Ineffable
Ambrose proposed loan move to Blackburn Rovers is being held up with work permit issues. With only 6 months left on his Celtic contract, the club are happy to see him leave on loan. It looks like the end of his Celtic career. Starting 2 games this season, he has not even made the bench since starting away to Astana on 26th July. At 28, he will want 1st team football.
Rarely can a player stimulate such divisive opinions as Efe Ambrose. With his lovely smile, gymnastic goal celebrations and charming interviews there is genuine affection for the Nigerian from the Celtic support. Then there are the “bomb scare”, and unkinder, sobriquets for a defender perceived to make too many, and high profile, errors.
Celtic Appearance Summary
Season Total Mins Games Started 90 mins Used Sub Unused Sub Not In Squad Injured Susp Internats 2012/13 3307 36 28 5 2 3 0 0 6 2013/14 4632 51 49 1 1 0 0 0 0 2014/15 3130 34 30 9 15 1 0 1 0 2015/16 2021 21 20 10 18 8 0 0 0 2016/17 180 2 2 0 0 38 0 0 0
Source: Transfermarkt.co.uk
Ambrose quickly became a regular pick by Lennon at DC. The 2012/13 season saw Ambrose contribute to 2 hugely notable victories, away Spartak Moscow (Celtic’s only Champions League Group Stage away win) and the 2-1 home win against Barcelona. His season got even better as he was part of the African Nations Cup (ANC) winning Nigeria team, playing 5 matches out of 6 at DR. He was named in the Team of the Tournament and described being part of the Cup winning nation as one of the best moments of his life.
The capricious nature of football was never better exemplified when on his return from celebrations on winning the ANC, he was selected by Lennon for the Champions League Last 16 1st leg vs Juventus. A clearly unfocussed Ambrose made errors directly leading to the 1st and 3rd goals as a street wise Juventus, under Antonio Conte, eased to a 3-0 win.
Celtic won the double that season and with Ambrose starting 36 games, he remained a mainstay of the team albeit only playing 9 mins in the Scottish Cup Final win against Hibernian.
The 2013/14 season saw Ambrose have his peak season starting 51 of 53 games including all 6 Champions League Group Stage games as Celtic again won the league. Forming a formidable defensive pairing with the prodigious Van Djik, and behind Fraser Forster, Celtic set a new record of 13 matches with a “clean-sheet” and 1,256 minutes without conceding a goal.
With the arrival of Deila, it started so well for Ambrose, I had him as Man of the Match away to KR Reykavik in game 1. It quickly turned south, being sent off in that awful night in Warsaw against Legia, trying to undo Van Dijk’s mess. In 44 minutes, he’d given away 6 passes and had 2 defensive errors and 2 critical errors.
Ambrose remained a match day squad regular for Deila but was increasingly an unused substitute, and when playing was alternated between DR and DC to cover injury to Denayer, Lustig and Matthews, starting no more than 10 games in a row. This trend continued in Deila’s 2nd season before Rodgers discarded Ambrose from 1st team duty after 2 early season starts.
A versatile player, Ambrose has played at DR, DL, DC and DMC. Remarkably he has not missed any games injured at Celtic.
(NOTE: A reminder I have player level game data from the start of Deila’s reign.)
Ambrose Defending
This is, largely, a decent radar. If you compare his radar to the 2016/17 DCs performances at Celtic (Simunovic, Sviatchenko and Toure – Boyata’s data sample is too low yet):
His passing accuracy is lowest of the 4 (even at 90%);
Ambrose PEI of 77% lags 4% behind Toure and Sviatchenko and 8% behind Simunovic;
Only Sviatchenko wins more challenges, although Ambrose tackle success of 73% is the lowest of the 4;
He completes more interceptions that the other 4 though his interception success is 6% less than Toure;
He concedes less fouls that Sviatchenko but more than the other 2;
He clears less than Sviatchenko but more than the other 2;
He executes slightly less defensive saves than Toure but slightly more than the other 2;
Defensive errors – at 1.16 per game, this is 3 times higher than Sviatchenko, 2.9 times higher than Simunovic and 1.3 times higher than Toure.
Ambrose is a decent defender by the standard of the current crop, although perhaps insufficiently a standout to warrant being a 1st team regular. But it is the error count that differentiates Ambrose.
Error Comparisons
(NOTE: the assignment of defensive errors and critical errors is a subjective one on my part. My self-guidelines are:
Defensive Error = an action highly likely to lead to an opposition chance/shot; and
Defensive Critical Error = an action highly likely to lead directly to an opposition goal.
You can conceivably find error in every goal conceded or chance created which, I believe, can negate the value of good opposition play. Overall I would say I am quite forgiving in assigning out errors, but whether I am harsh or fair, it is all my decisions and therefore there is consistency!)
I looked at the error rates of all Celtic defenders 2014 to 17 who have played at least 10 game s (i.e. 900+ minutes).
Taking defensive errors first:
As noted in an earlier article Tierney defends in a low risk style, prepared to concede possession to get the ball away from goal, as does Sviatchenko. Lustig is a particularly “safe” player, sound positionally, strong, and rarely takes risk. It is no surprise Van Dijk and Denayer are near the peak. Mulgrew has many minutes in midfield where there is less risk of conceding defensive errors.
At the bottom, we should be concerned for young Eoghan O’Connell although it is a very small sample and he had an absolute nightmare match vs St Mirren on 27th Sept 2014 accounting for 4 of his errors. Toure also suffers from a small sample (11 games equivalent, and 1 poor game vs Borussia Moenchengladbach).
Ambrose, over the equivalent of 59.2 games, is committing virtually 1 error every game on average. If Van Dijk is the “Gold Standard” for recent Celtic DCs, then his error rate is 88% higher than the Dutchmans.
(NOTE: I will consider Boyata in a companion piece to this article when I have collected the game data vs Hearts and Aberdeen).
Looking at Critical Errors:
Is it surprising having Ambrose and Boyata at DC in that horrid night away to Molde might, in retrospect, not have been a great idea?
Young O’Connell looks a real prospect in terms of his other performance data and hopefully being toughened up in English League One will knock the errors out of him. Considering Sviatchenko has not, in my opinion, committed any critical errors yet for Celtic, Ambrose rate of 1 every 5 games is a clear outlier (with Boyata) compared to the population of Celtic defenders 2014-17. This is 4 to 5 times the critical error rate of current 1st team regulars like Lustig, Tierney, Simunovic.
For completeness, I also record Defensive Saves = defensive action that stops a clear shot/chance. Goals are avoided simply by defenders standing in the right position, so this is more about those key blocks, last ditch challenges, goal line clearances and touches that take the ball away from the expectant scorer.
Two stand outs from this analysis:
DCs make many more defensive saves than full backs – this is logical as DCs are closer to the goal and play centrally where most chances and shots occur.
The same players who make a lot of defensive errors also make a lot of defensive saves!
Van Dijk and Denayer figure highly again and they were simply outstanding for Celtic. But at the top of the defensives saves table is Boyata and Toure. Boyata has the potential still to be an accomplished defender and Toure has been one for many years.
Ambrose is in the middle of the pack. His defensive saves perhaps not balancing out his defensive errors in the great ledger of defending. As mitigation is should be said that of his game time of 5331 minutes, 2259 mins (42%) has been at full back not DC and as we noted, full backs execute less defensive saves.
Why do those making the most errors also make the most saves?
Defending is about positional play, team work and communication, spatial awareness (assessing multiple players’ positions, the ball and space simultaneously) and decision making. The “stats” don’t record those aspects, rather we can infer from outcomes whether all those factors were well executed.
Good defending is therefore, if not hidden, certainly more difficult to discern from the data – in fact it may be the lack of certain data that infers consistently good defending. That is, a player who consistently positions himself well; who is aligned with fellow defenders and who is adept at tracking the ball and attacking movements simultaneously, avoids conceding goals through the lack of chances being created and shots being taken.
If a player has relatively poorer positioning, communication and decision making, he is likely to be having to take more direct and individual action to mitigate the risk of the opposition scoring. A recovery tackle, for example, may have been avoided with better positioning. He may rack up more errors and saves. You don’t need to take as many risks if you are usually in the right place.
Coming back to Ambrose, a quick player for such a tall, strong man, and who always looked comfortable on the ball (except when faced with the opportunity to pass the ball into the box from wide), his defending performance data shows that his contribution is simply not outstanding enough to make up for his high defensive error rates and relatively inaccurate passing. But what a guy!