Today's transfer nonsense: Vermaelen for Smalling

Let's take a brief break from the World Cup to talk about a transfer rumour involving two players currently playing (or riding the pine) in Brazil.

It has recently been suggested that Manchester United boss Louis van Gaal has decided to bid for Arsenal defender Thomas Vermaelen. The Belgian, who had a fine start to his Arsenal career, is now Arsene Wenger's "emergency defender", used most often when Laurent Koscielny, Per Mertesacker or Kieren Gibbs are unavailable (for Gibbs, read: injured).

At 28 and fending off Jan Vertonghen for his place beside Vincent Kompany at the heart of Belgium's defence, it’s logical to assume that Vermaelen wants first-team football; United certainly have money to burn that may raise his earnings above what the Gunners are willing (or able) to offer.

However, the rumour suggests Vermaelen’s current manager Arsene Wenger wants a return that includes a transfer fee and one of Tom Cleverley or Chris Smalling.

While possible, it’s unlikely that van Gaal would sanction such a move for several reasons. Both Smalling and Cleverley fit into the coach’s modus operandi as young players malleable to his methods; the Englishman is also four years younger than his supposed upgrade. Smalling is also United’s emergency right-back, while both midfielder and defender are English and therefore come at a FFP-induced premium.

While data rarely drives a transfer, especially in the case of Louis van Gaal, let's examine why the numbers don't like a Smalling-for-Vermaelen swap straight up – let alone with a transfer fee involved. Most of the data that follows comes from a pilot project that tracks each player's individual plus/minus throughout the course of the season. How each player’s team performed during those minutes give us an idea of how he compares to his team (and teammates).

Over the course of 2013-14, Chris Smalling played nearly 2000 minutes (1938) in 25 matches; Vermaelen managed 781 over fourteen in an injury-interrupted season. While this presents a difference in sample size, we can still draw enough information from the time each player spent on the pitch to compare them to their teammates.

Over the course of the season, United scored 64 goals and conceded 43 - or, 1.684 (scored) and 1.161 (conceded) per game. With games running to 90 minutes and players often not playing the whole game, this total can be compared to the rate of goals scored/conceded when player is on the field for 90 minutes.

You can see in the table below that in 2013-14 Chris Smalling was about as average a Manchester United player as you could find; when he was on the pitch, United scored slightly less frequently and conceded slightly more often than over the course of the entire year.

Player	G	Min	GS	GA	GD	Scored/90	Conc/90	GD/game	+/- per 90
Smalling	25	1938	35	25	10	1.625	1.161	n/a	0.464
Man United	38	3420	64	43	21	1.684	1.132	0.553	n/a

This resulted in Smalling's an individual plus/minus per 90 minutes of 0.464, slightly below United’s average of 0.553. In short, Chris Smalling – as a Manchester United player in 2013-14 and therefore beset by the utter lack of identity the team had under DavidMoyes – made his team slightly worse by his presence on the pitch.

The same cannot be said for Vermaelen. The typical Arsenal team of 2013-14 conceded 1.079 goals per ninety minutes in 2013-14 while scoring 1.800. Vermaelen's numbers are comparatively dreary; the Gunners scored only 1.382 and conceded 1.729 goals per 90 with him on the pitch. His personal plus/minus per 90 minutes stands at a catastrophic -0.346 – over one goal per game worse than his club plus/minus per game (+0.711).

Player	G	Min	GS	GA	GD	Scored/90	Conc/90	GD/game	+/- per 90
Vermaelen	14	781	12	15	-3	1.382	1.729	n/a	-0.346
Arsenal	38	3420	68	41	27	1.800	1.079	0.711	n/a

Not only were Arsenal worse off last year when Vermaelen took the field, they were almost catastrophically so. Selection bias can be blamed in part, as Vermaelen's subset of matches included multiple goal losses against Manchester City, Chelsea and Everton. This is partly balanced as these figures are also obviously reflected in Arsenal's totals. However, we can see here that Vermaelen made Arsenal a more porous defensive unit than his peers, while Smalling had only a small role in reducing United's redoubtability.

Numbers only tell half of the story, especially in this case. Figures can't account for a player’s ability to adjust to different styles of play, their locker-room leadership or a player's relationship with his coach. The stats used in this case aren’t able to necessarily predict the future but inform as to past performance. And these numbers are (relatively) clear-cut: in 2013-14, Chris Smalling was a better fit for United than Thomas Vermaelen was for Arsenal. Given the other intangible factors at play, the Red Devils should be extremely hesitant to countenance any kind of a swap deal with their re-tooled rivals.