The recent strong run by the Gunners has created this impression that they are a hard lot to beat. I am among the many Gooners who have noted and commented on the doggedness of this Arsenal squad. With that in mind, I thought it would be interesting to see if some numbers back the new found resolve that is being widely perceived and how other top teams stacked up when compared to Arsenal.
The stats for Points Gained From Losing (PGFL) positions and Points Dropped From Winning (PDFW) provide an interesting, if not definitive, benchmark of a team’s fighting spirit.
The following table compares the top six on these two measures over the last five seasons including the current one.
Arsenal are league leaders, including all teams and not just in the table above, when it comes to winning points from losing positions this season. And in terms of holding on to leads – i.e. not dropping points from winning positions – the Gunners haven’t done badly either as only Tottenham are above Arsenal among the top six. Chelsea match the Gunners closely on both parameters but the numbers for City, Liverpool, and United are surprising.
Simply looking at the above table it would seem strange that the team leading the Premiership and dominating most games is the one that is dropping the most points from a winning position. But instead of jumping to the conclusion that Mancini’s side is not quite as good at holding on to leads as other teams, one must delve deeper.
The explanation seems to be that City have the highest PDFW figure because they have been in winning situations much more often than the other teams. This is corroborated to an extent by looking at the sides which have done the best in terms of holding on to leads. Along with Newcastle, Bolton have dropped 0 points from winning positions. Clearly, this is not a indication of Bolton’s defensive strength but of the fact that they’ve rarely been ahead in games.
By extension of that logic, it seems safe to say that directly comparing the figures will only lead to illogical and baseless conclusions. So while Arsenal have gained more points from losing positions than any other team, we can’t really draw any meaningful insight from that alone.
In order to get some sort of a baseline I thought it would be interesting to compare the percentage of points gained and dropped. In other words, how many points were won from the total points that were being lost and how many points were dropped from the total points being won.
Mathematically, this would work out to Points Recovered (PR) = PGFL/(PGFL + Total Points Lost) and Points Surrendered (PS) = PDFW/(PDFW + Total Points Won). The following table expresses these metrics in percentage form.
Intuitively, this table makes a lot more sense. Even though City have only won three points by turning one loss into a win, they have recovered 30 percent of the points they were losing. Arsenal have a similar percentage because the Gunners have been behind in games more often.
It’s not easy to measure intangibles like the resolve shown by a football team but this table does give us some talking points.
First, let’s look at United as they have been in the top two in all the five seasons under consideration. It’s interesting to note that Ferguson’s side have rarely surrendered over 10 percent of the points they were winning. Last season was the only time they did so but that was due to the fact that most teams struggled in a highly competitive league. This resulted in the title being won with a relatively meagre 80 points. They Red Devils have also consistently recovered 25 percent or more of the points they were losing. The current season seems an aberration on that front but the number of games seems to be a factor.
Similarly, Chelsea have done well in terms of recovering points by clocking over 30 percent in four of the five seasons. Their numbers for surrendering points aren’t too bad either. Interestingly, Chelsea failed to hold on to the leads most often in the season when they actually did win the league, 2009-10. This tells us that there can be no simplistic linear correlation between these figures and the final League position. I think that particular aberration is explained by the fact that the Blues dropped many points towards the end of their League winning effort, which, indeed, allowed United to come within a point of the title. Chelsea would have run away with that title, as it appeared most of the way through the season, if they’d held on to the leads in keeping with their defending in other seasons.
Looking at other numbers in the table, it’s worth noting the outstanding recovery percentages that Arsenal (40.4) and Liverpool (44) achieved in 07-08 and 08-09 respectively. It’s not surprising that these were the years both came closest to winning the League. On the other hand, Liverpool’s form in the last few years has been poor and their rather disappointing recovery percentage neatly coincides with the drop from the top four.
Arsenal’s numbers are interesting when seen in the light of this analysis. This season the Gunners have surrendered only 9.4 percent of the points. This is distinctly better than the previous occasions. The five odd percent gap might not seem that big but one can only imagine where the title would have landed if Arsenal had kept this figure below 10 in 07-08. I believe it’s safe to say the numbers do support the notion that Arsenal are putting in gritty performances this season.
Even in the Champions League, in the eight games played thus far, Arsenal have recovered three of the six points from the only two occasions they were behind to get a PR of 50 percent. That includes a defeat against Olympiacos which came with the second string. And Arsenal have dropped only two points while winning 17 (including the qualifiers) for a PS value of 10.5 which is consistent with the League figure.
On the whole, I am convinced the Gunners are fighting harder in defence. It would be a stretch to label some of the performances as defensively solid but the number of mistakes, particularly of a casual nature, have gone down significantly. The defensive shape of the team has also been better largely due to the work rate and positional intelligence of the midfielders. Sustaining and indeed improving on this through the course of the season will be vital.
That won’t be enough though, and Arsenal must reduce the number of times they’re going behind. In the fifteen League games, Wenger’s side have conceded a lead on 6 occasions. That proportion has to come down or the percentage of points recovered will have to reach the 07-08 levels in order to have any title hopes or even to cement a position in the top four.
The Cup ties will be knockout games so percentages will mean little but tenacity and the ability to recover from losing positions will be needed in order to progress to the later stages.
There can be an argument against the way I have used the numbers. Some might say including the total points lost, for instance, brings in the element of drawn games where a side wasn’t behind so it might be better to include only points gained and lost from losing positions. That would certainly be another interesting way of looking at the numbers and probably not the only one. I don’t think this analysis conclusively proves anything but there are sufficient indications that justify an optimistic view towards the rest of the season. Revisiting the numbers at the end of the season will tell us more.