Week 1 is done, and from our point of view, that feels like a relief. We hope that we're at the end of our woes as far as server issues go, and the new infrastructure is in place to scale up our services as needed. We also discovered some DDOS attacks that had been made, so we're looking at ways to protect ourselves from people maliciously bombarding the site with queries simply to take us down. Seriously, why would anyone do that?

With that hopefully behind us, we can finally concentrate on content! As we promised last week, here's our look at the week 1 Statistical Picks, where we use various statistics as our guide for straight up and ATS picks.

Straight up:

Last year neither 'Best Defense' nor 'Best QB' performed particularly well in straight up picks, but so far, picking the team with the best defense from 2015 has looked pretty astute. Watching the likes of New Orleans, Indianapolis and Washington continue to give up big points showed that there's only so much roster turnover that can be done each season, so bad defenses won't get better overnight. From now on in, all statistical picks will use 2016 data (you know... now that there is some), so the next few weeks could well be bumpy as teams rebound or flatter to deceive.

The median score this week was 10-6, so for only two categories (Best Defense and Rushing Yards Allowed) to finish on or above that shows that there are certainly nuances to using stats in such a blanket way - especially offensive ones. Of particular note was the Patriots @ Cardinals matchup, which only the 'Best Kicker' category predicted correctly.

In straight up picks we also have the 'clutch' category, which filters out the games where there is a consensus heavily in favor of one team or the other. Of course, upsets are very possible (we'll keep an eye on these in a weekly upset-watch post), but clutch games are usually the ones that people have the most difficulty picking. In this case, there is one interesting development:

As you can see, picking the team with the best average passing yards per game is a better bet in close games than it is overall (8-8). We'll keep tabs on the clutch performance of all categories over the season.

Against the Spread

ATS is a different beast to straight up, and the hit rate needed to win vs Vegas is 52.4%, so anything above that means you're onto a good week. If it's substantially over that, we could be on to something...

Wow. To put this into perspective, a $10 on every game this week based on who threw for the most yards in 2015 would have returned a huge $49 (23.4%) profit ($209 returned from $160 stake based on standard odds). Before you go and bet the farm on such a thing next week, remember that this is only one week, and last year's reliable 'Best Defense' was less stellar, finishing at 43%. It may take a few weeks to work out which stats offer the best returns long-term, but the signs are encouraging so far.

Most interesting was the performance of ESPN's QBR rating. If we assume passing YPG and offensive PPG are factors in how well an offense and QB are performing as a whole, then QBR would seem to be a particularly out of step with those factors, out of step with 7 winning picks that chose the team with the best passing YPG.

As always, we like to make one thing clear about this data - it's not designed as a way of deciding which team is best, which QB or defense is better or any other type of subjective argument. It's specifically based on stats that you can use week in week out, providing they perform well. We'd advise giving these stats a few weeks to bed in, but equally, we all know that sometimes you have to go with the hot hand...