Obtaining An Obscure Soccer Stat

[Trinidad]

Does anyone know how or if it would be possible to find out the average time into a match that the first goal is scored for the different leagues?

i.e, on average, how far into an EPL match is the first goal usually scored?

[RickySteve]

It may be somewhere, but I've never seen it and I've studied a lot of soccer sites. The information should be relatively easy to compile yourself however.

[Wheell]

Ricky's right, you can gather a ton of data from a variety of soccer sites (and, surprisingly, Wiki). You'll realize very fast that while more goals are scored in the second half and that in each half more goals are scored in each minute than the last (ie, more goals in the 38th minute than the 9th). There are a few more things to pay attention to, but you'll figure it out.

[adriano]

At the bottom there is the goal time distribution by team.

http://stats.football365.com/dom/ENG/teams/ManchesterUtd.html

[69882000]

Soccerway is pretty decent.

[Data]

Does anyone know how or if it would be possible to find out the average time into a match that the first goal is scored for the different leagues?

i.e, on average, how far into an EPL match is the first goal usually scored?

A good approximation would be (90+t)/A, where t is the average injury time and A is the league's goal average. Likely, you are actually looking for the median, not the mean. The median is 0.693*(90+t)/A.

[Data]

You'll realize very fast that while more goals are scored in the second half and that in each half more goals are scored in each minute than the last (ie, more goals in the 38th minute than the 9th).

This may give a wrong idea of causation. The goal rate is slowly decreasing over time with each goal causing a shift up in the rate curve.

[Wheell]

Data: I'm sorry, but I think we have a failure to communicate.

There are a couple of things you might be saying:

1. Goals beget goals. IE, if there were 4 goals in the first 80 minutes that predicts higher scoring relative3 to the O/U for the game than a nil-nil game with identical inputs in the 80th minute.

2. You are talking about the likelihood of the FIRST goal being scored in any particular minute.


Now, if you are saying the latter I can't really argue. If you are saying the former I'll meet you halfway. Yes, goals do slightly predict goals, but as for the 80th minute being lower scoring than the 50th minute (again, with other inputs held constant), that is just wrong. I've gone over thousands of games and the data seems VERY clear on this (no pun intended).

[turnip]

Based on my data for EPL 07-08 (may not be 100% accurate, but is more than 99% accurate):

-26 of the 380 games were scoreless.

-Of the remaining 354, the median time of the first score was the 27th minute.

Here is the histogram: (It should read "31-45" instead of "30-45")

[Data]

2. You are talking about the likelihood of the FIRST goal being scored in any particular minute.

Yes, sir. I agree with you that in general the goal rate increases over time.

[Wheell]

Data: Ah, good, I didn't want to have each of us pull out their databases to see which one was bigger.

[vanman]

try optasportsdata.com

[Justin7]

I read a study on time of first score (I was looking at World Cup scoring props), but I don't remember the cite.

There were two things to consider though...
First, about 60% of all goals are scored in the second half.
Second, the scoring distribution is NOT a nice Poisson - the odds of a goal per minute go up after the first goal.