2008 Presidential Website Race

You know when you have nothing better to do (or want to avoid the things you need to do) when you spend all your time watching the coverage (including the "Fair and Balanced" Fox News) of the U.S. Presidential race, especially when you don't even live in the country!

Being in the field that I am I thought I would look at the websites of the front-running candidates^[1]. I took a loose group of parameters that I usually look at when analyzing websites and measured each candidate's site against those parameters. Let's see how they measure up (or you can just jump to the summary table):

The Analysis

Parameter 1: DOCTYPE used.
Declaration of the Document type is important for web browsers to correctly interpret the code that follows it. While I have a preference for using XHTML over HTML (more for well-formedness than any other reason), it's more important whether the adherence to standards mode is Strict or Transitional. The Results:

• Obama: XHTML Transitional
• Clinton: HTML4 Transitional
• Edwards: HTML4 Transitional
• Richardson: XHTML Transitional
• McCain: None
• Romney: XHTML Transitional
• Huckabee: XHTML Transitional
• Giuliani: XHTML Strict

Winner: Rudy Giuliani. Looser: John McCain

Parameter #2: (X)HTML Validation
There are very few websites that validate 100%, 100% of the time. But there isn't really any excuse. If you have more than one or two errors that have crept in due to CMS implementation then that can be acceptable. However, error-ridden code is just lazy and reflects poorly on the webmaster/developer/agency - so lets see how the candidate's websites perform:

• Obama: 182 errors. Despite nice clean code, there are a small number of critical errors that have a cascading effect creating other validation issues. Had one or two mis-step been cleaned-up this number would be in single digits.
• Clinton: 12 errors. A few un-encoded ampersands and open tags prevent a perfect score. The winner in this contest.
• Edwards: 21 errors. Some obvious WYSIWYG/CMS errors have made there way - but pretty good result.
• Richardson: 18 errors. Again some simple oversights prevent a perfect score - no excuse considering how simple they are.
• McCain: 79 errors. Failing on a number of critical issues, i'm surprised the number isn't higher. The only website without a DOCTYPE and the only one not using semantic mark-up. This site uses 5-year-old construction methods.
• Romney: 131 errors. I'm pretty confident that no human eyes have seen the code in this CMS mess. You only have to see the source code to know what I mean.
• Huckabee: 70 errors. Entirely CMS and ColdFusion related errors here. Accepted but not Acceptable.
• Giuliani: 16 errors. The fact that this is the only site that has a Strict DOCTYPE, I am not surprised that they have some errors - but the fact that the number is so low really is impressive. Easy to fix errors though.

Winner: Clinton. Looser: Obama. Honorable Mention: Guiliani.

Parameter #3: Use of inline CSS and/or Javascript.
Any website worth it's salt uses external CSS and Javascript files to take advantage of browser caching, thus increasing website speed performance. Additionally, the use of inline CSS and Javascript in the body of the HTML is hard to justify. I have used a school-type grading system where 'F' is fail, 'C' is pass and 'A' is outstanding (with the intermediate grades in-between) Here's how the candidate's sites handle this parameter:

• Obama: B - Almost all CSS/JS has been relegated to external files with only minor transgressions.
• Clinton: C - A good start with the majority of CSS/JS in external files then a fair smattering of inline code appears.
• Edwards: C - Bordering on a D, with <style> tags inside the body of the document.
• Richardson: B - minor use of inline CSS
• McCain: F - Yuck, ick, eewww. Yes, some JS files are called externally, and yes it does't use the dreaded <FONT> tag - but thats the only good things I can say...
• Romney: B - But bordering on C. Lots of inline CSS.
• Huckabee: D - Too much inline CSS and JS to pass this test.
• Giuliani: A - Close to perfect - a minor transgression only.

Winner: Giuliani. Looser: McCain.

Parameter #4: Use of a Splash page?
Splash pages or intro pages are universally recognised as bad design decision, often seen as a major impediment to the end user obtaining the information that they want. If a website here uses a splash page - it's not considered the main page that all the rest of the results are derived from. The results:

• Obama: Yes - but does use cookies to determine if the visitor has viewed the site before and skips the splash page if they have.
• Clinton: No
• Edwards: Yes - but uses cookie based skipping.
• Richardson: Yes, but uses cookie based skipping
• McCain: No
• Romney: Yes
• Huckabee: No
• Giuliani: No

Winner: 4-way-tie. Looser: Romney. Worthy Mention: Richardson for the worst (or best depending on which way you look at it) photoshoppery.

Parameter #5: Tidy URL's?
Essential for bookmarking, navigation, SEO and server performance, Clean URL's (webpage addresses) are pretty much standard practice nowadays. The Results:

• Obama: Yes
• Clinton: Yes
• Edwards: Yes
• Richardson: Mostly
• McCain: Mostly
• Romney: Yes
• Huckabee: No
• Giuliani: Yes

Winner: Too close to call. Looser: Huckabee.

Parameter #6: PageRank.
Mostly considered irrelevant except in terms of prestige, Google PageRank is based on a mathematical formula that measures a number of metrics to determine where on a logarithmic scale the site is ranked compared all other websites. The results:

• Obama: 6
• Clinton: 6
• Edwards: 6
• Richardson: 6
• McCain: 6
• Romney: 5
• Huckabee: 5
• Giuliani: 6

Winner: Too close to call. Loosers: Romney and Huckabee.

Parameter #8: Alexa Rank.
Alexa estimates and ranks websites on traffic volume, extrapolating the figures from users who have the Alexa tool-bar installed. Lower the number the higher the score. Results at time of posting:

• Obama: 17303
• Clinton: 32729
• Edwards: 69182
• Richardson: 213127
• McCain: 99639
• Romney: 63734
• Huckabee: 38119
• Giuliani: 92963

Winner: Obama. Looser: Richardson.

Parameter #9: Uses semantic markup.
Semantic markup is best described as the use of the appropriate tag for the appropriate content. In otherwords - using the heading tag (h1) for the main heading, paragraph tags (p) for paragraphs etc. The use of tables for anything other than tabular data is frowned upon. All candidates other than one use semantic markup (in various degrees), so here I will only name the looser:
Looser: McCain.

Parameter #10: Number of HTTP requests.
Based on the results from WebSiteOptimization.com, each site was test to get the total number of requests made from the browser to the server to retrieve information required to display the website. The lower the number the better - and less than 20 is considered optimal.

• Obama: 96
• Clinton: 247
• Edwards: 33
• Richardson: 47
• McCain: 117
• Romney: 68
• Huckabee: 20
• Giuliani: 61

Winner: Huckabee. Looser: Clinton.

Parameter #11: Page Size.
Relating to the index or main page only (not including splash pages) the same service as above measures the amount of data in terms of bytes that the files requested amount to to create the web-page. The rule of thumb is for your homepage to come in under 100Kb. The results are:

• Obama: 1174Kb
• Clinton: 978Kb
• Edwards: 154Kb
• Richardson: 215Kb
• McCain: 446Kb
• Romney: 914Kb
• Huckabee: 95Kb
• Giuliani: 521Kb

Winner: Huckabee. Looser: Obama.

There you go - what a pointless race. You might notice or ask why I didn't consider design. Well - i'm not a designer per-sé - So I'll leave that to the right-brained amongst us.

The Winner

Based on the above (giving points for placement and adding points for failure) the winner with the least amount of points garnered is Mike Huckabee Ron Paul!^[2], closely followed by John Edwards and Rudy Giuliani. The clear looser was John McCain (call me John!). Let me know your thoughts.

Inspiration for this post: 2008 Elections by Server