Tutorial: Rules and Their Consequences

Rules. We all have them, we all have to follow them. But how much do we think about our rules? Rules, by definition, are biased. If there wasn't someone who wanted to do something else, rules wouldn't be necessary. I'm not here to tell you how to write your rules. I figure it's your site, you should do what you like. Your rules should serve you, biased in favor of things you like. I'll give examples of rules and their possible consequences, and you can decide what rules you'd like to have.

Not that you couldn't do that without reading this tutorial... I'm just providing examples if you'd like a head start.

Occasionally I make suggestions in italics, which you are free to disagree with or ignore. That's why they're in italics. All dolls link to the page where they are displayed, where you'll find a convenient base credit link.

Simple Examples

Contest Judging
'Women Warriors': Why Rules Matter
"One Category per Doll is More Fair"... or is it?

Questions, Comments, Etc.

Terms of Use & Contest Rules

Very simple examples.

"You may only adopt my dolls with my permission."

Intended Result: You're aware of everyone who adopts one of your dolls, and you can make sure they credit you properly.

Possible Result: Many people won't go through the trouble of adopting your dolls, since there are so many great sites that only ask for a link back. Plus, since it takes effort to follow the rules, folks might be more likely to break them.

Comments: It's a trade-off between control and accessibility. It's up to you which you want to surrender.

"You must have a site to enter this contest."

Intended Result: Either 1. to ensure you get publicity when winners display their awards or 2. to guard against cheating. If the doll is on display somewhere, it's less likely to have been stolen from someone else.

Possible Result: Will probably lose you good entries, particularly from novice dollers. Dolls could still be stolen. Encourages folks to start sites before they have a clue what they're doing, thus inflicting bad sites on dollers everywhere.

Comments: Decide why you want to hold contests. Are they for you or your visitors? If you're more concerned about your visitors, why not waive the site requirement? However, if you've had security problems or have serious concerns about security problems, require a site and miss a few entries. Again, it's a trade-off.

Suggestion: If you're going to require site URLs, have the courtesy to actually link to them. Your contestants deserve the credit, at least once the voting is over.

~top~

Contest Judging

The way you decide to judge may be the most important decision you make about the contest. In fact, the judging system may determine the results. Don't believe me? Here's a (admittedly far-fetched) example.

Imagine I held a "Women Warriors" contest. I didn't really specify any more than that when taking entries. The dolls in this example are actually all by me or Sarah. Yes, I know the categories are contrived. The example is less confusing if the rankings within each category aren't arguable.

1. Arissa	2. Reyl	3. Merliss

Now, imagine I have three categories: Best Hair, Best Weapon, and Most Cute. I want to make sure even less-experienced dollers have a chance. I make a table, and rank the entries in each category, giving 3 points for first, 2 points for second, and 1 point for third.

Arissa definitely has the best shading. Merliss, being a sprite, has very little shading and only gets one point. Reyl falls in between, so she gets two points. On the other hand, it's hard to say if Arissa even has a weapon. That glass might count (she's a spy), so she gets one point. Reyl wins for three points, while the sprite, being a sprite, isn't as detailed and therefore only gets two points. However, little sprite Merliss is definitely the cutest, for three whole points. Compared to Reyl, Arissa is cute enough for two points.

Perhaps it will be clearer in a table. ;)

Hmm... it seems to be a tie. Why don't we compare the dolls in just one category and take the loser out of the running? That should make the results clearer. Imagine I pick weapons, because every woman warrior should have a weapon. Arissa has only one point in that category, so I remove the doll and the category from the list. We've already judged on weapons, after all.

Wait a minute... the sprite won? That hardly seems fair.

I have a confession to make. I specifically chose this elimination method so that the sprite would win. If the elimination round is about shading, Reyl wins. If it's cuteness, Arissa wins. If I'm paying attention, I can decide the winner and still pretend I did it within an objective system.

Sure, you don't judge contests like this. But the system you use is equally important, even if it isn't as obvious about it.

Every judging system has drawbacks. If voters use a point scale (say, 1 to 10), you have to contend with the fact that some people are stingy with high scores and some people are generous. Enough votes may balance out, but complicated systems attract fewer voters. If you judge by category (as I do), every judge may not understand the categories in the same way. If you run a Voters' Choice poll, people may cheat intentionally or accidentally. If you only accept votes by email, many qualified voters may abstain because they're nervous about giving their address out, or it's just too much trouble. Pick a system with drawbacks you can live with.

~top~

"One Category per Doll is More Fair"

An example using an actual system.

Intended Result: Each entry has a chance to win an award. The contest is happier and more fair.

Possible Result: The vote might get split for entries that are all-around good. The best doll could actually do badly, which raises questions about the validity of the awards. Do you really enjoy winning "Best Hair" when the winner for "Coolest Dress" has hair so much better it's obvious? I feel sort of guilty.

Comments: This one is getting popular. It's all very well, but it isn't fair. No contest is entirely fair, and we all have different preferences, but I grit my teeth when I read "to make it more fair..." when the system isn't any more fair than anything else, and can actually produce very biased results.

Imagine I ran a fantasy dolls contest, and received these entries:

1. Merliss	2. Valadine	3. Trinian

Next, imagine I asked judges to vote for three categories, only voting for each doll once. The categories are Best Shading, Best Hair, and Best Accents. Seven people voted, and I've shown the overall results in a table.

Considering that Trinian is clearly the best doll in each category, this is an odd state of affairs. The trouble was, the voters were divided over whether she should have "Best Hair" or "Best Shading." One voter, finding it impossible to choose between Merliss on Valadine on accents, selected Trinian because hers are obviously better. Merliss was clearly not a contender for "Best Hair," while Valadine's shading is noticably bad.

So, Merliss and Valadine go home with awards, but feel slightly guilty, and Trinian goes home empty-handed and probably feels cheated. If a doll really is the best, isn't it fair for it to do well? Furthermore, it's fair for lower-quality entries to do badly... it might not be nice, but it is certainly fair.

Suggestion: If your concern is giving folks a chance, and encouraging new dollers by giving them awards, why not think of awards that conveniently fit the entries? There isn't anything to stop you from making up all the awards you like, or just not listing all the extra awards won by the top dolls. Participants will be happier with awards they feel they earned, and they'll know you took time especially for them.

~top~

Questions? Comments? Additions? Complaints?

• appreciated/enjoyed/learned something from this tutorial.
• have questions or comments about this tutorial, or found something particularly confusing.
• find typos or formatting errors in this tutorial.
• can think of additional intentions or possible results for my examples.
• know a rule I just have to add as an example.
• have an objection you've thought about and can voice politely and respectfully.

• I'm a political science major. I think rules and their results are very interesting.
• I can be both sarcastic and silly.
• It is not my attention to criticize anyone as a person or a doller. You are free to use whatever rules you like, and I won't think any worse of you. I'm not trying to tell you what to do, I'm asking you to consider the possible results of what you do, if you haven't already. I've provided examples for folks who'd like a head start.

There are several ways you can contact me... pick whichever is easiest for you. Anyone who makes a substantial contribution will be credited. Thanks are due to Sarah for letting me use the retired version of her Reyl doll.

~top~

Individual Votes for Fair Contest Example

It's certainly true that I made sure these results were unfair. However, I found a plausible voting pattern rather than simply making up the numbers. This isn't an extreme case... it's entirely possible it could happen in a real contest. Not sure you want to take my word for it? Here's a rundown of the votes cast by all seven voters.

Voter #1 started at the beginning of the list and picked the doll with the best shading, Trinian. Then, from those remaining, Valadine had the best hair. Merliss got a vote for accents by default. This pattern makes a lot of since, since it starts from the top of the list, and two other voters did the same. (Voter #2 and Voter #3)

Voter #4 looked at all the entries and categories, knew she could vote for Trinian for only one, and decided that Trinian's hair was her most exceptional feature. (Trinian's shading is better than the others, but it isn't that exciting.) Next, she chose Merliss as the obvious remaining choice for shading, leaving Valadine points for accents because she was the only doll left. This strategic pattern made sense to two other voters. (Voter #5 and Voter #6)

Voter #7 couldn't decide if Valadine or Merliss had the best accents. Trinian's were obviously better, though, so she gave her the point in that category. Of those remaining, Valadine had the best hair and Merliss had the best shading.

There are many conceivable combinations of voters. This one is both possible and plausible, which is sufficient for my example. I have actually seen contests turn out with similarly odd results.

~top~