Bookmark the permalink.

2 Responses to Writers of the Future Contest 1Q29 Update: Statistical Torture

  1. rosieoliver says:

    I know this may sound over-simplistic, but why don’t you take the number of subs reported over the past year in duotrope and divide it by 4 to get the likely number of subs being reported on this database? The annual number of subs including what they call outliers for the year is 406. That makes is 101.5 subs per quarter. Let’s round it up to 102. And then take the obvous sums from there to give you a benchmark of how many rejections are likely in a quarter…

    • I think that is a perfectly legitimate approach. It also arrives at an answer eerily close to my estimate of 100.

      The only downside to that approach is that it does not account for variability from quarter to quarter. That said, all three of these approaches have their respective weaknesses. The right number is probably between 100 and 110.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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