The Crime Contest: What Have I Gotten Myself Into?

With the first round of the Crime contest a little more than half over, I just want to give contestants and readers a little perspective on what this contest is all about.  This is the third contest of this nature I've been in on Gather, and I think I've finally figured out at least part of how they really work.

First, the theory: Gather holds these contests as a promotion.  The contests bring in new people to Gather, and those new people bring in their friends to vote in the contest.  Gather has more people which means more advertising revenue. 

The contest in theory sorts out the top entries, which are then sent to a committee which picks one or more to publish.  The public gets a voice in what actually gets published by voting for their favorite entries.  Gather sorts through the entries that didn't get enough votes to get into the top 15 and picks 10 to advance to the next round.  That keeps good entries from getting overlooked because the author doesn't have a lot of friends in Gather.

What's actually going on:

So how well does all of this work?  Well, the popular vote part of it doesn't work particularly well.  I'll explain why I think that later.  The Gather pick part seems to work better.  Last contest Gather did a reasonably good job with their picks.  They didn't get all of the good entries, but the entries they did pick were generally quite good.

What about this popular vote thing?  What's wrong with it?  The main problem with the popular vote is that there simply aren't enough reasonably impartial readers involved to make the good entries rise and the bad ones fall much.  It would probably take about 10 times the number of readers that are currently involved for the popular vote part of this to produce meaningful results.  That's not Gather's fault, and I'm sure they would dearly love to see additional readers descend on the contest.

Let's put some numbers to the problem.  According to Rand's statistics, 56 non-contestant readers have commented on 8 or more entries so far in the contest.   I figure that's a good approximation of the number of readers doing more than just reading entries from their friends.  Those 56 readers have made a total of 826 comments.  If those comments were spread evenly among the 268 contestants, that would be a little over 3 comments per contestant.

In addition 60 contestants have commented on 8 or more entries.  That's probably a pretty good approximation of the number of contestants making a serious effort in the contest.  Those contestants have made 1856 comments so far.  That averages out to a little under 7 comments per entry.  

What does that all mean?  Well, assuming that all of the contestants are really rating their fellow contestants fairly, that would mean that on average a good entry would be about 10 votes ahead of a bad entry at this point.  Unfortunately, I'm reasonably sure that since the contestants are human some of them aren't giving scores based totally on the quality of the entry.  Some are playing the "I'll give you a ten if you give me a ten" game.  Some may even be rating good entries low and bad entries high in an effort to knock out good competitors.  Let's say that about sixty percent of the contestants are giving honest ratings.  If you work out the math, that would mean that on average a good entry would be about seven votes ahead of a bad one at this point.   Assuming that the readers and contestants keep up at this rate, that difference may grow to fourteen votes by the end of this round. 

Now the votes probably aren't distributed evenly over the entries, so this may understate the impact of the unbiased votes somewhat.  I start to read a lot more entries than I end up rating or commenting on, probably three or four times as many.  If something doesn't appeal to me I'll simply move on to the next entry without rating or commenting.  If that's common, which I suspect it is, the difference average difference between good and a bad entry might be in the 20 to 25 vote range now, and that might eventually rise to 40 to 50 by the end of the contest.  That's probably a pretty good ballpark figure.

Unfortunately those seven or twenty, or even fifty votes can be totally overwhelmed by some non-quality factors:

1. A lot of people brought cheering sections, some of them fifty or even a couple hundred strong.
2. Some people have been on Gather quite a while and have dozens if not hundreds of friends willing to vote for them.
3. At least a couple of contestants will play the sympathy card effectively.
4. Some contestants belong to small, but tight-knit companies or organizations where people will vote for them en masse.  It doesn't take more than a few hundred people voting in that manner to overwhelm the small number of unbiased readers.
   
5. If your entry doesn't totally stink you can generate votes at a fairly predictable and high rate by using Gather's social networking.  In the First Chapters Romance contest I got over 230 votes in each round, mainly by using the social networking facilites.  That was good enough to put me unofficially in tenth place in both the prelims and the semi-finals.  I found that the number of votes I got was pretty much a function of how much time I was willing to spend soliciting votes.  I averaged between two and three votes per hour over what I would have normally gotten for each hour I spent soliciting votes. 
   

You may be a little skeptical about that last one, since I've only gotten about 60 votes so far this contest, but there are a couple of reasons I'm not playing the social networking card this time around.  First, I don't have time.  I've been tied up in rural Wisconsin with no Internet access about half the time during this contest, and that will get worse the last half of the contest.  Second, it doesn't make a whole lot of sense to spend as much time as it would take to win this thing that way.  I figure I would probably gotten around 80 votes last time around without the social network marketing.  I actually got 230+.  Let's say the social networking got me 150 votes each round of the contest, and I spent on average an hour for every 3 votes I got.  That's 100 hours over the course of the contest.  And it would have probably taken twice that many hours to actually make the finals.  Since only 3 people get to the finals based on popular vote, all it takes to make all of that social networking futile is 3 people with more time to spend on social networking than you can spend.

Now to be fair, the social networking business does have a quality component too.  If something stinks, it will take more effort to generate votes for it, and a higher percentage of those votes will not be tens.  If something is really excellent it will require less effort and generate a higher percentage of tens.   I think I had prettygood entry last contest, and that did play a role in how successful my social networking was.

 Bottom line: Unless a lot of new readers join in, good entries will tend to rise slightly and bad ones will tend to fall slightly on average.   That tendency will be overwhelmed by the number of votes that aren't based on quality.  It only takes three contestants with mediocre entries and a willingness to apply one of those non-quality factors, and you end up with no publishable entries chosen by popular vote.  If that happens, the publisher will choose one or more of the Gather picks.

So what do we as contestants do about that? 

1. Understand what the real value of the contest is.  Most of us (at least 266 out of 268) aren't going to come out of this as published writers.  We can come out of it with more comments on our novels than we would normally get in a year or two.  We can come out of it with a lot of exposure for our writing.  We can come out with contacts with other writers.  This is all free and very much worth it.
2. Don't obsess on winning.  A lot of people do.  I did both times I entered before.  I'm trying not to this time.  If you obsess, I can guarantee that by the end of the 3+ weeks you will be burned out and the contest will be a bad experience for you.
3. Vote honestly on as many contestants as you can.  I'm not the one to be guilt-tripping you on that because so far I've only been able to give to 20 to 25 entries a full review.  I have read at least a couple of paragraphs of around 150 more and picked out around 50 I want to download and read on my laptop while I'm trapped up in Wisconsin.  It really is up to us as to whether this contest is just a popularity contest or not.
4. If you know any avid readers, point them to the contest and turn them loose.
   

The vote honestly part of that sounds kind of scary.  What if you give someone a 10 and that's the difference between them getting into the next round and you getting into the next round?  When you think about it though there is only one situation where that would matter.  If you end up in 16th place and you gave number 15 a ten you could potentially mess yourself up.  On the other hand I'm guessing that if someone is in 16th place by one vote they are probably going to be one of the Gather picks unless they're pretty bad.  No guarantees of that, of course.

The ways that how you vote help or hurt your chances can be rather subtle.  For example: Let's say you come across a really good entry.  Well, if you give it lower score than it deserves that may actually backfire on you and hurt your chances.  If you don't get into the top 15 by number of votes, your only chance of making the semi-finals is through being a Gather pick.  Any good entry that isn't in the top 15 will be competing with you for one of those ten Gather pick slots.  If you aren't going to be in the top 15 yourself you want as many of the good entries as you can get in those 15 slots so you don't have to compete with them for the Gather 10.  Make sense?

Another tactic that can backfire badly: I'm pretty sure that at least a few contestants are systematically giving strong entries very low scores in an attempt to 'bury' them.  That could very well backfire in a couple of ways.  First, while the contestant may not know who gave them low scores, Gather can if they want to.  They have the information.  They have to track it in order to keep people from voting twice.  I'm just guessing here, but I suspect that Gather could easily pick out contestants that do that sort of thing and simply quietly disqualify them or not pick them as Gather picks.

 The other way this can backfire is more subtle.  If a good entry is buried down in the bottom third of the rankings and Gather finds it as a Gather pick, that entry is going to stand out in the minds of whoever Gather uses to make their picks.  That makes it more likely that the entry will also be a Gather pick in the semi-finals.  In the last contest the eventual winner was in the bottom third of the entries at the end of the first round, probably partly because people tried to bury it.  It was an outstanding entry and probably would have won anyway, but the fact that it was so low in the ratings probably gave it a boost in Gathers' eyes.

I hope this helps put the contest in perspective.  If any veterans of FC1 or FCR read this and want to chime in, feel free.  You may have some different perspectives. 

 I revised this aritcle on December 2nd to reflect some things about the contest I hadn't thought through earlier.  I also want to emphasize that nothing in this article is intended to minimize the hard work and accomplishments of the authors who eventually reach the semi-finals and the finals by having a good entry and marketing it honestly.

-------

 Just so you know about my potential biases:  I am a contestant in the Crime contest.  My entry is Char, a present day police procedural with a science fiction twist.