Agree with Wayne, but for reference I'll have a large, good quality image at around 70kb and a standard image at around 20 - 30kb. If I can get it in for less without it looking terrible I will.

If you have photoshop it shouldn't be much of a problem playing with the save for web setting and seeing how much you can trim off. 60% is a good standard for jpg files.

So you use robots.txt to disallow indexing of anything under the /bet/ folder, you link to 'bet/XYZ' using nofollow and 'bet/XYZ' has a redirect on it?

I'm going to go with safe. It's a fairly common practice.

You want to track individual pdfs?

Off the top of my head adding ="javascript: pageTracker._trackPageview('/spoofpage')>anchor text will spoof a page view to your spoofpage in analytics.

That will require individually tagging each pdf though.

Let me have a think if there's anyway to change the 'spoofpage' to the href 'url' to allow you to add the same code to each.

1. Doesn't matter too much.

2. Yes - http://www.youtube.com/watch?v=GIn5qJKU8VM

3. Yes.

Things to maybe think about though is HTML 5 and the

<header>tag for sections. Read - http://html5doctor.com/the-header-element/ - and - http://www.alistapart.com/articles/previewofhtml5

At the moment there will be some accessibility issues with screen readers when using multiple

s but that should get better.

Headers aren't a big ranking factor - http://www.seomoz.org/ugc/the-five-least-important-ranking-factors - http://www.seomoz.org/article/search-ranking-factors#metrics-6

A vast majority of the web isn't coded correctly and the above points to it not really mattering, however, it's worth thinking about your structure and still only really using one

if you can.

</header>

robots.txt isn't a requirement, indeed it's only voluntarily followed by spiders (as in they can choose to ignore it), so I think you'll be fine without it. The default is to 'allow all' and 'follow, index', so they should still be crawling the site correctly.

Check in Webmaster tools by fetching as Googlebot or alternative find a page and put cache:pageurl.html into google and see if it's cached it correctly.

That said returning a 500 instead of a 404 may be causing an issue that isn't obviously apparent and 500 is a bit too generic a message to say specifically what, but I would try and solve it as quick as possible. The benefits will depends on what you put in your robots.txt file

Tildes are okay these days, but 'unsafe'.

http://www.cs.tut.fi/~jkorpela/rfc/2396/full.html#2.3

Tilde was (begrudgingly) added to the unreserved character list a while ago, so Google should treat them fine without encoding.

However, if you can avoid using them I would, so leave the old addresses but from now on I'd use a hyphen (in preference to an underscore, still) instead of a tilde if you can.