Wednesday, September 22, 2010

Goal Trends

We’re now 5 weeks into the new season and while it’s perhaps a touch hasty to reshape our team based on the early trends, it is never too early to look for indicators that might show where future value may lie.

I have been tracking the way goals have been scored and who has scored them and some interesting trends have emerged that might be worth considering when deciding who to captain and which defenders to play each week. The data will be less useful for transfer decisions, but even then it can’t hurt to know a bit more about the player you are getting’s prospects over the coming weeks.

Scoring/conceding analysis by position
I have highlighted some of the strongest trends seen for certain positions scoring goals and for positions that tend to score against certain defences. For example, if Everton scored 5 goals and midfielders scored 4 of them, they would show up as Midfield 80% below. Equally, if Arsenal conceded 10 goals and forwards scored 5 of them, they would show up as Forward 50% in the defensive trends below.

Scoring trends
Everton – Midfield 100%
West Ham – Midfield 100%
Bolton – Forward 86%
Wigan – Defender 50%
Birmingham – Defender 43%

Defensive trends
Arsenal – Forward 100%
Blackburn – Midfield 100%
Chelsea – Midfield 100%
Man City – Forward 100%
Man Utd – Midfield 86%
Blackpool – Midfield 75%
Sunderland – Defenders 60%
Fulham – Own Goals 33%

These rankings can be used in conjunction with the clean sheet rankings to get a better idea of your players’ chance to score, given the position he plays. For example, Everton at home to Blackpool looks like a dynamite fixture for, say, Yakubu, but given that an Everton forward has yet to score this season and Blackpool are generally conceding to midfielders you might want to look elsewhere.

Equally, we might downgrade Sunderland against a team like Birmingham or Wigan (whose defenders have accounted for 43% and 50% of their team’s goals) considering Sunderland have conceded an incredible 60% of their goals to defenders.

Finally, the trends provide some assistance with longer term transfer decisions as you may wish to reconsider that Everton forward or that Bolton midfielder.

One word of caution is the fact that you will notice that the better sides in the division dominate the defensive rankings. This is because you are looking at a percentage of a much lower number and so while Chelsea have conceded all their goals to midfielders, this is actually just a single goal from Scott Parker so you wouldn’t therefore upgrade your midfielders when they take the daunting trip to Stamford Bridge.

Scoring/conceding analysis by goal method
I have also looked at how goals were scored (crudely breaking them down into ‘open play’, ‘set piece’, ‘long range’, and ‘penalty’) with the highlights set out below:

Scoring trends
Aston Villa – Open Play 83%
Tottenham – Open Play 83%
West Ham – Set Pieces 67%
Stoke – Set Pieces 60%
Newcastle – Set Pieces 60%
Sunderland – Penalties 40%

Defensive trends
Arsenal – Open Play 100%
Chelsea – Set Pieces 100%
Sunderland – Set Pieces 60%
Man City – Penalties 50%
Man Utd – Set Pieces 29%
Man City and Chelsea – Open Play 0%
Arsenal, Blackburn, Man City, Newcastle, and Wigan – Set Pieces 0%

I think the scoring trends are less useful on an individual basis given that we pick players we rate and who are performing so even a poor score in one of these rankings does not preclude a successful fantasy season (if a team scores 60 goals with only 20% from forwards, if all those goals belonged to a single player he would still enjoy a useful 12 goal season).

I do however like the scoring/defensive trends as another layer to build onto the clean sheet rankings. If for example you see that Arsenal have yet to concede from set pieces, then suddenly that tricky road game to Stoke or West Ham looks less daunting given those teams’ reliance on set piece attacks.

Equally, we can also use this data to try and strip out anomalous results that may suggest where future value lies. For example, Sunderland (and hence Darren Bent) are unlikely to benefit from 2 penalties every 5 games and so his and their goal-scoring prowess must be doubted. We could also suggest that a team like Fulham are likely to have improved defensive fortunes as their 2 own goals conceded in just 5 games must surely be an aberration that will regress to the mean over the course of the season (whether or not that means less goals conceded is arguable and outside the scope of this analysis).

Given that these numbers can be a bit abstract, I will incorporate them into the weekly rankings as well as specifically referencing them in my write up rather than producing an exhaustive list which would prove hard to use. The lesson is that just because a team concedes goals, how and to whom isn’t always clear and we should try and figure this into our analysis when selecting our teams.

No comments: