By Tony Attwood
A comment made recently on Untold had me pondering the statistics issue again.
I had written about Arsenal’s finances vis a vis marketing income, and noted how the marketing income was rising, while patting myself on the back for having predicted this while an eminent economist had suggested it would not happen.
In response came the comment
Now we only need an adept manager capable of competing and winning at the highest level. Not someone finishing the league on average 13,57 points off the league winner over the past 13 years, apparently irrespective of what players he brings in (özil, sanchez, Mustafi, xhaka…) and the (growing and massive) amount of money he has at his disposal.
There were two problems here. One is the mention of players brought in at cost over three years compared to the points deficit over 13 years (thank you Jammy J for making the point in the comments section) and the fact that the word “only” was used at the start.
Clearly only one team wins the league each year, and yet there is an implication that somehow Arsenal could have had such a manager at some time in the past 13 years. One can argue against this in two ways – that the money to buy the players mentioned in the comment has only recently become available – and has only become available now because Mr Wenger kept us in the top four all the way through, and thus enhanced the club’s income. And second that even with the money, bringing in a new manager does not guarantee success.
Manchester City and Manchester United have spent a lot more than Arsenal but are not at the top of the League. Chelsea spent a lot more leading up to last season, and had to put up with a mediocre mid-table finish – a finish which enabled them to by-pass the Champions League (no problem because they don’t need the cash as Arsenal did during the period of paying for the stadium.)
So why can’t we settle all this with statistics?
For years on Untold I have tried to extol the statistical route, and many who are much more adept at numbers than me have helped with stats that reveal the inconsistency of referee decisions, the link between referee organisation in the PL now and in Italy during its match fixing days, the way teams that are highly lauded by the media use long ball hoofers, the fact that big money transfers don’t normally work in the first season, the issue of why November is a bad month for Arsenal, the question of why we have more injuries than everyone else, and on and on.
But lots of people don’t believe numbers. Just before Trump was elected 68% of his supporters said they did not believe the economic data published by the government. In the UK, a research project by Cambridge University and YouGov found that 55% of the population believes that the government “is hiding the truth about the number of immigrants living here”.
If S Robson on Talk Sport had been asked why last summer he was still shouting that Arsenal had more injuries than anyone else, when all the figures said this was untrue, he would probably have said, “you can prove anything with statistics”. It was true, it seems, because he said it.
So we are now in the era of not trusting experts and as Michael Gove so joyfully announced, and not trusting numbers. Numbers – the one thing that is supposed to be the ultimate definition of how the universe works – slippery buggers and not to be trusted.
What we find is that people respond now to the issue of today – as in last week’s result – and highly simplistic issues like “top four is not a trophy”. But they don’t like statistics.
As the Guardian says, “People assume that the numbers are manipulated,” – although they don’t point out that their own senior journalists can do this, as with my oft quoted (and it is oft quoted because it was so blatant, and anyway I find it one that is easy to remember) comment that Arsenal had only two players whose goalscoring was in double figures last season. True, but the “only” was misleading, since it was better put that Arsenal was one of only five clubs that had two players in double figures. Statistics as a way to mislead – a perfect example – but not a reason to give up on statistics.
The Guardian’s conclusion on the lack of faith in numbers is that, “All of this presents a serious challenge for liberal democracy.” And of course what their senior football reporter says is not really in the same league, no matter how widely she is respected, but still, such a commentary does not help.
Anyway, the Guardian’s article tells us that “The declining authority of statistics is at the heart of the crisis that has become known as “post-truth” politics” and that certainly seems to be the case.
Statistics allow us to analyse issues (in our case, in relation to football) in ways that can be considered beyond pure opinion. And at this point I am reminded of the aaa website that the Daily Mail constantly links to, that claims (or claimed, I have not looked for quite a while) that it was the place where you could “have an opinion, not get one” – or something along those lines, which always seemed very curious to me in that it seems to imply that having an opinion or being given an opinion is all there is. Definitive facts no longer exist.
Statistics, analysis and logic are the tools we can use to help us understand what is going on around us and then to make reasoned decisions as to cause and effect, and subsequent action. Yes you can prove pretty anything with poorly gathered statistics and/or statistics that are used without reference to the broader context. But that is why analysis and logic is there as well.
Unfortunately, while football clubs and educated managers use statistics more and more to explore and explain what is going on in matches, it seems some managers and certainly most media-employed commentators use statistics less and less, if at all. So we can have Paul Merson and co shouting on Sky about the fact that Hull have employed a foreigner as a manger rather than an Englishman, without considering any evidence as to how well English managers do in the Premier League vis a vis foreigners.
With football it sometimes seems that even simple counting is no longer an option on the table, as with the argument over Giroud celebrating his goal against Bournemouth. He was roundly criticised for not getting the ball back to the centre spot fast enough, seemingly without any of the commentators bothering to check just how long it took for Bournemouth to restart, vis a vis the Alexis goal, where he ran back and placed the ball on the spot for Bournemouth.
As Untold pointed out, it took longer after the Alexis goal – as it generally does when a player puts the ball on the spot. The team that has just conceded and is seeing a lead slip will not willingly oblige the opposition by taking the kick off quickly – most certainly not when the ball is deliberately placed for them. Giroud did the right thing – and got roundly criticised for it.
We have more statistics available in football now than ever before, and yet the data seems to be ignored more and more, or indeed so disregarded that the opposite conclusion from that which the numbers point to, is drawn.
It’s a sad state of affairs, but just as we should not believe the ravings of politicians who predict £350m a week more money for the NHS once we leave the EU, so we should not believe the people who just make statements in football. There are facts out there, and even though the media long ago turned their back on them, we should not.
Untold Arsenal and the Arsenal History Society… recent stories
Untold’s “Money League” predictions concerning Arsenal’s finances come to pass
Fifa preparing to change football’s rules.
Football in crisis. Football in chaos. Football as a fairy story.
Arsenal need a new striker. I know this because last summer lots of journalists told me so
Would Arsenal be higher up the league if we had Pep Guardiola instead of Arsène Wenger?
The value of Arsenal’s players is not what you think. Or at least not what we thought.
April / May 1937: Arsenal slip back and Man City triumph – for the moment
Football might be the biggest sport in the world but its management is chaotic
The Index of major articles about Arsenal players throughout the club’s history: A to K L to Z
Tony
Found an interesting piece of info on pgmol and the media yesterday. It seems Dermot Gallagher doesn’t only step in for Sky and the Sun once the weekend’s action is over.
Webb mentions in his book that part of his short-lived job, post reffing, at pgmol involved sitting in a booth with Gallagher in Salford media centre to watch all the games live. From where they would highlight any incidents they thought worry of discussion, and communicate with Sky, BT and talksport, answering any queries and compiling a package of clips to distribute which they deemed important.
Their day would end on the Saturday with a little meeting with the Match of the day crew to discuss the day’s action.
It’s not a perfect means of control. Their partner broadcasters could disagree with them, and presumably choose different incidents to discuss,etc, but it does highlight how involved they are at every step, and also how unaware your average viewer would be of that fact.
When Gallagher appears on Sky, is it even mentioned that he is a key pgmol employee rather than an independent authority.
* should try being accurate in these things. Mistake for me to say Gallagher writes for the Sun. He might do but only two Newscorp jobs I’m sure of are appearances on SkySports News and the column he does on sky website giving his verdict on weekend’s key incidents.
I love stats maybe because I am a numbers man by trade. I had some free time in the summer and worked out where we dropped the most points last season and it was against teams that finished 6-9. We picked up 15 points against the top 5 in 8 matches but only 8 points in 8 matches against the teams that finished 6th-9th. My point is the numbers make things clearer- i.e. where to improve, etc. The flip side is that numbers/stats are all too easily manipulated. Sky published an analysis of goal keeping in the PL this week with a view to focussing on Claudio Bravo following mishaps at Goodison last Sunday. An info graphic placed Bravo at 20th in the league for Save percentage but if you look closely, the data reviewed is from December 1. Any stat can look worse or better when examined over a shorter spectrum.
Just been on Legrove, asked the question can you build a stadium and still be successful? Made some good points.
Tony,
I believe the issue is that today people (or what looks like a majority of them) will only look at number/statistics/facts that confirm what they believe and do not want/like/know how to deal with contradiction.
Just look at the press conference or speech Trump gave the other day where he just stated that 86 million americans where out of a job and that he would make sure they’d get one. I did not hear much of an outcry.
The number is most probably right. Just that it includes pensioners, sick, students, kids, inmates etc etc etc. Yet millions of americans are happy with it and will carry on arguing with it, none of them asking himself if it is possible that more than one out of every 4 americans is jobless ?
So in our world where everyone is paranoïd, we see people cherry picking bits of ‘truth’ or should I say bits of partial truth instead of embracing the challenge of thinking things through.
We live in a world where people zap from channel to channel until they agree with what is being said/shown, do the same on facebook feeds and in the end build a sort of personal Potemkin village that suits them.
Guess too much information kills information…
Fat Sam used data analytics to evolve his teams’ style of play,
and that was over a decade ago!
I may or may not like his style (ignoring the time wasting and OTT hacking which reduces any consideration) I can’t argue against the data.
Statistics can only be believed if the people or organisations compiling them do do with honesty and integrity.
To your point about people not believing the stats produced by government about the level of immigration, this may well have been because for years, most notably under the Bliar and Brown governments, we were told the statistics for this didn’t exist. If this is the case you can’t possibly plan for the provision of housing, education and health, all of which have been and continue to be under pressure.
You can’t have it both ways.
I really enjoy the blog and think it serves a real purpose but I do wish you would keep the blog to football matters and stop shoe horning in these lower sixth level political comments.
People are becoming lazy, it’s complicated to think about statistics and numbers to prove something and it’s a lot easier to have an opinion. It’s as simple as opening your mouth in front of tv !
I joke but it is a real issue in a lot of domain, a worrying issue !
The Cambridge University research was a little flawed as far as I’m concerned, because it portrayed people’s opinions as facts. It suggested that people “not believing” the government’s statistics meant that those statistics were false. A little bit like the comments of one of the Professors in that research who suggested that we “learn Urdu and Polish” to help migrants feel more welcome. Well, my research (with my many friends who live in the UK, but are from Poland, Hungary, Latvia, Romania, Lithuania, Norway, Sweden, Germany and France); suggests otherwise. I find that ALL of those friends can speak almost perfect English, some are a lot better at it than many Brits.
The Cambridge University research, in my opinion, was unnecessary, a waste of resources, and condescending to migrants.
In an attempt to look like they know what they are talking about (as they don’t use numerical data themselves), I just love the way pundits just highlight players and draw lines on pitches to prove their point. Of course, with no baseline, we have no way of comparing what they say about a particular passage of play to know whether it is better, worse or the same as many other plays.
My favourite which is “oft used” is the one where to prove how well a group of three players are playing, a pundit highlights the position of the three players, has a line drawn between each of them and extols the effectiveness of playing in triangles. The TV companies must think we are totally stupid if we can’t work out that it doesn’t matter where the hell three players stand on the pitch, they will always form a bloody triangle (although, of course a straight line may occasionally occur, before someone points it out!). Yet they do it again and again and are so smug about the point they believe they are making……………..
@ ClockEndRider
It’s interesting that you claim that Tony makes “lower sixth level political comments” yet allude to the idea that successive governments have actually done any planning around housing, education and health………
GM,
A fair point. Personally I hold almost all politicians in contempt and hold no party political affiliations. However the point I’m trying to make is that the Labour Party in particular constantly seeks to make play of the current state yet conveniently forgets it’s major role in the genesis of that situation, seemingly happy to insult the intelligence of the populace in an Orwellian ” we have always been at war with Eurasia” manner.
two points;
1) It’s wrong to suggest that our club no longer rely on the CL money that 4th place gains us. Although it’s true that our world won’t fall apart without it, it’s still a big chunk of money that our club would miss badly.
2)Having spent many years in a business which revolves around stats I must point out that the most difficult part of Statistics is gathering the exact data required to prove certain results. It really needs to be very specific unless your just looking for a ‘general idea of’ type results. We could look for stats for the average height of a PL LB but without very clear definitions the figure would be wrong. Gabby played there for us but should he be considered a LB. Similarly, if Chavski play with a back 3, is their wingback going to be considered as a LB or is the CB on that side of the 3 considered as the LB. The usual issue is that the figures in general circulation are collected and published as ‘fluff’ to give a little entertainment, but then grabbed by people that want it to show a certain result. The detailed information that does or doesn’t prove their theory is rarely available without a price-tag being involved (and they don’t want to pay).
So some of the Stats on Untold are very effective at proving points (despite being ignored by a loud minority), but there does seem to be a portion of them that take general numbers and try to prove a specific point which the data shouldn’t be used to support… Even if the original theory is sound.
Sorry, Gabby did of course play RB for us, but the point remains that the data used must be defined correctly before being collected.