Saturday 8 September 2012

Newcastle's 2011/12 "Lucky" Season.

There's little doubt that Newcastle's 2011/12 season was a huge success for both the players and management. They finished five points and one place shy of the fourth place that in any typical year would have guaranteed Champions League football at the historic Sports Direct Arena. Their success was built on an impressive strike force of Cisse and Ba and a tactical approach that largely shunned possession, while remaining pleasing on the eye.

It's therefore understandable that many should take issue when their lofty finishing position is attributed to luck. Much of the debate revolves around a confusion of terminology. Luck implies a reward reaped by someone, whether deserving or not, through events that they have no ability to influence and that doesn't really seem to apply wholly to the results of a series of football matches. Cisse's wonder goal against Chelsea at the tailend of last season certainly appeared to have a component of luck about it, but there was no doubting the skill involved in striking a ball with such venom and swerve, while also avoiding a severe dislocation of the right knee.

So how should we define luck in sporting contests ? Again a coin analogy is the best starting point. A fair coin has the talent to land on heads 50% of the time, but that doesn't mean that in every 4 tosses, two win be heads and two tails. The rate at which a coin displays it's "talent" only becomes apparent after a much larger number of trials and the proportions of heads begin to trend towards 50%. The distribution of successful tosses in much smaller runs will jump about alot, sometimes hitting lots of heads, giving the appearance of being in form and sometimes not. That's all part of the random nature of events.

The same randomness that prevents a coin from continually alternating between a head and a tail is also present during a football match. It pays to have the best talent available to you, but if your star striker lines up a last minute spot kick where history tells you he converts such chances at a rate of 78%, he cannot guarantee you a goal. His success rate would have to be a bona fide 100% for that to be the case.

If penalties are to be missed you would prefer them to be missed 86 minutes into a 6-0 romp, rather than in the last seconds of a stalemate, but the random component of a skill means that you are largely powerless to ensure that the best case scenario always happens. Being skillful is plainly good, but being skillful and have random manifestations of that skill fall in a fortuitous pattern, (as may have happened at Newcastle) is an extremely potent mix.

One way to look at how "randomly blessed" Newcastle were in 2011/12 is to look at the high and low points from each game during the season. Games states are constantly changing, usually only slightly by dint of time elapsing and sometimes dramatically through goals or red cards and at any point in a match a team will have an associated points expectation derived from their likelihood of winning or drawing the game. By looking at these highs and lows we can better see what the average seasonal up and downside was for a side. One of the biggest swings against Newcastle came at home to Wolves. Leading 2-0 going into the 50th minute, the home team would average 2.95 points from that position, but at full time their points expectation had collapsed to 1.

Newcastle's High & Low Points During Each 2011/12 Game.

Opponent. Highest Points Expectation Lowest Points Expectation.
Arsenal. 1.07 1
@Sunderland. 3 1.09
Fulham. 3 1.92
@QPR. 1.23 1
@Aston Villa. 1.06 0.30
Blackburn. 3 2.24
@Wolves. 3 1.25
Tottenham. 1.25 0.18
Wigan. 3 1.45
@Stoke. 3 1.02
Everton. 3 1.60
@Man City. 0.61 0
@ Man Utd. 1 0.13
Chelsea. 1.11 0
@Norwich. 1.29 0
Swansea. 1.90 1
WBA. 1.72 0
@Bolton. 3 1.25
@Liverpool. 1.56 0
Man Utd. 3 0.83
QPR. 3 1.79
@Fulham. 2.11 0
@Blackburn. 3 1.36
Aston Villa. 3 1.43
@Tottenham. 0.54 0
Wolves. 2.95 1
Sunderland. 1.65 0.15
@Arsenal. 1.39 0
Norwich. 3 1.81
@WBA. 3 1.21
Liverpool. 3 1.23
@Swansea. 3 1.23
Bolton. 3 1.59
Stoke. 3 1.88
@Wigan. 1.41 0
@Chelsea. 3 0.63
Man City. 0.95 0
@Everton 1.16 0
Actual Points Total. Best Case Average. Worst Case Average.
65 83 31.5

The first point to make is that these game states are purely theoretical. You can't take 1.16 points from a single game against Everton. But if you replayed enough scenarios from that particular game state, then your average points haul would be around 1.16 points. So by summing the various game states to get a best and worst case points total we have combined both real and theoretical points totals. It is a useful device to get an average indication for what might have occurred had the cards fallen more or less kindly for Newcastle last term.

If everything had gone right last year, Newcastle could have probably secured Champions League football with an average of 83 points. Their actual points haul in each "lucky" season would be centred around 83 points, even  including the slimmest of slim possibilities that they may have won all 38 games.

Of much more relevance to the 2012/13 Magpies is the range between last year's best and worst case scenarios. The mid point is 57 points compared to their actual total in 2011/12 of 64. Might the 57 points be a better indication of what Newcastle "should" have got during last season's campaign and might that be a better indication of what a luck neutral Toon side may achieve this term ?

We can repeat the process for Chelsea, the team who finished just behind Newcastle, but are legitimate title contenders and top four constants, as well as being the current holders of the Champions League trophy.

Chelsea's Best & Worst Case Averages for 2011/12.

Points For Chelsea.
Best Case
Points Average.
Worst Case
Points Average.
64 93 45

Chelsea's best and worst case averages are over 10 points above Newcastle's respective totals, despite the London side gaining one less actual point than Newcastle and their midpoint is 69 points. So despite finishing line abreast with Newcastle, "in game" situations may indicate that bad luck, much of it self inflicted may have prevented Chelsea from gaining a more representative 69 points. 

Perhaps tellingly,current points projections for 2012/13 have Chelsea ending the campaign with 74 points and Newcastle with just 54, numbers that are closer to their expected midpoints from last year than their actual figures recorded in May.

Every team experiences random variation from their true ability over a short 38 game season and a team which finishes just outside the top four and isn't part of the EPL's major six sides has very likely benefited from some good fortune combined with excellent play. 

Further supporting evidence is present in the records of teams who finish just out of the top four positions. Over Premiership history, the average goal difference for sides finishing 5th to 8th who aren't one of Man Utd, Man City, Chelsea, Arsenal, Liverpool or Spurs is plus 7. The average points total is just over 58 and the average finishing position is 6.6. In the following season, average goal difference drops to minus 3, points total is 50 and finishing position is 11th. Four such teams were actually relegated and 75% of them recorded lower final points totals. So the overwhelming evidence is for a return to earth following an over achieving season previously.

If Newcastle of 2012/13 manage to breach even the 50 point barrier, they will have bucked a very strong trend.


No comments:

Post a Comment