Premier League Projections, Week 8

Another week, another post with the same top four. Liverpool played terribly but escaped with three points somehow from Loftus Road. Arsenal created chance after chance but their second-unit defense conceded a couple easy shots and Hull was clinical in earning three points. One would think that could open a window for someone to make a move on fourth place, but once again only Southampton took advantage. Spurs were beaten solidly by Manchester City while United continued its march of mediocrity with a draw at the Hawthorns.

At some point, you have to start taking Southampton seriously. I said when I first ran my projections that the Saints seemed too high, but they've done little since then but win football matches. The question for any smaller side is how well the club can handle injuries as the season progresses, and perhaps Southampton are headed for a fall. But thus far, the Saints' 51 shots attempted from the danger zone leads the entire Premier League, slightly ahead of Manchester City and Chelsea. They've been the Premier League's fourth best club, and only my continued use of 2013 data and estimated payroll data is keeping the Saints from a better position.

You might have noticed that Arsenal and Liverpool swapped places in that graph. The xG numbers for the two teams were so disparate that Arsenal actually improved in their underlying quality projection, and Liverpool lost so much ground, that the xG effects outweighed the real effects of Liverpool taking two more points from the weekend.

In the table at its current state, huge opportunities are available to any of Southampton, Spurs, United or Everton if they want to seize it. But in the league so far, only Southampton has played good enough football to take advantage. United are still running behind the points total an average team would have taken from their schedule, and Tottenham have been at best maddeningly inconsistent. Hopefully in the coming weeks I can start offering different commentary, but better football has to be played first.

Premier League Projections

• Full metholodogy, don't forget to click on "Methodology" for lots of math.
• EPL Advanced Statistics page, with many of the component stats for these projections.
• Because of rounding, the numbers may not all add up quite right.

• Last week I talked about how QPR's league-worst defense made them solid contenders for the terribleness title, but then they went and outplayed Liverpool solidly. So I'm holding off. Right now Hull City have by far the worst underlying statistics in the league, but it's hard to go all in on a club that keeps getting results. From watching them against Arsenal, I think that they're some finishing pixie dust covering the club, and it's probably not going to last. But they might have enough points anyway to survive.
• Newcastle y'all. They don't suck. Spurs project to beat them, but that doesn't mean it should be easy. The Magpies have a solidly above average attack (and a below average defense), and Spurs will likely need a couple goals to secure the points.
• Hey, you know what, I'll just put my Spurs-Newcastle projection here. Spurs go over 20% for top four with a win. That's not bad.