In the aftermath of the AZ Alkmaar first leg, Celtic’s home-grown full back, Ralston, continued his “local underdog comes good” narrative arc with a performance that included some notable interventions.
Bad Habits
On The Huddle Breakdown last week, recorded after the AZ game, (follow the link to listen, and please subscribe to the channel – the other guys are excellent!), I fell into a trap that I usually try and avoid.
Namely, I had, on the pod, stated roughly along the lines of how wonderful Ralston had been versus how awful Taylor (and Starfelt for that matter) had been.
This binary conclusion without recourse to studying the 90 minutes of performance data reminded me of one of the primary reasons I started Celtic By Numbers.
Binary Analysis
Despite what some may have you believe, when watching any Celtic match, I don’t have a spreadsheet open! I have the same rational/ irrational, emotional, excitable responses to what happens on the field as most others. The second or third watch is when the data is collected and calm reflection applied.
My take on watching the AZ Alkmaar tie was broadly:
Ralston = last ditch intervention - wow, taking players on up the line, headers, challenges = brilliant
Taylor = continually giving the ball away, losing it in his own third, too small, too slow = rubbish
Because, like all of us, I cannot process 100 actions x 22 players and remember all the contributions. When watching a match and thinking back on it you form your own Match of the Day highlights reel. And that becomes your reality.
I started Celtic By Numbers as, when studying my sons team, it became obvious this brand of analysis served virtually no benefit to anyone because you were remembering a tiny % of a performance and ignoring all the other stuff that didn’t involve e.g. goals or mistakes.
So, I started to capture all the players’ actions. Painstaking but doable.
Back to Ralston and Taylor.
Ralston and Taylor
Spoiler: I am not going to conclude Ralston had a poor game, because he did not. But it had many facets.
His Defensive Action Success rate was 50%, not terrible, but ”ok” for a full back (Taylor’s was 55%). But he was error free and cleared the ball 9 times, the most of any Celt.
Ralston had 1 key defensive save (the second half intercept) and 1 error (giving the ball away in a dangerous position in the second half), whilst Taylor had 1 error (giving the ball away in a dangerous position in the first half). These rare events have a huge bearing on perception.
He was 2nd in the team with 19 recoveries. Taylor managed 15.
He was also second (again to Welsh) with 6 occurrences of wining the ball back in his own defensive third. Taylor did that twice.
Ralston completed the least passes of any defender (40) and his pass success rate was 80%, 1% “better” than Taylor who “continually gave the ball away”. Taylor completed 52 passes.
Ralston had 2 Key passes (passes that led to shots) whilst Taylor had 1.
But Ralston only completed 5 Pack Passes (forward passes taking opponents out the game) – the lowest of the defenders by 7. Taylor had 14 – the joint most with McGregor.
Ralston excelled with 6 progressive runs, the most in the team. Taylor had 0.
Neither had a shot at goal but both received forward passes that took out opponents – Ralston 9 times and Taylor 12.
In the end, Ralston’s xSC (expected scoring contribution) was 0.56 compared to 0.05 for Taylor.
But Taylor had the highest packing score in the team at 117 – 20 more than McGregor and 28 more than Ralston.
Ralston was also booked.
Summary Data

Performance Summary
Ralston’s performance was by any measure very good. But it stood out because of moments. The goal saving intervention, the runs into the box to hit the by line. His “moments” tended to be positive.
Taylor’s “moments”, the ones you keep in your internal Match of the Day reel, were negative. Giving the ball away under no pressure, losing challenges in his own defensive third.
These were all seconds of a 90-minute performance. In the second half, Ralston completed 6 open play passes and gave 6 away. Yet Taylor was 18/7 and much more involved.
Taylor’s “rest of his performance” meant he was the Celtic player who most drove Celtic up the field either through passing or receiving forward passes. He had the highest Usage Rate of any player – 14% - meaning 14% of all on field actions for Celtic players were with him.
That wasn’t the conclusion I formed on first watch and communicated on the pod.
Which is why I wanted to catch myself on, but also hopefully provide an insight into why ALL the data points matter to some extent and that ALL performances have nuance that we (I mean "I") often overlook.