Thursday, August 9, 2012

What to expect from the promoted teams

I am somewhat disappointed to admit that I don't follow the Championship as I used to as some of my favourite footballing memories come from Bolton's time spent chasing promotion from the old First Division*. Alas, the move to Canada and the availability of countless other sports to fill the hours have taken their toll on my viewing hours.

*Incidentally, you can get an idea of someone's age in England by what they call the second tier of English football. To youngsters it's the Championship, to anyone in their late 20s - 30s it's the First Division and beyond that it's the Second Division.

So it's with each new Premier League prediction season that I approach with some dread the issue of the newly promoted sides. I generally don't know much about teams unless they have yo-yoed back from a couple of years ago, and even then my knowledge of the individual players involved tends to range from decent familiarity (particularly with Allardyce's Bolton 2.0 side) to having a vague recollection of someone playing for Stockport in Championship Manager.

The solution, as with most unknowns on this site, is to apply averages. This approach will not be perfect and it will not help you identify the next Charlie Adam or Michel Vorm (we will get to individuals in a separate piece). Generalising does however allow us to predict how a team's success in the Championship should translate to the Premier League and thus we can get a handle on how many goals they are reasonably likely to score/concede which helps with all matter of prediction tools from weekly forecasts to captain stats.

Taking the data from the past 10 seasons, we can see how each promoted side faired in the Championship comapred to their first season in the Premier League. This gives us a 'factor' which we can take the average of and use to help us predict the totals for this year's teams. For example, when Wolves were promoted in 2002-03 they scored 40 goals at home (1.74 GPG) and followed that up with 23 goals (1.21 GPG) in 2003-04 in the Premier League. Simply taking one over the other, we can say that Wolves scored 70% as many home goals in the Premier League as they did in the Championship. Perform the same calculation for every promoted side and we have a data set which is reasonably reliable.

Taking the average of all the factors gives us the below:

   hFor
   hAgainst   
 aFor      
aAgainst
Average
  0.69
1.72
0.65
1.71
StDev
  0.16
0.59
0.25
0.37

A couple of points to note here:
  • The impact on home and away performance appears to have been approximately the same
  • The impact to goals scored and goals conceded also seems to have been approximately level. For example, we expect a team to score 69% as many goals at home and concede 72% more.
  • One key difference though is in the standard deviation line. The 'goals for' factors have been fairly consistent with that ~65-70% marker applying to a majority of the promoted teams tested. This suggests that 'goals for' carries some predictability and thus we can be surer in the forecasts below. It's worth adding here that the larger variances noted have been where teams have significantly underperformed the average such as Sunderland in 05-06, Derby in 07-08 or Wolves in 09-10.
  • Defensively however, the numbers have been quite varied, particularly at home. In 2007-08 Stoke conceded 1.17 GPG at home. On promotion to the Premier League they simply cut that down to 0.79 GPG, a factor of just 0.67 compared to a league average of 1.72. Conversely, Newcastle, a useful home defensive team in 2009-10 (0.57 GPG) came up and were quickly exposed, surrendering 1.42 GPG at St James' Park, giving them a factor of 2.51. We're therefore more comfortable in suggesting a team like Southampton might be able to sustain their strong home offense (2.13 GPG) than to suggest how badly (if at all) their defense might decline.

Once we crunch the numbers we get the below predictions for the promoted teams for the upcoming season:


   hFor   
hAgainst
  aFor  
aAgainst
Reading
 25
31
  21
39
Southampton
 34
31
  23
48
West Ham
 28
45
  26
38

These numbers are encouraging for a couple of these promoted teams (at least from a fantasy perspective) with both West Ham and Southampton forecast to score goals at a rate that would have placed them in the top half of all teams last season. The story for Reading doesn't look so rosy offensively and none of the team project to do particularly well at the back (although as we've seen that these numbers are less stable).

I'd suggest that for the foreseeable future you are going to want to regard a fixture against any of these sides as a green light game for your attacking players. None of them had a particularly good defensive record either at home or away last year and I don't see anything by way of new manager, players of philosophy to suggest right now that things will change when they face the big boys every week. Offensively, West Ham and Southampton look set to do well and facing them won't be a slam dunk clean sheet for your defenders.

As a final point, I recognise that I have ignored transfers to this point, which in theory could have a major impact on a team's reversal of fortunes on promotion. I have done this because (a) it would be almost impossible to account for them individually and (b) I don't see any one player of collection of players for these promoted sides that is likely to have a ground breaking impact on their fortune. Perhaps Mobido Maiga gives West Ham another option up top, but given that we are already projecting them to be a mid-table offensive side, I think it would be a stretch to say that he will transform them into a top-6 unit.

I'm working my way through the promoted teams now to try and identify some individuals we might want to target, so if anyone with a deeper knowledge of the lower leagues has any tips, post them below or send them to @plfantasy.

9 comments:

Jeb said...

I'd love to see some insight on the rising teams as well.

It would certainly help the members of the independent Fantasy EPL league I am running. 20 teams - 18 player rosters - weekly head to head schedule.

Would love to have you and/or some of your readers join us.
http://fantasyepl.proboards.com/

GuĂ° said...
This comment has been removed by the author.
Gummi said...

So based on this goals scored at home is the most reliable number and goals conceded carries the biggest risk for us to use?

Would that mean, based on these results, that out of all the players in the promoted teams Ricky Lambert should be the best bet (ignoring value)?

gurka said...

Lambert is one to watch probably the most exciting prospect of the 3 promoted sides

Ayoob said...

Chris, you are the best analyst! Great stuff. But it is difficult for me to look away from Pogrebnyak who's valued £5mil at the official game no matter what the numbers indicate.

chris said...

Hi Chris- love your blog.

Sorry but I don't believe your results- can you double check them? I think you have far too many goals predicted for the promoted teams.

For example: you predict West Ham will concede (on average) 83 goals is crazy. That's saying they are the worst defence the prem has seen in a long time.

Also, Wolves conceded 52 goals in the champ 2008-2009. And then, 56 in the prem, not the 88 or so that would be predicted by your model.

You say things don't look so good for Reading, but you predict they will score more goals than 8 prem teams did last season.

Have I misunderstood your goal predictions table?

Or I wonder whether a mistake has been made somewhere?

footballfan said...

I agree with Ayoob - Pogrebnyak seems to be a decent signing for the Royals and could help them a fair bit in their bid to stay up

HunterP said...

Chris -- You do have a mistake here that is inflating the numbers. I have been running the same projections for the past few years and my "factors" almost exactly match yours. (Mine are based on 11 years results now.)
The problem?
You are converting your predicted goals for/allowed based on the 46-game Championship season, not the 38-game EPL season. So, no, West Ham won't ship 83 goals, more like 65 perhaps.
My numbers (GF-GA; home, road):
Reading 20-25, 17-31
Southampton 28-25, 18-37
West Ham 23-36, 20-29

HunterP said...
This comment has been removed by the author.