| Positions, Points, Performances and QPR's season to date. 12:42 - Apr 13 with 3141 views | NewBee |
Although I don't pretend to understand exactly how it's calculated, I have become a big fan of Expected Goals (xG) analysis as a better underlying indicator of a team's overall performances than a snapshot of its league position on any given date.
I use FiveThirtyEight for this, an American site which covers many different sports (also politics). And while reviewing Brentford's progress the other day, I took a quick look at QPR, which was very significant imo:
https://projects.fivethirtyeight.com/soccer-predictions/championship/
The front page shows their prediction of where they predict teams will finish the season - as of today they have QPR in 11th place on 65 points.
However, if you click on the SPI button (just under "Team Rating"), this shows their assessment of each team's current performances. ("SPI" stands for Soccer Power Index and basically equates to a combined xG, both For and Against).
And this season the disparity between QPR's results/league position and their underlying performances has been quite striking at times.
During August, when QPR were top five, their SPI rating was 8th or 9th, albeit with a very small sample size.
In September and October, the two correlated much closer, with the team standing between 6th and 10th in the table, with an SPI of 7th to 9th.
QPR's SPI remained the same during November however, while QPR climbed the table as high as 3rd (29/11).
And this disparity widened during December and January, when the team was consistently 4th in the table, while their SPI fell to 9th or 10th.
February/March saw a continuance, with the club still in/around the playoff places, while their SPI fell to 13th.
By 20th March QPR were down to 8th in the table, while their SPI was showiing 15th and as the club has dropped to 9th and then 10th since, so has their SPI fallen, to 16th and then 17th (today's rating).
In short, while the club was flying high around the turn of the year, FiveThirtyEight suggests this was a false position, as measured by underlying performances.
Which in turn explains their recent run of defeats, as results finally begin to align closer with their actual performances (reversion to the mean). |  | | |  |
| Positions, Points, Performances and QPR's season to date. on 13:16 - Apr 13 with 3006 views | CliveWilsonSaid |
We don’t need xG websites to tell us we’re going to finish in 16th.
Where do they have us in the justice league btw? |  |
| |
| Positions, Points, Performances and QPR's season to date. on 13:17 - Apr 13 with 3004 views | T_Block |
As long as we acknowledge the issue is the attack in general and injuries and not the manager as the stats just give a suggestion the manager has enabled/organised/cajoled a team to over -achieve. | | | |
| Positions, Points, Performances and QPR's season to date. on 13:25 - Apr 13 with 2963 views | ozranger |
Interesting. Like all statistics there are questions. Take a look at the predictions for the upcoming games. Why are all games with a 25-30% chance of being a draw? That's a bit ridiculous. I generally, like many fortunately, take these sort of data with a grain of salt. |  | | |
| Positions, Points, Performances and QPR's season to date. on 14:11 - Apr 13 with 2859 views | Myke |
I am absolutely useless at analysing data, so probably making an overly simplistic (and completely inaccurate) observation here. But... if as you claim, the SPI is now more closely aligned with our recent (Okay, not so recent at this stage) run of results, than it was in January, how come the gap between our actual position today and the SPI (10/17) is actually wider than it was in January (4/10)? It seems to be lagging behind by several games and probably won't 'catch up' until the season is over, by which time I will have a decent stab myself at where we will finish! | | | |
| Positions, Points, Performances and QPR's season to date. on 14:53 - Apr 13 with 2786 views | NewBee |
| Positions, Points, Performances and QPR's season to date. on 14:11 - Apr 13 by Myke |
I am absolutely useless at analysing data, so probably making an overly simplistic (and completely inaccurate) observation here. But... if as you claim, the SPI is now more closely aligned with our recent (Okay, not so recent at this stage) run of results, than it was in January, how come the gap between our actual position today and the SPI (10/17) is actually wider than it was in January (4/10)? It seems to be lagging behind by several games and probably won't 'catch up' until the season is over, by which time I will have a decent stab myself at where we will finish! |
Fair question, but if I understand it correctly (stress the "if"), then just as a snapshot of the league table can be misleading, so can the SPI at any one point.
In QPR's case, the gap between SPI and league position is currently higher than before because you still have early season points in the bag which don't suddenly disappear, even when you're losing (almost) every game. But unless you start to pick up points again, then those early points will progressively count for less, as you get overtaken by other teams below who do manage to put a few wins together.
Ultimately the key to picking up points, whether immediately, or after a lag, is underlying performance. And right now, the latter isn't looking too good for QPR.
On which point, I am reminded of a saying in The City, which Warburton will doubtless recognise, namely "The trend is your friend".
The difference being that in The City, you can profit equally from a downward trend by selling early, as by buying early in an upward one.
Whereas in football, no-one benefits from a downward trend except the other teams. [Post edited 13 Apr 2022 15:04]
| | | |
| |
By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions.
