Review of rolling data

During pre-season I undertook an investigation into the extent to which teams’ attack and defence strength weightings fluctuated throughout last season. I found that exercise illuminating, so I’ve taken the opportunity afforded me by the latest International Break to repeat it, to see what interesting trends have emerged during the early stages of the current season.

Of particular interest to me this time around is what insights can be gained about the promoted sides. My starting point supposition is that my method of best guessing their relative attack and defence strengths overestimated how well LEE, WBA and FUL would transition to the Premier League.

First off, a quick reminder to you that I ran parallel models last season to test which number of gameweeks enabled my spreadsheets predictions to achieve the strongest correlation to actual scores, and I established that an 8 gameweek data range was optimal.  Ever since then, my spreadsheet’s predictions have been based on each team’s last 8 home or away games, whichever was relevant for the upcoming gameweek.

Every team is given a weighting for attack and defence strength, home and away, based on their last 8 home and last 8 away games. In theory, these ratios represent how many xG teams would be predicted to score or concede against an average defence or attack with a rating of 1.00. They are updated after each gameweek, and by charting the fluctuations in these ratios we can observe in which direction teams are trending.

The best to worst 8 game sequences are coloured on a scale of blue to red, making it easy to see where in the season each team’s best and worst xG form was leading into gameweeks 1 to 9. For example, the most noticeable trends in the home attack strength table below are the upward ones achieved by LEE, TOT, and AVL, and the downward ones sustained by BUR and MUN.

MUN‘s decline in home attack strength (see table above) is the biggest experienced by any team in any of the four categories. Their predicted xG vs an average defence with a rating of 1.00 has fallen since the start of the season from 1.23 to 0.73 (see graph below).

This is not encouraging news for all those bringing Fernandes into their teams ahead of a home game vs WBA, and/or planning to captain him. In better news though, as will be shown in the last of the four tables, the Baggies away defence strength weighting is deteriorating at the third fastest rate overall.

MUN decline in home attack

Meanwhile, the previous manager of MUN has TOT trending in the opposite direction.

TOT have the fasted rate of improvement in attack among non-promoted teams

The bad news for MCI, the next visitors to the Tottenham Hotspur Stadium, is that Mourinho also has his team’s defence trending in the right direction (see table below), reducing the xG conceded on average vs an attack with a ratio of 1.00 from 1.16 down to 0.80.

They join SOU, ARS, BUR and BHA as the standout improvers in this department. Conversely, SHU, FUL and CRY experienced the biggest declines.

A little over a year ago, SOU conceded 9 goals at home, and their troubles at St Mary’s over the past couple of seasons have been well documented, so the transformation this season is especially creditworthy. And as the graphs below show, the Saints’ defence has trended in the right direction overall on the road also. The rate of improvement at St Mary’s is the fastest recorded by any team (home or away), and SOU are responsible for the biggest combined (home and away) improvement.

SOU currently have the most improved defence

Given the injury problems that have plagued the CRY backline this season, it is not altogether surprising that it is the Eagles whose home defence strength weighting has suffered the most so far.

CRY defence has deteriorated more than any other

With regards to attacking strength away, it is LEI and WHU who have registered the most significant rises in these weightings.

A key finding from my review of last season’s rolling 8 gw data, was how much better the performances of LEI were on their travels compared to playing at home, and despite the absence of supporters in the stands this has seemingly continued to be the case this season, with the rate of improvement in their attack strength weighting quite astonishing.

LEI have the most improved away attack, but…

It must be remembered, however, that LEI have been awarded a ridiculous number of penalties already, and 6 of the 8 they’ve been given came in 3 of the 4 away games they’ve played. This will undoubtedly have distorted my model’s reading of the data, so the meteoric rise in LEI’s away attack strength should not be taken at face value.

The fact that the next 2 fastest improvers were promoted sides, LEE (at home) and FUL (away), is probably indicative of nothing more than my best guess weightings for them at the start of the season erring too much on the side of caution.

TOT‘s rate of improvement in home attack strength was next fastest, followed by 2 teams of particular interest to FPL managers right now, given their immediate, favourable, fixture swings. They are WHU away, and AVL at home (see graphs below). Grealish and Watkins have been amongst the fastest risers in the FPL transfer market this week, and my model’s rolling 8GW data for AVL endorses their credentials.

The standout improvers in the away defence strength department are AVL, ARS, WHU and SOU (see table below). The suggestion earlier was that the home attack strength of LEE had been underestimated prior to the season beginning. Well, the reverse would appear to be true for the away defence strength of LEE, who along with another promoted side, WBA, have experienced one of the sharpest declines.

In fact, the rate of deterioration in the weighting of LEE’s defence strength is the fastest of any team, home or away. Please understand though, this is not to say their defence is one of the worst in the league. Far from it, in fact. Indeed, there are only 5 teams with a superior away defence strength currently. Although, “currently” is the operative word, and it remains to be seen how far their fall from grace will continue.

Once again, the explanation for this contradiction is likely to do with the initial weighting my best guess method assigned them (second best in the league!) which, with the benefit of hindsight, massively overestimated them. Unfortunately, for me, I based my starting keeper (Meslier) on this wholly unrealistic interpretation of last season’s championship form. Needless to say, I’ll not be using my best guess method ever again!

There are 7 teams with a worse away attack strength weighting than WBA at the moment, but that won’t remain the case for long judging from the Baggies rate of deterioration (third fastest), and the fact they were given the artificial headstart of an overly generous initial best guess weighting.

Having highlighted the impressive improvement in SOU’s defensive process earlier, it would be remiss of me to not do likewise for ARS. The Gunners have registered the second biggest combined improvement (home and away), and the rate of improvement in their away defence strength weighting is the fastest recorded by any team.

The other 2 teams shown making significant progress in the table above are AVL and WHU who, as I stated earlier, are both about to embark on favourable fixture runs. The Villains have actually improved their away defence strength weighting by more than ARS, and are yet to concede on their travels, albeit having played one game fewer, while the Hammers deserve high praise indeed for recording the third fastest rate of improvement in this category, despite their opening four away fixtures being against ARS, LEI, TOT, and LIV!

Finally, my hypothesis that my best guess method of weighting the promoted sides’ attack and defence strengths overestimated how well they would transition to the Premier League seems to have been borne out by this exercise. An initial best guess weighting that meaningfully underestimated their capabilities was found in only 2 instances out of a possible 12. Namely, the home attack strength of LEE, and away attack strength of FUL.

I have personally found this investigation into emerging patterns very helpful in terms of crystallising my perception of where teams are at with their attacking and defending processes. I hope the same can be said to be true for you.

Eight gameweeks down, thirty to go! May your arrows be green.

Coley aka FPL Poker Player @barCOLEYna

P.s. My 6GW spreadsheets with sortable columns continue to be in high demand with customers all around the world, and are available for purchase at a ‘cheap as chips‘ cost of £2 each, or the heavily discounted price of £30 for a season ticket. Please feel free to DM me any questions you may have about acquiring your own copies, or click here to go directly to my PayPal link.

4 thoughts on “Review of rolling data

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s