By The Numbers: Shot Quality and League Position

youtube

Shots on target: if you’re doing them, they’re good; if they’re the opposition’s, they’re not. It’s a pretty straightforward assessment of one of the simpler data sets you can find. A team that registers a lot of shots on target not only has a greater chance of scoring, but it also suggests that their build-up play and/or decision making might be better, because they’re shooting from better positions or working the ball into good positions in the first place; you also need the ball to be able to shoot on target in the first place and, as we saw in my last column, there is some correlation between possession and position in the table.

This season, Liverpool (131) and Tottenham Hotspur (124) lead the way in shots on target (SOT) so far, followed by Manchester United (120) and Chelsea and Manchester City (both 109). At the other end of the table, Middlesbrough (50), Sunderland (58), and Burnley (61) have registered the fewest.

Conversely, giving up SOT suggests that a team’s midfield is porous, perhaps that the defence is disorganised, or the goalkeeper is not in command of what’s happening in front of him; at the very least, we can safely say a team that is giving up a lot of shots on target isn’t doing very well.

Unsurprisingly, the sides that have faced the most shots on target are Sunderland (130), Hull City (121), and Burnley (119). The tightest sides are Chelsea (50), Spurs, (57), and Southampton (59). Of course, a shot on target is not a problem if your goalkeeper is in fine form, and it’s worth noting that for proportion of save/shots on target, Ben Foster’s West Bromwich Albion come top of the league with 76.1% (70 saves of 92 SOT), with Tom Heaton’s Burnley (with a little help from Paul Robinson) in third with 73.1% (87 saves of 119 SOT); Hugo Lloris/Michael Vorm’s Spurs are in second, in case you wondered. At the bottom of the pile are Southampton with 54.2% (32 saves from 59 SOT), Crystal Palace with 58.6%, and Swansea with 59.8%. It is instructive to see that two of those sides are in trouble, perhaps unexpectedly in both cases given previous seasons and goal scoring ability, and one (Saints) are under-performing against last season; goalkeeping is important.

But, shots on target and shots on target faced are not, in themselves, clear guides to league position. The table of SOT faced shows the sides in 12th and 13th in the league are 17th and 18th in the table for SOT faced (with 20th being worst); clearly those sides are compensating some other way for giving up a load of shots. Southampton are 10th in the league but have given up the third fewest shots on target; they must be deficient in other areas (and one we’ve already noted above).

Similarly, while Liverpool and Spurs are first and second in the shots on target achieved table and second and third in the league table, first place Chelsea are only fifth for shots on target achieved, while eighth place West Brom and 12th place Burnley are right down the bottom, in 17th and 18th for shots on target managed. And relegation-threatened Swansea, 19th in the league, have managed 85 shots on target, putting them 9th in the list.

A ratio of shots on target achieved to shots on target faced might be a clearer indication if we were a looking for some kind of overall measure of how shots relate to league position, a balance of attacking and defensive ability. But, this still doesn’t quite work: Chelsea top our list with a SOT/SOT faced ratio 2.180, with Spurs (2.175) and Liverpool (2.079) next, in a reversal of league order. But Arsenal are down in seventh, despite being 5th in the table, behind Southampton, whose superb SOT faced counts over-heavy in their favour. Similarly, Burnley are 19th in the ratio list but 12th in the table, and West Brom are 17th in the ratio list but eighth in the table.

The missing factor is shot conversion, which I use including blocks (because not doing so doesn’t make sense). Southampton’s Shot Conversion Rate (SCR) is a measly 6.51%, the lowest in the league by about 1.7% (so much for not needing an out-and-out striker, eh?); West Brom’s is a superb 13.79%, second only to Arsenal this season (the Gunners are on 14.43%, which is very, very good. This is the factor that accounts for how a team who are defensively sound and creating chances, but not scoring them, can be so ranked highly when only SOT faced are looked at (Saints), or how a team that is defensively sound but doesn’t create much (WBA) is doing so well in the actual league table.

If, then, we factor in SCR, we can devise a little equation. Let’s call the value Q, for quality, because that’s a nice, vague word. If we say Q = (SOT/SOT faced) x SCR, we get something pretty interesting:

Team                       Q value         Actual league position
Chelsea                       30                             1
Liverpool                   28.4                          2
Spurs                          23.4                          3
Man. City                   22.8                          4
Arsenal                       19.9                          5
Man. United              17.6                          6
Everton                       14.5                          7
Bournemouth            10.7                          9
Southampton             10.4                        10
Crystal Palace            10.2                        17
WBA                             9.6                           8

The top 11 places, when ranked by Q, correspond up to seventh, and only one side not in the top 11 has a Q value in the top 11. That, incidentally, suggests that Palace have a factor affecting their position that’s not included: that could be goalkeeping errors, as noted above, or bad luck (yes, that happens in football); they are a bit of a mystery, perhaps. Otherwise, including SCR irons out issues by factoring in whether SOT volume actually achieves anything (Saints waste theirs, WBA over-capitalise on far fewer opportunities).

Lower in the table, things are a bit more confused: while Sunderland and Hull occupy the bottom two places for Q value and are 18th and 20th in the league respectively, West Ham are 18th for Q but 13th in the table. Similarly, Swansea, 19th in the table, actually have a Q value that puts them 13th: their SCR of 9.96 is a major factor, again showing why playing Sigurdsson as a false nine in key games possibly wasn’t the best plan, and they give up more SOT faced than they probably ought to. Nonetheless, the bottom five Q values are Burnley, ‘Boro, West Ham, Hull City, and Sunderland, probably a pretty fair reflection of actual quality, with Burnley overachieving (in large part because Tom Heaton keeps out so many shots) and West Ham an oddity (along with Palace and West Brom, possibly the hardest side to work out).

Having run a quick correlation on Q for league position, it shows a strong correlation as well (R=0.881, if you’re interested).

The Q value also points to where teams can improve: Palace need to give up fewer shots (and save more, probably); Saints need to do better with their chances. It also suggests that West Brom are overachieving with an unusually high SCR, which you might expect to regress to the mean and West Ham are over-performing given the numbers for reasons I am, as yet, unable to explain.
Nonetheless, if you want a decent, straightforward way of working out roughly how good your side are relative to their league position, the Q value works. Quite what we do with that next, remains to be seen.

Alex Stewart on twitter
Alex is a Southampton supporting writer based in north London. He writes about sports and coffee for a variety of magazines and websites. He likes hermits, reading, drawing, forests, and graphic novels.

He creates By The Numbers and provides scripts for the Whiteboard Football: Tactics

Share this:

22440

Categories:

Tags:

2 Comments on “By The Numbers: Shot Quality and League Position”

  1. Tom says:

    This is fascinating, as always. I haven’t had a chance to sit down and think this through properly, but I had a glance at the numbers and came up with something that potentially explains the West Ham anomaly, and also highlights why WBA underperform a bit for Q/ why their SCR is so high.

    “Shot on Target” is obviously a fairly broad term; 30-yard potshots that trickle through to the keeper are measured the same as a 3-yard tap-in. It seems to me that the type of SoT you’re producing is important as it affects how efficient you are under the SoT/SoTFaced x SCR formula – if you had, say, a player who takes lots of long range pings alongside the more effective stuff you do, it would distort the overall value.

    So, I had a look at headers because, by their nature, they appear to be good quality chances as hitting the target with one necessitates being in close proximity to the goal. West Ham’s general SCR is pretty bad, but if you break down their conversion rate for different type of shots, it’s quite revealing. They convert about 15% of their headed chances vs ~6% of all other chances (total shots – headers). Just over 20% of their total shots were headers, which is comparatively high for the league, scoring 11 goals from them (almost half of all their goals this season). However, they take a lot of non-headed shots – which they’re horrendous at scoring from – which skews their SCR and could possibly explain why they look bad in terms of Q? They create quite a lot of chances they’re good at scoring from, but also take many shots they’re bad at scoring from.

    Whereas, nearly 30% of West Brom’s total shots have been headers (only marginally less actual headers than West Ham’s – 56 to 58) and they’re even more effective at converting them, with just under 20% of their headers result in goals. Their SCR for non-headed shots is twice as good as West Ham’s at 11% but they don’t bother taking a high volume of those types of shots – obviously they’ve clocked they’re good at headers and don’t try to make as many alternative chances as it’s not worth focusing on. It seems like that low total number of shots is what gives them an overall lower number for Q, while the fact that a high proportion of the shots they take are the sort that they’re effective at taking explains their high SCR.

    Interestingly, given the players at their disposal, 18% of Southampton’s total shots have been headers but they’re awful at converting them (~6%). That might go some way to indicating why their shot conversion rate is so poor and explains the discrepancy between their fairly high volume of shots and their lack of goals: they’re creating a lot of the type of chances that they’re not very good at scoring/ that they’re not properly equipped to convert.

    Anyway, that’s quite jumbled and confused, but I think there’s something in there, especially on the West Ham front. Might be worth considering in the future, or at least bearing in mind when thinking about this sort of thing.

    1. Alex says:

      That’s really interesting, thank you. I think you have a strong point re WBA and West Ham.

Leave a Reply