Premier League Power Rankings and Projections, Week 9: The ballad of bad wins

That was a bad win. Tottenham Hotspur were home to Hull City. We escaped with a win by the slimmest of margins, on the back a penalty that in many games you would not see awarded. People have pointed out that Hull City are a capable defensive team, and this is true. Steve Bruce has them well-drilled and Hull look like a club that could reasonably pull off a 14th place finish maybe. But that does not a good club make. Here's a comparison of shots on target and big chances for other top clubs in their matches against Hull:

Hull aren't bad, but when other good teams have played them, Hull have conceded significantly more and better scoring opportunities.

Now, I want to argue for some amount of perspective here. Spurs won their fourth 1-0, but this 1-0 was somewhat different from previous 1-0s. Spurs really didn't create much against Hull and the penalty was on the dubious side. Against Crystal Palace, Swansea City and Cardiff City, Spurs created a good number of chances and won much clearer penalties. So this 1-0 isn't exactly the continuation of a trend. The other 1-0s were better than this. Spurs averaged 3 SiBot (2 central, 1 wide) and 2.5 big chances in the other 1-0 wins. So in the other matches our problem was finishing the good chances we created. This match the problem was creating chances in the first place.

To me, the latter issue is much more worrying than the former. I see a match where Spurs put several shots on target from good positions inside the box and are credited with a couple of big chances, while stifling the opposition, and I feel good. I don't feel good about the numbers from the Hull City match. At least it was just one match.

Power Rankings and Projections

I found a small coding error in my projection engine this week. I had meant to phase out the subjective projections more quickly than the previous years' data as the season went along. I figured that I needed your subjective opinions of team quality at the beginning of the year, but I'd rather be using objective data as much as possible by the time we get to ten weeks. I had instead been phasing out the subjective side of the preseason projections equally as quickly as the objective side. So I fixed the error, which led to a couple of significant changes. (1) Manchester United take a hit. (2) Southampton jump up. (3) The returning clubs that have been really bad early (Sunderland and Fulham especially) take a hit. So for most of the numbers this week are normal, but a few might look a little weird—how did United lose 10 points of projection after a win, for instance—and so this is why.

Do remember that because of rounding, not all the numbers necessarily add up quite right.

• Expected Goals methodology
• Season Simulation methodology

• Fulham got the crap beaten out of them again. They put no shots on target from anywhere against Southampton, while allowing five shots on target inside the box to the Saints. I don't want this column to turn into MCofA's betting corner, especially because as you can see from the disclaimer above these numbers are very much still a work in progress. But I do like looking at the betting odds just as a check against my own calculations. The bookies have Fulham's relegation as a 4-to-1 bet, while Bloomberg (more on them later in the week) have Fulham just 15% to be relegated. I think they're in much more serious danger than that. This is a bad club playing badly who have taken a couple lucky wins.
• Southampton just keep winning and winning convincingly. They're edge a little closer to top four contention with every victory. There's a lot of talent there, in particular I think Morgan Schneiderlin should be in the conversation for best central midfielder in the Premier League this year. They have good depth over most of the pitch and aren't competing in any European tournaments that could run down their fitness over time. My belief is growing.
• Manchester City remain in good position after playing a good, close match with Chelsea at Stamford Bridge. Chelsea put four shots on target from inside the box to six for Manchester City. Chelsea's chances were better (three of four from central positions including one from very close range, compared to two of six from central positions for City), but the numbers project a very close game. City drop a little but stay top of the projected table so far.
• Watching Crystal Palace on Saturday morning, I think I'm ready to make the call. I'm not saying they played a poor game against Arsenal. Precisely the opposite! I think Crystal Palace had one of their best performances of the season, they were well organized and hard-working, they made Arsenal work harder to create chances than the Gunners have had to in recent weeks. And Palace still lost a home match by two goals. The sign of a bad team isn't always what they do when they're off, but instead what they do when they're on. Bad teams lose even when they play well. That's what happened with Palace this weekend. They're terrible.