Last week, a different podcast spent a few minutes complaining that often Spurs get their first goal "too soon", thereby mussing up their chances of getting victories. Part of the reasoning is that the team lets up a bit, allowing the opponent to get back into the match. They snatch a goal, and then Spurs have to go back to work to salvage the three points.
Luckily, this is testable. So I've taken every PL match this year where Spurs scored a goal (we can name off the top of our head which matches don't meet this criteria) and noted when the first goal came, what the change was to the result at the time (did we pull ahead? Did we get an equaliser? Did nothing change since we were that far behind already?), the final result, and whether the first goal scored by Spurs "sealed" the tie—that is, we shut out the opposition, or it ended 1–1, with our goal's guaranteeing the draw.
I then also chose 10 PL matches at random from both Chelsea and Woolwich, our nearest competitors, where they score. I compiled the statistics for those matches as well, to give a kind of baseline for comparison. So when I mention, below, Woolwich and Chelsea statistics, remember that they are based on a random sample, and not on the population as a whole.
On another statistical note, please remember that a lot of the differences may seem significant, but could, actually, still be within the realm of random fluctuation.
First, Spurs' first goal comes, on average, in the 41st minute. This is interesting because after the first ten matches, we'd only scored in the first half three times. In the meantime, we've started scoring earlier, to bring that average down. Woolwich, on the other hand, score their first goal, on average, in the 35th minute. Chelsea in the 29th. Both clubs have also scored more goals than we (I think), which can account for that timeframe coming down a bit. So we don't seem like we're particularly eager to score early.
On the other hand, Spurs scored first in 20 of 28 matches where we do score. Assuming who scores is random, 71% is already well into the realm of statistical significance (a coin will come up heads 20 (or more) out of 18 times only 1.7% of the time). So we can say that Spurs are good at scoring first, even if it's not early. Both Woolwich and Chelsea are in the 70% range as well, but because each sample is smaller, the probabilities are higher (17% vs. 1.7%).
This makes perfect sense. These are elite teams. Elite teams get wins, and part of getting a win is scoring first. Spurs won 14 of 20 matches after scoring first (three draws and three losses), which is a 70% win rate, compared to 53% win rate over the season as a whole. Woolwich and Chelsea, on the other hand, won only 50% of the matches where they scored first. This suggests that Spurs are good at scoring first and subsequently grinding out the win, even if it's usually of the 2–1 variety (only four times has that first goal been enough—25% of victories).
So on the whole, nothing wrong seems to come from scoring first.
But scoring "first" is not the same as scoring "early". And Sunday is a good example of "typical Spurs" in the mind of some podcasters: we get a goal, but it's too early, and, subsequently, throw points away, settling for the draw instead of the win.
Let's split up the game into quintiles (18 minute chunks). Spurs have scored their first goal nine times in the first 18 minutes, twice in the second 18, five times in the third 18, and six times in both the fourth and fifth 18. It's interesting that lack of initial scoring in the second 18 (between the 18th and 36th minutes). Either we score early or late. That second chunk, though…
Anyway, what ends up being the result depending on when we score our first goal?
Looks like that other podcast is full of shit. When Spurs score in the first 18 minutes of a match, they win almost 80% of the time! So much for scoring "too early". How does it compare to the Woolwich and Chelsea samples?
As you'd expect, the likelihood of a loss or draw gets greater the later we get our first goal, though Woolwich and Chelsea are better at pulling together wins in the final quintile than we are. On the other hand, we're more or less better at winning when we score early. Anyway, the graphs speak for themselves.
What about when Spurs score first and early? The sample size actually doesn't change much. Only once have we scored in the first 18 when we were already trailing the game (Reading at home). For Chelsea and Woolwich, it's no times.
Again, then, there's no indication that scoring "too soon" has anything but a great effect on our league results. And the same is true for our competitors. Bring on the early goals, then. Panic when we're having trouble finding the net in the final 18 minutes…
As a final note, and an avenue for further investigation, in the Chelsea and Woolwich sample, 30% of the time that first goal "seals" the tie… as in either they shut out the other side or they get an equalised in a 1–1 draw. Spurs' have only "sealed" the tie with the first goal in six of 28 matches (21%). So our issue is that we're giving up more goals than our nearby competitors… not that we're scoring too soon. And that was the case against Everton. We let them score twice, which was a far greater crime than enjoying Adebayor's moment in the sun in the first minute!
Luckily, this is testable. So I've taken every PL match this year where Spurs scored a goal (we can name off the top of our head which matches don't meet this criteria) and noted when the first goal came, what the change was to the result at the time (did we pull ahead? Did we get an equaliser? Did nothing change since we were that far behind already?), the final result, and whether the first goal scored by Spurs "sealed" the tie—that is, we shut out the opposition, or it ended 1–1, with our goal's guaranteeing the draw.
I then also chose 10 PL matches at random from both Chelsea and Woolwich, our nearest competitors, where they score. I compiled the statistics for those matches as well, to give a kind of baseline for comparison. So when I mention, below, Woolwich and Chelsea statistics, remember that they are based on a random sample, and not on the population as a whole.
On another statistical note, please remember that a lot of the differences may seem significant, but could, actually, still be within the realm of random fluctuation.
First, Spurs' first goal comes, on average, in the 41st minute. This is interesting because after the first ten matches, we'd only scored in the first half three times. In the meantime, we've started scoring earlier, to bring that average down. Woolwich, on the other hand, score their first goal, on average, in the 35th minute. Chelsea in the 29th. Both clubs have also scored more goals than we (I think), which can account for that timeframe coming down a bit. So we don't seem like we're particularly eager to score early.
On the other hand, Spurs scored first in 20 of 28 matches where we do score. Assuming who scores is random, 71% is already well into the realm of statistical significance (a coin will come up heads 20 (or more) out of 18 times only 1.7% of the time). So we can say that Spurs are good at scoring first, even if it's not early. Both Woolwich and Chelsea are in the 70% range as well, but because each sample is smaller, the probabilities are higher (17% vs. 1.7%).
This makes perfect sense. These are elite teams. Elite teams get wins, and part of getting a win is scoring first. Spurs won 14 of 20 matches after scoring first (three draws and three losses), which is a 70% win rate, compared to 53% win rate over the season as a whole. Woolwich and Chelsea, on the other hand, won only 50% of the matches where they scored first. This suggests that Spurs are good at scoring first and subsequently grinding out the win, even if it's usually of the 2–1 variety (only four times has that first goal been enough—25% of victories).
So on the whole, nothing wrong seems to come from scoring first.
But scoring "first" is not the same as scoring "early". And Sunday is a good example of "typical Spurs" in the mind of some podcasters: we get a goal, but it's too early, and, subsequently, throw points away, settling for the draw instead of the win.
Let's split up the game into quintiles (18 minute chunks). Spurs have scored their first goal nine times in the first 18 minutes, twice in the second 18, five times in the third 18, and six times in both the fourth and fifth 18. It's interesting that lack of initial scoring in the second 18 (between the 18th and 36th minutes). Either we score early or late. That second chunk, though…
Anyway, what ends up being the result depending on when we score our first goal?
Looks like that other podcast is full of shit. When Spurs score in the first 18 minutes of a match, they win almost 80% of the time! So much for scoring "too early". How does it compare to the Woolwich and Chelsea samples?
As you'd expect, the likelihood of a loss or draw gets greater the later we get our first goal, though Woolwich and Chelsea are better at pulling together wins in the final quintile than we are. On the other hand, we're more or less better at winning when we score early. Anyway, the graphs speak for themselves.
What about when Spurs score first and early? The sample size actually doesn't change much. Only once have we scored in the first 18 when we were already trailing the game (Reading at home). For Chelsea and Woolwich, it's no times.
Again, then, there's no indication that scoring "too soon" has anything but a great effect on our league results. And the same is true for our competitors. Bring on the early goals, then. Panic when we're having trouble finding the net in the final 18 minutes…
As a final note, and an avenue for further investigation, in the Chelsea and Woolwich sample, 30% of the time that first goal "seals" the tie… as in either they shut out the other side or they get an equalised in a 1–1 draw. Spurs' have only "sealed" the tie with the first goal in six of 28 matches (21%). So our issue is that we're giving up more goals than our nearby competitors… not that we're scoring too soon. And that was the case against Everton. We let them score twice, which was a far greater crime than enjoying Adebayor's moment in the sun in the first minute!
