Despatches From The Underground Glove Fraternity

Alex Smithies has not been Town's regular goalkeeper for a long time now - since about this time last year. I was, initially, sceptical about Ian Bennett's ability to deputise long term, but he seems to be entrenched in the shirt, and fair play to him for making it his own. My favourite feature about Smithies, and if you watch him play, you'll notice it pretty much straight away, is his speed of distribution. When he gathers the ball, he gets rid of it very, very quickly indeed, and positively, too, to put Town on the front foot. This, in itself, was a key feature of a game I've come to regard as being a watermark of Huddersfield Town under Lee Clark - the 6-0 win against Wycombe. This distribution is not something that Ian Bennett has about him - which is perhaps because he did his goalkeeping apprenticeship in an era when the number ones had the ball at their feet a lot less often than today.

There's been a lot made of Victor Valdes' performance in the last Clásico between Real Madrid and Barcelona (Barcelona won it 1-3, at a canter in the end - Alcântara, if you will) and the way he interacts with the team. Valdes, despite a calamitous error, gifting Madrid possession and the first goal, continued to play football alongside Barcelona - distributing generally to Puyol, Pique and Abidal - indeed, he has the highest pass completion rate of any goalkeeper in La Liga (85% after the Clasico), and a fair few outfield players, too Barcelona build their attacks from the back, as I'm sure you've noticed. It is unrealistic for Huddersfield Town to play like Barcelona, but I'd like you to remember the points about Victor Valdes.

My favourite Premiership player, however, wears the number 25 shirt, and hangs around between the posts on Merseyside. Jose Manuel (Pepe) Reina. I could watch Pepe play in goal all day. I accept that he does make mistakes with his handling - I would say that this is generally when the ball is in the air, something more dominant - but I would contend that his distribution would, in a normal team (rather than the currently profligate Red Men) more than balance that out (after 11 games it was 64.04% - 2nd to Petr Cech). I would imagine that this ability to turn defence into attack (he has more assists - 4 - than any other Premier League goalkeeper since 2006, and created had created 2 goalscoring opportunities by game 11) explains a lot of the effectiveness of Glen Johnson - not as good as Martin Kelly, though - and Jose Enrique (fantasy league gold) in the Liverpool team. Both players often operate as auxiliary wingers, and it allows Liverpool to play with narrower forward players; Kuyt, Bellamy, Maxi Rodriguez, so there's more bodies in the box to score,- just as Barcelona can use Fabregas, Xavi, Iniesta, Villa, Messi, Pedro, Alexis Sanchez, Thiago... because they know Dani Alves and Abidal will be bombing forward, getting the ball from Victor Valdes.

I'm sure you can see where this is going now. With Ian Bennett between the sticks, attacks are built up a lot differently to having Alex Smithies keeping goal.. I would assume, from what we know, that if Alex is giving the ball to his full backs closer to goal, it will lead to Town having more possession, more players in the attacking third, and more attempts on goal per game than if Ian Bennett is getting rid of it in a more industrial way. Because none of Jack Hunt, Gary Naysmith or Calum Woods is Dani Alves, I would also expect them to give the ball away when they have it from time to time - something that wouldn't happen if they were bypassed. I noticed it, particularly, during the Wycombe game when the cameras were on Ian Bennett as he cleared the ball. It was a punt forward, and he often turned his head away from it once it had left his boot. There was a couple of times he looked as if he might roll the ball out, but then didn't, like he was just waiting for people to get further upfield.

This is, as the graph below shows, exactly what happened. Ian Bennett may be consistent over the two seasons in this respect (11.36 & 11.71 shots for Town on average) and facing around 9 - more last season than this, which suggests the defence has improved, but Smithies shows improvement in all of his three seasons so far, leading to a peak of 12.86 in the games he played last season, although the same is true of his shots faced - reaching as high as 10.82 shots against him last season. I've highlighted the two sets of bars that are 'competing' against each other, as I have for all the graphs you'll see today.

There's an element of Charles Reep (or Graham Taylor, if you take it further along that path) to that attitude - the further the ball is away from your own goal, the less likely you are to concede; so the quicker it gets there, the better. This is balanced by the argument that you cannot score without possession of the ball yourself, so keeping it, even if 20 yards behind where it might be, is better than allowing a 50/50 challenge for it, when the opposition could just as easily end up with the ball. Huddersfield Town, more often than not, have more possession than their opposition. Sometimes it feels like they're trying to stifle the opposition, strangle them at source, and nick a 0-1; sometimes it feels like they're a blue and white tide that the opposition's goalkeeper's attempts to stop are as futile as that of King Canute. Both are pretty good ways of gaining points.

I would expect, again, to see Town's opposition having their increasing amount of shots, though, from less possession. All this would lead to the following. Working according to my theory, games in which Alex Smithies plays instead of Ian Bennett have more goals, for and against, and as a result, more decisive outcomes (wins either way), and fewer draws. It doesn't take a large leap to note that that is, very much, what has happened over the last season or so - Town lost games in the Autumn of 2010, when Alex was playing, but won them too. The common reason people give for the failure to get promotion is the drawing of too many games in the unbeaten run - something that has continued this season - all when Ian Bennett is playing. So far so good.

You know me well enough to know that I'm going to add some accurate numbers to back up/disprove my theory. So let's have a look. Firstly, I was surprised to find, during my research, that Alex and Jack Hunt have played only four games together. Those games ended with a record of W3 D1, F4 A1 - not bad by anyone's measure; that's 2.33 points per game. Four games is nothing to go on, mind you, so here's the numbers in a little more realistic format.

I've taken (as the previous graph showed) the league games played under Lee Clark, and divided them up, season by season to allow trends to appear - in total, Alex Smithies has made 93 appearances under Lee Clark, Ian Bennett 39 and Simon Eastwood 1 (a 1-1 draw). All Bennett's have been the last two seasons (24/25), and Smithies' games are split 25/46/22/0). Last season was the only one the goalkeeping time was split (so far) so that will prove the most interesting (except as to seeing trends away from how each goalkeeper performed), like we can see Smithies development above.

Look at the next graph, representing the goals scored and conceded while different goalkeepers were playing. Yes, Simon Eastwood's game looks deeply impressive, but there's some more to be gained from that - particularly the two highlighted parts of the graph. Alex Smithies, last season, is slightly higher than Ian Bennett in terms of goals scored (and I mean slightly, 1.68 v. 1.67) which is far closer than I would've expected; and a fair bit lower than the 1.78 of the season before, and the 2.04 of this season. This only serves to point out how last season's Town team were a little more reluctant to attack - its also pertinent that Smithies didn't feature much (8 games) in the unbeaten run, so his figures are bound to be skewed a little against him. It also goes to show that Town are less likely to concede when Bennett is between the sticks than when Smithies is - and Smithies' figure has been around 1.3 goals allowed every season he's played. It would be very interesting to see how that changed if and when he got a chance this 2011/12 season.

Smithies this season, would be playing behind a defence far more amenable to having quick possession (Jack Hunt) and doing something with it than before,

As you'll remember, I brought those numbers out to indicate how likely drawing was (given the 'We'd rather Win 2 Lose 2 than Draw 4' argument - which is completely valid in terms of points, but I would question it in terms of momentum) when each goalkeeper was playing - disregarding whether the non-drawn games were won or lost.

Conclusive, I think we can say, is the best way to describe this graph. Alex Smithies rarely draws games. Pity poor Simon Eastwood, with his 100% ratio, but its noticeable - and falls very neatly in with what I predicted - that Smithies draws far far fewer games, and always has, than Ian Bennett does. While the team are more likely to win than lose, this is a good thing - not drawing brings more points (because it means more wins), but when the team is more likely to lose than win, the opposite. This is another trend that would be interesting to follow if and when Alex gets back into the team.

Which leads me to the point of possession. Town, as I've mentioned before, average more possession than their opposition (77% of the league games under Lee Clark, 110/143 - maximum 67%, minimum 39%) - under Lee Clark, the average is 52.34% possession for Town, though it is as high as 53.88% this season - as evidenced by the graph below. When playing with the same team (i.e. last season) Smithies has by far the edge - indeed, look at his three columns there; Town average about 5-6% more possession than their opposition every season Smithies keeps goal - however the team are performing; keeping the ball, as I predicted before. This is, really, where we might see a benefit of Smithies and Jack Hunt playing together - it was a promising 51.5% average for the four games they played together last season, lower than Smithies' overall average, but up on the 49.8% he averaged in the spell he kept goal over Christmas; add in Calum Woods instead of Gary Naysmith, and there's plenty of options for the game to be built from the back. The prospect excites me.

I don't quite know what to make of the figure Ian Bennett has attributed to him this season - when bearing in mind that Town have already been to (deep breath) Charlton (48%), Hillsborough (41%), Bramall Lane (54%), Stadium:MK (53%), Griffin Park (56%) - away games at the big boys; I can only see that figure improving; there's a 2.5% average increase between away games and home over the 143 games (53.63% v 51.13%); that can only mean good things in terms of the second half of the season.

All of this leads me to one final point (and I ought to point out at this juncture that the number I am using came from ESPN's game reports - counting the saves that each goalkeeper made in each game) about which goalkeeper is performing better in terms of stopping shots. During the Sheffield Wednesday game, it looked like Ian Bennett was allowing every shot taken at him into the goal (he didn't, of course, but he was only recorded making one save), which is what prompted this post in the first place. Ian Bennett has, to my eyes, been getting rave reviews for his shot-stopping, and well-deserved, too. Having only seen him once this season, at Brentford (and now at Wycombe) he wasn't particularly over-employed, and I wondered if it represented a decline to any extent.

I wasn't entirely sure how to work this out, but I figured the best way was to balance the amount of saves a goalkeeper made with the amount of goals they let in - not all shots are created equal, by any means, but a ratio of shots that needed saving against those that did but weren't should allow us to see which keeper is making more stops.

This graph came as a pleasant surprise to me - perhaps working further against the defence this season, in a way that I wasn't quite expecting. If I use Simon Eastwood's 1 appearance as a basic example - he made 6 saves for every 1 goal he allowed (6 saves - 1 goal, in fact, in that 1-1 draw with Orient) - so you can see how impressive Ian Bennett was towards the end of last season - averaging 5.63 saves for every goal he let in. To me, this is indicative that this season, with the same personnel in goal, as much as his form might have dipped, the quality of shots - given there's actually fewer shots allowed (9.75 - 8.36) at Ian Bennett this season - is better, so the defence is letting the opposition get into better positions before they get their shots away. Part of me wonders whether this is the flip side of having such a progressive full-back as Jack Hunt playing - there is that much more space behind on the right flank, or Scott Arfield covering that spot means there's a body missing in the middle of the park. What it does indicate, however, is that when Alex Smithies does return, the amount of saves he makes will be consistent with what we have seen so far this season, and if that possession figure does increase, there'll be fewer of them for him to make, so Town should, I would reckon, concede less goals - albeit a pretty small amount fewer.

What conclusions can I draw from all of this, then?

It is as I thought in terms of Town's goalkeepers - Smithies playing means more possession, and a little more likelihood of goals being both scored and conceded, but less chance of a draw - the difference between the two is small, though. It also suggests that Ian Bennett WAS in very good form last season, and would have been - probably - the right man to play in goal regardless of Smithies' fitness. If Alex comes back and continues to develop the way he has done thus far, though, we will be looking at him being superior to Bennett by the end of the season. I say 'Give Smithies the shirt, give Hunt the ball, and play at the opposition'. The hardest away games are done, and there's nothing to fear from the rest. Aye, we might lose a few, but it'll be more exciting to watch and more rewarding at the end.

Oh, and I am aware that there are countless other factors that affect all these things; not least team selection and formation, but they're dealt with elsewhere. This is strictly comparing what happens with the different goalkeepers.