Lies, Damn Lies & Statistics

Anyone who has read my blogs or my forum posts will know that I love a good statistic. My head is full of them – I’m one of those people that can remember who scored a goal in an obscure match but can’t remember what he had for breakfast this morning. I even had a statistic on my recent podcast debut regarding Celtic having never won in Inverness by any more than two goals and even then that was six years ago. Well, until last Saturday of course when they equalled their best ever result up there.

And before you correct me, yes there was a 3-1 win prior to that, but Inverness were groundsharing at Pittodrie at the time. See, my memory remembers all sorts of random stuff!

As far as I’m concerned, those are good statistics as they are an indication of how tricky a game it is for Celtic historically and that they may struggle to get a positive result. But the key in that sentence is the word “indication”. There’s is a terrible habit in this country – though not confined to this country – for people to take statistics and throw them about as the facts themselves. Not only does that get on my nerves, it can be quite dangerous depending on who’s doing it.

Ebbe Skovdahl was never a hit at Aberdeen, but he did come up with an absolutely gem of a quote which sums it up perfectly.

“Statistics are just like mini-skirts, they give you good ideas but hide the most important things.”

I tried to prove it on Twitter at the weekend. It back fired a little bit, so I’ll explain the whole thinking here.

If you look at the Rangers v St Johnstone game from Saturday afternoon, you’ll note that there is one big statistic that looks somewhat eyebrow raising at first glance. The number of bookings in the game was two to Rangers and three to St Johnstone, yet the fouls committed statistic quite clearly said Rangers had committed sixteen to St Johnstone’s six. Or to put that another way, St Johnstone were booked for 50% of their fouls committed while Rangers were booked for just 12.5% of theirs. To put that in a little bit of context, in the earlier match there were cards shown to Inverness 28% of the time and Celtic 33% of the time. So, on the face of it, there looks to be a massive imbalance in Rangers favour here.

I tweeted that as my “favourite statistic from Saturday”, with a hashtag to suitably stir things up of #thatswhywereparanoid (I hate that hashtags don’t allow apostrophes), and left it at that for a while. When I returned several hours later, I found that I had got one of four reactions.

Many Celtic fans retweeted it, including Phil Mac Giolla Bhain interestingly. Many Rangers fans replied suggesting that maybe I should shut up about refereeing decisions given the red card Inverness had received earlier the same day. One of my mates suggested that my favourite statistics really should have been “Rangers goals: 0”. My mate was of course spot on, although St Johnstone also having 0 goals was slightly disappointing! I’ll come back to the fourth reaction shortly.

The Rangers fans reaction amused me. As usual, rather than answering the specific point being made the opposition fans turned it around and asked something about my own team. This happens a lot – in both directions to be fair – and has got to the point there’s a ridiculous word doing the rounds to describe it. The “whataboutery” of football. Whatever you want to call it, it’s usually a great indication of “we know we are wrong but we don’t want to talk about so we’d rather talk about why you are wrong because it makes us more comfortable and you less comfortable”. Try it yourselves, you rarely get a straight answer out of anyone when you confront them about their own team.

The Celtic fans reaction amused me as well. I knew perfectly well what I was saying and trying to do, but they didn’t. The fact Phil Mac Giolla Bhain also retweeted it is where it backfires though, because I’m still having to reply to people who take me to task about what is effectively half a point! Still, it appears that Celtic fans merely lapped up the statistic and spread it about like it was yet more proof that Rangers get all the benefits, especially at Ibrox. For me it was proof that football fans have extremely short memories.

Back in October – less than one month ago – I read statistics regarding the “bookings per fouls” ratio in the SPL. These statistics showed that the Saints Mirren and Johnstone were being booked about once every five fouls, while Rangers were seeing about one in seven, Dundee United were one in nearly eight and Celtic were a whopping one in thirteen. Rangers fans of course took that as being proof that Celtic’s campaign against the SFA last season had worked and the referees were leaving them alone. So much so they were still talking about them at the start of November. See what I mean about short memories?!

When I first read these statistics, I thought it confirmed what I already knew. Celtic are very lightweight this season. The tackles Celtic have been putting in are soft, and the fouls being conceded are equally as soft. I’ve seen them concede enough goals recently where I’ve moaned about them not putting tackles in. But that’s statistics for you, you can take what you want from them.

The fourth and final reaction I got to my twitter test was the correct one – and it came from our forum administrator himself. Craig’s response was to point out exactly why St Johnstone had received yellow cards. One of them was for booting the ball away. Another was for a deliberate tug back. Both instant bookings, no matter how many fouls you’ve committed previously.

That’s precisely the point I was trying to prove. You cannot compare the number of fouls committed with the number of bookings received. It doesn’t add up, it never has. You could, in theory, have a match where one team doesn’t commit any fouls at all but a player boots the ball away every time a goal kick is conceded to waste time. Zero fouls, several bookings. More realistic than that though, one foul is not equal to another. That’s why sometimes you see a straight red and other times you see just a yellow. Without context, the statistics being banded about are useless. But it doesn’t stop people using them to prove whatever point they want to prove.

Take another statistics that has been doing the rounds the last few days. Celtic haven’t won a match against eleven men since mid-September. Technically correct, but in exactly how many of those games was the sending off an important factor in the outcome of the game? The 2-1 win over Aberdeen was already at 2-1 when Ryan Jack was sent off in the closing minutes. Ivan Sproule was sent off at 4-1 up in the League Cup match. Rennes were 3-1 down when they went down to ten men for the final few minutes of their game at Celtic Park. Motherwell played out the last seven minutes of their defeat with ten men, already 2-1 down. In fact, Saturday’s sending off of Greg Tansey is the only game Celtic scored after the sending off, and if anything it continued to be an even contest until Celtic made a second half substitution – suggesting that was the major turning point, not the sending off in itself.

Last week even proved just how dangerous statistics can be. The Scottish Government, a body still intent on pushing forward a highly ill-advised bill for which the supporters of many teams have massive objections, put out statistics for religiously aggravated crimes. The media even helped them push the point further by choosing to highlight the football specific figures. Of course, the one that everyone immediately pointed to was the 47 arrests that took place at Celtic Park – the highest of any single ground in Scotland. Must be Celtic that’s the problem, right?

Wrong. Celtic’s response to this was to put out the full figures – to put the statistics in context if you will. Of the 47 arrests at Celtic Park, 33 of them were from the away support. 14 people out of over 1 million home supporters were arrested. That would be 0.0014%. Celtic have banned these 14 people. When you consider the away support total is 35,000 people even that percentage is 0.09%.

I don’t have figures for the other grounds because Celtic were the only team to release theirs. But the statistic that should be focused on more than anything is the one that says 67% of all religiously aggravated attacks have nothing whatsoever to do with football. Even that statistic is questionable since only 90 of the 693 offences actually took place at football grounds.

Why the Government are pushing this bill through is a question for another article, perhaps one from a supporter of another team that is not quite so polarised in popular culture as mine. I note with interest that Motherwell fans have been particularly vocal in opposition of this bill recently, something that was quickly followed up by some reportedly heavy handed treatment of their supporters in Aberdeen by the police.

Statistics are great little tidbits for in the pub, or for use in blogs by us amateur writers. In the right context, statistics can even be useful or “statistically significant” to use the correct technical term from my school days. But for the most part statistics should be taken with a pinch of salt.

After all, 74% of all statistics are made up on the spot.