Celtic Stats Corner: Should Craig Gordon or Dorus de Vries be the Starting Keeper for Celtic?

By foxsports
Published

Celtic goalkeeper Craig Gordon has had quite the roller coaster ride during his time at Celtic. When he first was brought in, most doubted the Scottish keeper’s ability to stay healthy for a full campaign (not to mention some hard feelings about his willingness to come up for a corner against Celtic while at Hearts).

However, in that first season, Gordon proved that he could stay fit and was one of the best players for Celtic that year, making critical saves in both the domestic and European campaigns. Gordon put up a save rate of 83% that year and won the Scottish Writers Player of the Year award, helping Celtic to the SPFL Premiership title.

Last season, the general consensus from Celtic supporters was that Gordon’s form dipped. This belief mostly comes down to some high profile mistakes Gordon made in European matches that cost Celtic points and possible Champions League group qualification and the knockout stages of the Europa League. Yet, despite the perception that Gordon had a bad year last year, he ended the season with a 74% save rate, one of the best in the league.

More from The Celtic Bhoys

    Both years, Craig Gordon was among the best keepers in terms of save rate in the SPFL, yet it has been shown that save rate is not the best way to judge a goalkeeper. Save rates can be volatile and bounce up and down throughout a year. For example, in European campaigns Craig Gordon had a save rate of above 80% in 2014/15 and above 40% in 2015/2016. Was Gordon that significantly worse in the two years?

    There have been a few attempts to find more effective stats to judge a goalkeeper’s performance. Those numbers use the “expected goal” stat that we have for attackers and apply it to keepers, coming up with “Expected Goals Against” (xGA) and “Expected Saves”. Based on the shot location and type, these stats come up with how many more or less saves a keeper should have and goals he should allow.

    I have previously used those stats to judge the keepers of the SPFL Premiership and found that last year Craig Gordon both the highest amount of saves above expected and the best actual goals against minus expected goals against in the SPFL. He also had the best “Goalkeeper Rating”, which is the number of saves made in a season divided by the number of expected saves multiplied by 100.

    When Brendan Rodgers was named the manager of Celtic this summer, the rumor persisted that Rodgers was looking for a keeper that was better with his feet than Gordon. The stats clearly show that Gordon is performing up to the standards one would have for a Celtic keeper. In fact, Celtic historically has struggled to find a consistent keeper and have only recently found quality keepers with names like Artur Boruc, Fraser Forster, and now Craig Gordon.

    The idea that the keeper needs to be competent or better with his feet has become a popular one recently. Tom Payne discussed on The Ringer about Pep Guardiola unceremoniously loaning England’s #1 keeper Joe Hart to Torino to make room for Claudio Bravo at Manchester City. Payne described Guardiola’s ideas for the keeping position as:

    To implement his pass-first, possession-dominating style, Guardiola demands a strong passing ability from every player, including the goalkeeper. His Bayern team averaged 66 percent possession in the Bundesliga last year, and when you pair that with a well-coordinated pressing system that often wins the ball back as soon as the team loses it, the goalkeeper’s net is rarely tested. (Bayern conceded the fewest shots in Germany in each of Guardiola’s three seasons at the club.) It makes sense for the Guardiola keeper to be used in other ways; otherwise, for large parts of each game, he’d be no different from a spectator who’d wandered onto the field.

    You can see the similarities to Celtic in that quote about Manchester City. Celtic often enjoys much of the possession in a game in the SPFL. Celtic conceded the fewest number of shots last season. It makes sense that if Celtic’s keeper does not have as many shot stopping responsibilities as an average keeper in the SPFL, he should be able to bring something else to the team when on the pitch.

    This is all in addition to the free flowing, high-octane attack that Brendan Rodgers wants to bring to Celtic. So it was not a surprise when Celtic signed Dorus De Vries. De Vries was promised to be better on the ball than Craig Gordon and the perfect fit for Rodgers’ system.

    Above are the stats for all the keepers in the SPFL thus far. It is still very early, so I would expect many of those keeper’s ratings to regress back towards the 100 range, which is “average”. However, you can’t help but notice Craig Gordon and Dorus de Vries’ numbers in similar minutes played this season. Gordon has easily the better numbers.

    Yet, like we would expect the keepers from the rest of the league to regress, we can expect de Vries’ numbers to improve as well. De Vries has only faced 3 shots on goal, saving one, in his 188 minutes. I can only think of Rangers’ only goal as one that I thought he could do better on for the 2 goals he has conceded.

    Craig Gordon might feel hard done to lose his starting job after putting up two seasons and two games worth of good stats. However, if Dorus de Vries can put up average shot stopping numbers and improve distribution, he could bring another element to Celtic and further improve their already potent attack. If de Vries can’t raise his shot stopping

    Want your voice heard? Join the The Celtic Bhoys team!

    This article originally appeared on