I once subscribed to the view that Charles Reep was a villain of the English game. For those who do not know who he was, he collected statistics of thousands of games after the war and through the 50s and part of the 60s and deduced that it is better to create chances with as few passes as possible. His findings were attributed to the introduction of the hated long ball game.
I now see him as a hero. A visionary who was ahead of his time. His findings were not relevant for the modern game because he sampled games where skill levels were low. When this is the case, he was right. If a team on average struggles to string 4 passes together, they need to try to create chances by making no more than 1,2 or 3!
There are many in football, including still some managers, who don’t realise that statistics provide so many answers. Jose Mourinho once bemoaned English fans for celebrating getting a corner as if they had scored a goal. He knew the numbers proved this shouldn’t be the case – It is a case of selective memory. You remember the few corners that result in goals and not the many that don’t or even the corners that result in goals for opponents beacuse you have committed players deep into your opponent's area. Having said that, a team playing us in recent years where we defended corners poorly, could justify some excitement when they got one!
You will get a load of people in the game saying that the most dangerous time in terms of conceeding is when you have just scored a goal. Where as in fact, statiscs have shown this is not true - actually you are a bit less likely. But you remember the times when you do. People in the game still believe this is the case, despite it being proven not to be so.
I’m sure I will be ridiculed by some, but I will be trying to appraise our results and prospects from a statistical standpoint. Albeit a bluntish one! There will be some assumptions, which if I had the time I could firm up and be more scientific but hey ho. To elaborate, tomorrow we play Shrewsbury at the Valley. I would say that we are in this sort of game a 1 goal per 8 chances team. Our opponents are probably a 1 goal in 10. So statistically, to score one goal, we need to create on average 8 chances. These numbers need to change depending on the defensive qualities of the opposition of course.
That doesn’t mean we will, teams can have 40 chances and score none and their opponents 1 and win. This does happen and whilst some will say, this proves using statistics is wrong, it doesn’t. The law of probability, says this will happen – the question is, does it happen more or fewer times than it should. Studies of the top leagues in Europe over a number of seasons have shown these freak results happen pretty much as many times as you would predict they will.
So lets say we create 16 chances to Shrewsbury’s 4 tomorrow. This will firstly mean that Shrewsbury would be most likely to score 0 if they are a one in 10 chance team! If we are a one in 8, we are more likely to score 2. So that is the result you would expect to occur the most if you played the same game with the same number of opportunities a hundred times. Of course as a one off, it is down to probability and luck how it turns out – but if we create 16 chances, we will be unlucky to not score one and Shrewsbury would be lucky to score at all. Of course sometimes you are lucky and over a hundred games, there will be some Shrewsbury wins.
Ok, so what is the point? Well, I stated before the Euro Championships the issue with England was that they didn’t have enough goals in the side. So if a team restricted England’s chances, their lack of goals per shots becomes an issue. If you are lucky as a manager, you have centre halves and midfielders who get their share which improves your chance to goal ratio. But if not, you have to use this information to influence you selections. It isn't about just who you think are your best players. Now moving on to us, we would clearly benefit from a goal scoring central midfielder and a way of creating enough chances to make the odds work for us. Tex, would be a valuable player to keep as he is a centre half who knows where the goal is. If I was a manager, I would pay to get all the analysis done properly, and try to make the numbers work for me. Any manager, who doesn’t use statistics in the modern game is a dinosaur.
6
Comments
Bury -
Chances for 8, Chances against 11 - verdict - a bit unlucky not to get a 1-1 but defeat highly possible outcome (Luck scale -1)
Northampton
Chances for 17, Chances against 7 - statistically most likely score 2-0 or 2-1 - verdict - a little unlucky not to score one more and concede one less.(luck scale -1)
So we have been a bit unlucky but nothing major so far! After 2 league games my luck score is reading -2 (average -1 per game).
I'm guessing that goals/corners ratio changes markedly as the quality of the defenders diminishes. As, you say, it seemed like teams scored twice every game against us from a corner last year! (And usually in the 44th minute.)
But our inability to defend corners adequately pre-dates the current regime and goes back to at least 2012 (when, I think, Morrison, Taylor and Devrite were the central defenders). In a way, it has been a measure of our decline.
So, (unless a team is playing Charlton), on that analysis, getting a deflection off a defender and the resultant corner is pretty much a waste of time. But perhaps a throw-in would be a different matter (see below)?
Here's an interesting blog analysing Northampton Town's goals. https://footballperformanceanalysis.com/tag/throw-in/
It claims Northampton score over 50% of their goals from set plays and 40% of those are from throw-ins. It also highlights that many aren't from long throws near the goal line but right back to the half way area and that nearly half of the resultant goals were scored with feet rather than from a header.
I hear that Magennis is a long-throw specialist? Statistics tend to suggest that with proper coaching and preparation, it might be a skill that should be used if we want to score more.
The only time I ever encouraged my kids to jeer at a football match was on 23rd November 1991 as Cambridge United left the pitch. It was at Upton Park and they had just inflicted a 1 - 2 defeat over Charlton.
We knew what to expect of course as the manager’s reputation went before him. In the warm up, his team of beanpoles split into two and lined up facing each other on the touch lines. They then proceeded to pump 40 / 50 yard lobs at each other. Their K.O was also bizarre. They lined up like a rugby side - three guys out wide on the left who charged off at a rate of knots on the refs whistle, hoping to win the ball - and if not that, a dead ball situation - any dead ball situation, be it free kick, corner or throw in.
John Beck adopted any stifling tactics that statistically gave him an advantage, and sadly to some success (two promotions and two F.A. Cup quarter finals). The only problem being, it was killing football, who on earth would want to witness this dross on a weekly basis?
The only comfort I take is that (imo) it will never ultimately succeed against teams and players of true quality.
My system is a bit of fun. I would need a team of people to work out accurate goals per chance ratios and how they are affected by qulaity of opponents and playing home or away - so I have to apply a judgement. But I think it is an interesting way to look at games from a supporters perspective. I suppose it could be argued that if the -1 luck was +1, we probably would have 4 points. Anyway, just another way of looking at things.
I don't have the exact passages on Reep to hand, so I'll have to borrow from the Wikipedia article (very sorry): EDIT: Pressed send too early. But I think that while stats are of course all very well and good - read them right first.
Good idea for a thread Muttley. I don't know the numbers but I'm not sure if there is a difference that would be the reason. For worse defenders lower down the leagues you get worse strikers and worse corner takers. I read somewhere it's only a 3% chance of scoring from a corner. Or another way 97% of corners will not result in a goal. So as Muttley wrote why do we get excited when winning one? It's quite a weird phenomenon and I think just another of those things that football fans perceive to be of value. A bit like moaning about a ball being passed backwards rather than forwards.
Question? Where did you get the average goals to shots ratio? After two games it can be bad luck but if after, say, ten or twenty games if we are still finishing -1 on each game does that mean we are unlucky or that we need to ha he the ratios? Also are you getting the ratios for the other clubs from somewhere or are they assumptions?
https://www.theguardian.com/football/these-football-times/2015/mar/04/praise-sam-allardyce-manager-bolton-wanderers-premier-league
So according to BBC we had 12 shots to their 14. But a higher percentage of their shots were hardly worthy of the name. Having said that, We probably over achieved statistically getting 3 in that game. The thing was, we were rampant when we got to 3 but took our foot off the gas after that a bit. Anyway, I would say that getting at least one goal was about a 98% chance, two goals around 80% and 3 around 45 to 50%. Shrewsbury getting a goal was about 80% and two was around 50%.
So for a change we got what we deserved. We were a bit lucky to get three - statistically (it was a comfortable victory) and Shrewsbury were unlucky not to get one. I would say luck was neutral - 0. I apply luck to the outcome rather than goals scored. So after 3 games our luck is at -2. That is an average of -1.5 per game. If luck had been neutral for every game, we would have 7 points.