Stats and Projections for Norwich, Plus: Goals from Open Play PANIC

A wholly lackluster performance in the North London Derby gave way to an unhappy deadline day and nearly two weeks of a mostly blah international schedule. Even though Spurs have a very winnable fixture tomorrow, taking on Norwich City at home, there seems to be lot of angst in the Internet Spursfan Community. First, it's just sort of what we do. Second, despite six points from three matches, it's easy to identify problems with the results so far. I want to talk a little bit about one of the primary concerns that's been voiced here and across the internet.

No Goals from Open Play

Tottenham Hotspur have scored only two goals in 270 minutes of play, both on penalties. That would be an unimpressive record of goal scoring even if both were from perfectly constructed totaalvoetbal attacking moves. I get the concern. At the same time, most stat-based analysis of football runs on the basic assumption that goals are overrated. Not in the aesthetic sense, in the objective sense. If all you have is a 270-minute sample, your goals scored number will be too heavily influenced by random variation to be a strong estimate of team attacking quality. You would rather look at a variety of underlying stats. And in the underlying stats, Spurs are, well, ok. Not particularly strong, but, you know, ok.

I think maybe the best way to express what I mean by "ok" would be by writing up a big data table in html. Everyone loves html data tables. These are all the key stats I look at when analyzing the teams each week.

So on the one hand, just four shots on target inside the box is very poor. On the other hand, five big chances and two penalties drawn suggests that Spurs have been creating a respectable number of chances inside the box, but we just haven't been shooting well, missing the target on nearly every big chance so coded. Three of our big chances fell to Paulinho against Swansea, and he did not place a single one on the target. Jermain Defoe and Gylfi Sigurdsson also each scuffed an goal-scoring opportunity in the Palace match on opening weekend.

What do we make of this? Paulinho has shown good finishing ability for Brazil in international matches, and we all know Jermain Defoe can shoot. It's hard to say from a sample of five that Spurs lack finishing quality. One of the consistent complaints of Spurs fans watching the opening matches has been the lack of service for striker Roberto Soldado. The fact that we've created five big chances but none for Soldado does suggest that the Spurs attack has been running at a relatively low level of efficiency so far.

The other notable stat there is our eight shots on target from outside the box, second most in the league. Last season, Spurs were massive outliers in shots outside the box on target, with 109 on the season. Second most in the league were QPR's 76. We also had quite a few shots on target inside the box, but our 135 placed us well behind the best attacks in the league (MCI, MUN, LIV) and toward the back of the second pack (ARS, CHE, EVE).

As I have argued, not all shots on target are created equal. Shots from outside the box on target are converted at a rate nearly 1/3 that of shots on target inside the box (~35% for SiBoT, ~12% for SoBoT). An attack which ends up taking lots of shots from outside the box, even when they have the opportunity of an open lane and the shooting ability to put the ball on target, is producing far less valuable chances than an attack which is creating opportunities closer in.

My takeaway here is that the record is mixed. There are good signs and not so good signs. We'd rather be Manchester City or Arsenal, creating and converting lots of good chances from close range. But there are also strong teams with equally doubtful underlying stats, like Manchester United and Chelsea. And there's Everton, with good underlying stats but too few goals to show for it. After three weeks, you want to be wary of drawing conclusions from pretty much any statistics at all. Sampling issues are overwhelming.

There are enough positive signs in the data for Tottenham, I'd say, to conclude that it's way too early to panic about this attack.

Stats and Projections for Norwich

I haven't had a lot to say in the writers' prediction league things. It's mostly because I'm not creative enough to come up with a running bit. But it's also because the league right now hasn't really defined itself beyond some broad strokes. We'd got contenders, we've got mid-table, we've got just above relegation, and we've got a relegation race. Norwich are one of a large handful of clubs that currently project just above relegation. When Spurs play these teams, we should win, and there's not a ton else to say, unless you have something funny to say. And that takes work.

The top four race remains tight enough that Spurs should stay in the 50% range regardless of the outcome. They'll push a little toward 2-in-3 with a win. They should win.