Watford drew the curtain on their latest pre-season campaign this weekend with a semi-impressive draw in this year’s prestige friendly against mother/sister/great aunt club Udinese. The Hornets haven’t pulled up any trees with their performances in the last month, leading some to openly wonder if we are heading into the season lukewarm, in first gear, or simply as a not very good football team. ‘I know you can’t tell from friendlies but…’ has been a popular refrain in the last few weeks, usually followed up by criticism of Beppe Sannino.
So as we anxiously twiddle our thumbs waiting for Saturday to come around and another nine month cycle of anguish to begin, ITWM decided to put their Maths GCSEs to the test and see if pre-season results really are meaningless, and if the warm-up for the 14/15 campaign has actually been that bad.
One of the reasons that the results of friendlies are so denigrated is that it’s nigh-on impossible to compare them. Below is a table, compiled with the help of Jon Sinclair of the Watford Mailing List, summarising each of Watford’s last ten pre-season campaigns.
So, how much of a correlation is there between pre-season results and how the side have performed throughout the league season?
Taking the “points” gained throughout the campaign wouldn’t work due to the differing number of matches, so here’s what it look like when the average points per game is plotted against Watford’s final league position:
As we can see from this, the two best pre-seasons results-wise in recent Watford history have been in 2006/07 and 2013/14. The former, you’ll remember, saw us reach November before registering our first league win in the Premier League, and ended with Aidy Boothroyd’s Watford slumped at the foot of the table with a paltry five wins, wistfully dreaming of that 4-0 win over AS Horsens in July.
And a spate of non-competitive victories last summer was just a tantalising starter for a bland, deflating and ultimately bitter main course that saw the back of Gianfranco Zola. Despite dispatching representative sides from Val Venosta and South Tirol by an aggregate of 13-0 to kick start the season, even the most forgiving of Hornets fans could be found despairing at the flat performances in the depths of winter. How soon we forget.
In fact, when comparing these two data sets there is a correlation of 0.155 – i.e. bugger all.
But it’s not about the result is it? Anyone can beat a team from the Conference South or some Spanish side so lowly it’s quite possible that the club’s PR department have made them up. The sign of strength and preparedness is how you beat them.
So how does it look when we plot league position against the goal difference per game, showing the average amount of goals each game was won by?
Yet again pre-season results would have us think that last season was going to be a cakewalk. Franco’s free-scoring charges netted twenty three times and conceded just twice. Clearly we had retained our attacking prowess from the previous season, despite the loss of Matej Vydra, and with a few shrewd signings had shored up the occasionally-porous defence that had denied us promotion.
In 2011/12, Sean Dyche’s first (and only) chance to subject a Watford side to his favoured brand of gruelling post-summer labour went down poorly. The acquisition of a bunch of northern, and heaven forbid even the odd Scottish, journeymen to go with our young squad meant that expectations were low, and finishing our preparations on a goal-scoring par after a schedule of Borehamwood, Wealdstone, AFC Wimbledon, Brentford and Colchester did nothing to dispel that nagging feeling that we’d better start looking up the way to Rochdale.
As it happens, after a spate of form after January, when the side looked fresh and energetic, we flirted with the playoffs and everyone was rather happy with that thank you very much.
Again, if we calculate a correlation between the two factors we arrive at 0.096 – even less than the last one. So while it means nothing to grind out results for imaginary points in the pre-season, it seemingly doesn’t help to turn sides over. There’s no morale rollover; joie de vivre doesn’t win promotion, at the beginning of August everyone starts on nil points.
Of course one reason that judging seasons by the friendlies that precede them doesn’t work is that the standard of opposition varies every year. You can look at league tables and know that every game was against teams of the same ballpark standard – 13th last season was poor compared to 3rd the year before, and switching Bristol City for Bournemouth didn’t do much to sway that.
But there’s a difference between beating Northwood 7-1 and battling to overcome a Serie A side like Chievo or Parma. Without that information the above table is completely meaningless (sorry about that).
I did start to calculate the average league position of each opposition for the last decade, but somewhere during the 2009/10 season when trying to find out where CF Balaguer finished (9th in the Tercera Division and thus 71st in Spain) I gave up and decided to just compare each year’s Borehamwood result instead:
There’s a correlation of 0.378 between how much we beat Borehamwood and how well we then go on to do that season, so while friendly results mean absolutely nothing in general, we might do well to take note of the result at Meadow Park each year.
As I said earlier, when you’re deep into December and the struggle of league football is all-encompassing nobody remembers pre-season; nobody cares how fluid you attacked on a Tuesday night in St Albans.
What matters then is fitness levels allowing you to keep fighting until the last minute and not letting that winger get past you and put that 89th minute cross into the box. What matters is having an answer when the opposition have crowded the midfield and you haven’t gotten into the game at all.
Reports out of the camp this summer suggest that Sannino has been working the players hard. He told the Watford Observer that the performance at Borehamwood suffered as the squad had been in the day before for a double session.
Though this clearly shows a lack of research into the slight relationship between Borehamwood result and success, it does mean that Beppe acknowledges the need for greater fitness amongst the squad. Alarm bells rang last summer when players were excitedly reporting that they’d got the balls out on the first day back. Championship players are Championship quality, and kicking a ball about for a month isn’t going to change that dramatically, but the margins of sports science are so precise in the modern era that maximising fitness can make a huge difference to a side’s fortunes; especially if it leads to your best players not missing months of the season through injury.
When he arrived at the end of last year, our players were – in the eyes of the fans – ill-disciplined and unfit, now Beppe has done his best to create an identity of defensive solidity (by his own admission to the initial detriment of some of our attacking verve) and is now trying to remedy the propensity of the Pozzo era Watford of gifting away points in the death throes of matches and fading in the closing months of the season.
He is also making sure that when we meet brick walls we are able to find our way around them, rather than just bashing our head against them until something gives. So convinced were we this time last year that we would just waltz through teams with crisp passing that we didn’t prepare to do anything else. We called for a change of shape, but real life isn’t Football Manager and you can’t just throw players out in a different formation and expect it to work.
In the past month the side has gained match practice in at least three different systems and players have been used in a variety of roles, not only do we not have a few different plans of attack, but players such as Tommie Hoban have been encouraged to work on different parts of their game and adapt to new positions. I don’t expect that the “Irishman” will start against Bolton at right back, but if he needs to fill in there he’ll be more prepared to do so. Rather than shoving square pegs in round holes at a moment’s notice in March, we’ve sanded a bit off now so that they fit that little bit better.
None of which means that we’re heading for success. By all accounts we haven’t looked at all good against competent opposition, and Vydra – a player we need to rediscover his form – has been ineffective in both a wide position and on his own up front. And of course a lot rides on whether Troy stays or goes.
But that means nothing now. It only means something if it keeps going into the season. However if we’re getting hot at the turn of the year and running teams into the ground then it would be good to remember the work that’s been done behind the scenes to improve players and team this month, and to not pin too much on how we look next July.