The purpose of this blog is to try and assess the success, or otherwise, of the rankings given by AutoManning.

I’ll get more onto that in later posts, so, first up, a few first thoughts and some useful info for getting the most out of the site.

The first thing to note is that this is very much a work in progress and, as such, far from perfect. I kept making numerous tweaks to the formula and it would improve results in some games and make others worse. So I figured it important to get something up and usable in the interim and I can improve on it over time.

And, with that caveat in mind, the formula is very much usable. I’ve done some very cursory scans over the first few weeks and the better games, by and large, undoubtedly rise to the top of the rankings.

So far it seems as though the number one game is always (without exception?) a good game: a good game simply being defined as one that is worth watching, isn’t all over by midway through the third quarter or worse by halftime.

Interestingly, some games that appear, on paper to be very similar (for example two games that finish with a similar number of points and similar points differential) might be ranked ten, twenty or even thirty points apart. This is obviously because the games also consider likelihood, comebacks, and evenness. Nonetheless, its sometimes not to easy to tell which of these games is actually better – even when looking at the post game reports.

As such, the number one ranked game is definitely not always the best game that weekend, but the top two or three almost always is. It’s therefore useful to use the full rankings over the number one pick and use your discretion. If you see Bears @ Packers with 230 points and Browns @ Raiders with 200, your probably better off going with the lower ranked game if you happen to prefer those teams, because there’s unlikely to be huge differences between them: both should be good, which is, of course, the most important thing.

One glaring anomaly has jumped out from a quick browse of the results: the game between the Chiefs and the Rams, which was not only regarded as the best game of the season, but possibly the best game ever, ranks as only 6th best in Week 11. This is definitely a bit of an outlier but far from ideal nonetheless!

When you look at how the game’s are ranked it’s kind of understandable that it didn’t rank as high as you’d think. For, even though it was a massively high scoring game, it was devalued by unusual timings of scoring (thereby not appearing as close as it was) and the fact that the team winning for most of the game, won in the end (which effectively ranks it at 0 for comebacks). Finally, the team everyone confidently thought would win before kick off, did win, and so it was poorly ranked on that metric too.

So it’s both understandable but not desirable. I have been thinking about a solution for sometime, and have a few ideas, but I need to go back to the drawing board in earnest and this will take some time.

So, for now, a very workable, if imperfect AutoManning. Enjoy.