By Tony Attwood
I have often explored the argument that this season the balance of home and away wins has changed in the Premier League.
And because my favourite question is always “why?” I have indeed tried to ask that.
It is a question that is all the more interesting not just because the stats it reveals are very odd, but also because as I have tried to explore the issue so other publications have come out and denied that the issue exists at all.
Now obviously people can write what they want in the UK, but when it is the Guardian and the New Scientist magazine claim there has been no change in the numbers, it is worth taking note. Have I really screwed up big time?
New Scientist is an important science magazine in the UK, and their statement about home and away wins was totally out of line with the truth. But they went out of their way to publish a special report on how there was no change this season in home/away bias when there clearly is.
To me that smacks of a deliberate attempt to head off a discussion – and we can immediately see what that discussion is…. for the only explanation that has an evidence base to show why home and away figures have reversed in crowdless games is because of the way referees respond to crowds.
I’ve been through that evidence lots of times and the articles are still on this site (see below) so now I just want to update the data.
This season there have been 122 home wins, 128 away wins and 78 draws. Below is a chart comparing that in terms of percentages with recent seasons along with goals per game.
| Season | Home win | Draws | Away win | Home goal | Away goal | Both scored |
| 2020/1 | 37% | 24% | 39% | 1.33 | 1.32 | 48% |
| 2019/20 | 45% | 24% | 31% | 1.52 | 1.21 | 51% |
| 2018/19 | 48% | 19% | 34% | 1.57 | 1.25 | 51% |
| 2017/18 | 46% | 26% | 28% | 1.53 | 1.15 | 49% |
| 2016/17 | 49% | 22% | 29% | 1.60 | 1.20 | 51% |
What this shows us is that the percentage of home wins in this season without crowds is considerably lower than ever before, but before year on year the relationship is similar. In 2018/19 and the number of away wins rose, but this is what we would expect given that three clubs drop out of the league each season, and we don’t get the same teams winning the league each season.
So some fluctuation is expected but this season it is more than that. The home win figure which has always been the highest percentage is now the lowest. And away wins have taken over.
Now if this were just a normal shift of figures – the ebb and flow you get as clubs recruit new players and new clubs enter the league – we might expect home wins to become draws or draws to become away wins… but here it is a total reversal – home wins go down away wins go up, draws stay in their normal range.
There is one other factor in this lock down season – goals.
The number of goals scored by the home and away teams are now almost equal because the number of home goals has dropped considerably. From 1.57 goals per game in 2018/19 to 1.33 goals per game.
VAR is of course a new factor this season, and that might have changed the number of goals scored, but why would it have taken the average percentage of away wins which was running at around 30.5% across the last four seasons, up to 39%?
Unless we argue that VAR gives an away advantage on its own, the key factor that has changed, of course, is the lack of crowds. Which is what we have been arguing all the way through. The crowds affect the referee decision making.
The articles linked at the end of the article give you all the data and findings as we have investigated this, but if you don’t want to read it all, here is the simple fact.
A series of experiments conducted via one of the most prestigious universities in the UK took a group of professional referees and showed them each a series of full matches. Half the group had silence on their headphones, half had the actual crowd noise. At each contentious decision the video was stopped and the referee in the experiment had to give his decision. The referees with crowd noise showed a distinct bias to the home team throughout.
These figures above now show what happens when that bias is removed.
The conclusion is simple. Referees are not neutral arbiters of each game. This doesn’t prove that no other bias is creeping in, but it does prove the crowd noise bias is at least one factor in games.
But here’s a funny thing. Figures from other countries don’t show such a change during a period of games without crowds: only England. Which is why when the figures have been reported in the English media they have always cited figures from “across Europe”. In other articles they rarely do. In issues relating to crowdless stadia, they invariably do.
Here’s the background.
- How even a serious science magazine has used fake data on football
- Someone is trying to fake football stats – and doing it rather well,
- Crowdless stadia stats reveal fallacy of PGMO claims
- Proving unconscious bias by referees
- How clubs manipulate referees through their tactics
- Referees are not 98% accurate but only 75% accurate
I think one has to allow for the possibility that players’ behaviour may also be being influenced by the lack of crowds. For example, bad fouls against a home player may engender a greater wave of opprobrium from a packed stadium, than that on a visiting player. This in turn may affect the mood and determination of the players in different ways. Perhaps away teams are less cowed & more bold as result
Of course, as you say, the refereeing will be influenced in similar ways. How the refs respond is either down to a pre-formed agenda or by crowd noise. Or both.