I recently came across this informative and useful .pdf article by Jim McMillan regarding the many nuances of autoguiding. When the general literature gives the impression that it is a straightforward and relatively easy process it is reassuring to read Jim's article revealing that it is far from being so as I have spent the last year discovering. With the feedback I'm getting from Metaguide my last few sessions have clarified just how much the seeing affects autoguiding performance. I usually try and start imaging as soon as it gets dark but in reality this is often when the seeing is quite poor and it is only after a few hours that it might improve. Although an accuracy of around 2 arcseconds does seem possible under good conditions it is often 6 arcseconds or worse on some evenings.