Celtic are conceding less goals than last season, averaging 0.74 per game versus 0.89 last season. Many factors to this, but how is Gordon contributing to this improved defensive performance?
Saves
In the mainstream like Sky Sports the analysis of goalkeeping performance is restricted mainly to Save % - i.e. what percentage of shots on target are saved. Unfortunately, this tells you little about the quality of the goalkeeper. This is explained very well in this article by 11tegen11. I can illustrate further by showing you Gordon’s save % record over the last five seasons.

You’d conclude that Gordon is 1. inconsistent and 2. in some sort of terminal decline. As a shot stopper our eyes tell us this is not true.
And this can be illustrated by comparing the saves Gordon makes with the shots the opposition are taking. Firstly, by sheer volume.

This shows that the number of saves Gordon is making per 90m is broadly in line with the number of shots the opposition are having on target. That is, he is consistent. And by the way, it is encouraging that the trend for opposition shots on target is reducing.
As mentioned in the 11tegen11 article we need to look at Expected Goals (xG) against versus actual goals conceded. I only have this data for the last 2 seasons thus.

Last season Gordon conceded 42 goals. The xG models suggests he should have conceded 41.39. This season the respective numbers are 27 goals against and 29.49 xG against.
Broadly Gordon is conceding the volume of goals you’d expect given the quality of the chances the opposition have created. Now you might wish that Celtic, as the dominant team in Scotland, has a goalkeeper who is conceding less than expected as De Gea (Manchester United) and Allison (Liverpool) do. But what you cannot say is Gordon is under performing in the shot stopping department.
Distribution
The other major component of goalkeeping is distribution. If you are a bit old school, and let’s admit it, who isn’t, you are thinking “Goalkeepers are there to save shots not play like bloody Beckenbauer”. Admirable sentiment though that is, the reality is that whilst a ‘keeper may have to make 2 or 3 saves a game (if they play for Celtic) they have to attempt 20 – 30 passes. For a possession-based strategy, that is key. The first pass of the attack.
I described at length how Gordon had successfully adapted to Rodger’s requirements for a play making ‘keeper in Old Gord, New Tricks. That was January 2017 (it’s a super article – read it)! So, time for a refresh.

Gordon’s distribution has been on an upward trajectory under Rodgers. Last season he managed 80% pass completion and managed the most passes of his Celtic career per 90m – 27.01. This season both the % complete and volume of successful passes have dropped off.
Additionally, the number of forward passes that take out opponents has dropped from 2.38 to 1.63 per 90m.
In short, Gordon’s distribution is regressing.
But against what benchmark? Well, the highest rated goalkeeper in the English Premier League currently is Allison Becker of Liverpool. He is particularly revered for his ability to play out from the back. On WhoScored, Allison’s equivalent stats are 26.7 passer per 90m completed @ 79.3% completion. On that basis. Gordon has equivalent numbers albeit against a much lower standard of opposition.
Summary
Gordon is saving the shots expected of him (arguably, no more). His distribution is regressing but using the bench mark of England’s top passing ‘keeper, he is performing at a similar level against poorer opposition.
It is another position in the squad requiring succession planning. But there is life in the old Gord yet.