Ten Things: A Football Nerd’s Guide to the 2020 Crewe Alexandra Team
A deep dive into last season, and a look forward to the next one.
Who wants 5,000 words of analysis about a (soon to be) League One football team?? Well, since no-one asked, here it is. And what a team to be writing about. The 2019/2020 Crewe Alexandra football team were a joy to watch. If you polled 1,000,000 football fans and asked them what they would most want from their team, you’d get answers like:
· Winning lots of games
· Playing attractive football focused on attacking
· Using young players, especially academy products
· Having a good cup run and beating bigger teams with bigger budgets
· Beating their local rivals
Crewe Alexandra last season emphatically provided the first 4, and still shat the bed against Port Vale. Nothing’s perfect, eh?
Still, with promotion to League One officially secured through the romance of a weighted Points Per Game formula, things are on the right track for the Railwaymen. With a quiet pre-season expected, there is no better time to run a hot analysis bath, really soak into last season’s success and start to get a bit excited about next year.
I have shamelessly ripped off this article’s format from Sheil Kapadia’s brilliant NFL writing on The Athletic. Data is from FBref/Statsbomb, WhoScored, Mark O’Haire, Experimental 361, Gareth Cooper, Wyscout and transfermarkt. Age plot is built using Robin Wilhemus’ app. Thanks to Blues Breakdown for helping with the PPDA data. Let’s jump in.
1. Crewe prove you can win things with kids.
As my collaboration with Playmaker Stats explained, Crewe built their success in 2019/2020 on a core of 6 young players – Will Jaaskelainen (now 21 years old), Perry Ng (24), Harry Pickering (20), Ryan Wintle (23), Tom Lowery (22) and Charlie Kirk (22).
Neatly, they all feature in the top left of the team’s 2019/2020 age plot pictured above, as the 6 players who played the most minutes for Crewe last term (across all competitions). This team did a lot of intelligent things this season. The simplest thing they did was pick their best players every game. When you’ve got young, extremely talented players, you let them rip. Crewe manager Dave Artell did just that.
As the long ‘tail’ for each player in the age plot above shows, these youngsters have all been with the club for several years. This was critical to their success in 2019/20, which is something I predicted last summer. Here’s the key section:
“(Research) shows that, on average, it takes a signing three years to hit their peak at a new club…Crewe’s squad have (cohesion) in spades. When considering that most of our team are academy graduates and have been with the club from a very young age - they greatly benefit from this potential continuity.”
Continued faith in young players paid off. Among League Two teams, Crewe gave the highest share of playing time to young players, with over one minute in every three going to someone aged 21 or younger. Experimental 361’s graph shows how they compare:
Crewe’s business model for staying afloat relies on the youth academy producing 1-2 talents every few years, putting them in the ‘shop window’ by playing them in the first team, and then selling them to bigger clubs. After a lean few years for producing young talent, 6 have come along at once.
2. They will lose a great deal of experience this summer.
Crewe supplemented their young core with a small group of older players. These grizzled veterans all played important roles during the 19/20 season, but most won’t be around next term. Shown below is a bar chart of the Alex’s squad for 2019-2020, ordered by career league appearances. Those in orange are leaving this summer. This clearly shows the upcoming experience loss within the squad:
Hunt and Nottingham departing leaves thin centre back depth. I’ll go into more detail about Nottingham later but just a quick word for a now very slow Hunt – cheers. You were clearly showing your age and it is rightly time to say goodbye. Still, your calm leadership didn’t go unnoticed. You also had the highest pass completion % among all League Two defenders (82.4%). Respect.
In midfield, Paul Green was a helpful role player and an exemplary role model for his younger teammates. His stats don’t wow in any area, but he was a reliable presence. Green was an ever present in big games against opponents at the top of League and very rarely had a poor game. He will reach 600 career league appearances very soon, which is testament to his fitness, longevity and commitment.
Another veteran midfielder is leaving. Despite being only 25, James Jones was one of the most experienced squad members and has left for Lincoln. Jones came through the academy and played over 150 games for the club. At his best, he is an all action midfielder with undoubted talent.
3. The Railwaymen’s style of play is more reliable than the train network.
Crewe built their tactical approach on continuity and balance. They used a 433 shape about 90% of the time in the 2019/20 season, with clearly defined roles for every player, which played mostly to individual strengths.
The shape was unwavering and made little adjustment to opponents. The only adjustments made were in personnel, but even then, there wasn’t much variation from week to week. Only 14 players were on the pitch for more than 1,000 minutes for The Alex last term:
• Ryan Wintle – holding midfielder
• Perry Ng – right back
• William Jääskeläinen - goalkeeper
• Harry Pickering – left back
• Charlie Kirk – left winger
• Tom Lowery – central midfielder
• Chris Porter - forward
• Daniel Powell - right winger
• Nicky Hunt – centre back
• Paul Green – central midfielder
• Eddie Nolan – centre back
• James Jones – central midfielder
• Chuma Anene - forward
• Michael Nottingham – centre back
I polled Crewe fans on Twitter to ask them about their favourite team performance from 2019/2020. 57% chose the 3-1 home win against eventual champions Swindon. The average positions (taken from SofaScore) of the Crewe team that day are below:
As shown here, the centre backs sit much deeper than the rest of the team (number 6 Eddie Nolan, and number 15 Nicky Hunt in the image).
Both full backs have license to go forward. Fullbacks Perry Ng (number 2) and Harry Pickering (number 3) are probably the 2 standout players, and are critically important to how the team attack and defend.
Wintle (number 4) was nearly ever present at the base of the midfield, progressing the ball and shielding the defence. He is a jack of all trades and has turned into the master of many of them. His talent on the ball was especially important to the team’s strategy last term.
With Crewe’s attacking endeavour on the left-hand side, Artell provided some caution by shading the hard working and unselfish Green (number 12) over to this side, especially in games against strong opponents.
He was slightly more withdrawn than Tom Lowery (number 16). The right sided central midfielder has license to create, drifting into the #10 position or the right ‘half space’.
Wingers in the system have license to attack. On the left, Charlie Kirk (number 10) drifts inside on his favoured right foot, allowing Pickering to overlap on the outside. The right winger, in this game Owen Dale (number 19), stays wider, with Perry Ng as likely to ‘underlap’ as overlap.
Lastly, the central striker stays high, and as central as possible. Chris Porter (number 9) normally played this ‘target man’, though Crewe prefer to use shorter passing. Nonetheless, his physical and aerial presence remains especially useful when they decided to go long.
Crewe led League Two in possession (average of 58.3% per game). They had more of the ball in 34 of their 37 league games last term. Patience was the name of the game in their build up. The team’s balance meant that threats in possession came from all over the pitch, with wing play proving an especially useful tool in their attacks - 35% of attacks came from the left side, with 25% in the middle, and 40% from the right side.
They occasionally used 4231 as a more attacking shape, using a prominent number 10 against weaker opposition. Both Tom Lowery and Callum Ainley looked effective in this role, though it weakened the defensive structure and build up through Ryan Wintle in the deepest midfield role.
Using a 433 shape, possession heavy football, a low-medium defensive block, and relying on their extensive technical talent all over the pitch to unlock opponents was the formula in 2019/20. You can count on it being the formula again in 2020/21.
4. They used a new keeper, and it turned out he was a keeper.
One of the bravest decisions that Crewe manager Dave Artell made in 2019/20 was naming Will Jaaskelainen as his goalkeeper at the start of the season.
It was a rocky start. Jaaskelainen looked tentative and a bit undersized (think Kepa) in the first few games of the season. He was hesitant to come for crosses and wasn’t vocal enough. Unwavering, the manager kept his faith, and his goalkeeper grew in stature (literally) throughout the season, becoming one of the key success stories. His assured play out from the back meant he often became the first form of attack (think Ederson).
He wasn’t perfect by any means. Using standard goalkeeper measures, Will Jaaskelainen was middling by League Two standards. His save percentage was 70.2%, which ranked 10th among starting League Two keepers. In January, so around ¾ through the season, analyst Gareth Cooper assessed League Two goalkeepers by Goals Saved Above Average. This measure tells you exactly what is says on the tin – how many goals a keeper has kept out compared to the average goalkeeper (based on where the shot is heading). Again, Jaaskelainen was around average, ranking 12th in League Two.
Where he excelled was with the ball at his feet. His comfort building from the back was a crucial, underappreciated asset. In a possession focused side, properly building attacks from back to front starts with the goalkeeper. He set the tone. He played just 4.8 long passes (passes over 25 yards) per game, which is less than any other League Two goalkeeper. It was even less than a few of his outfield team mates which is unusual, with Wintle (7.3), Nolan (5.9) and Ng (5.3) all playing more. His ability to play incisive, shorter passes showed in his accuracy. Jaaskelainen’s pass completion was 64.7%, the highest among starting League Two goalkeepers. It really isn’t an exaggeration to say he is better with his feet than some League Two defenders.
5. Their defence weren’t exposed very often last season, but their sheets didn’t stay clean.
In open play, Crewe conceded 0.7 xG per 90, which ranked joint 5th best in League Two.
Crewe’s main defensive tactic was keeping the ball. However, when they lost it, opponents found it hard to penetrate the team’s final defensive line. Their opponents only averaged 6.11 shots in the box per game against them (ranked 2nd best in League Two). They forced opponents to shoot from distance – 44% of shots conceded were from outside the box, the highest percentage in League Two.
Their issue was a lack of clean sheets. They kept clean sheets in just 21.6% of games, joint 17th most in League Two. They never got more than 2 league clean sheets in a row. The data doesn’t provide a lot of answers for how they managed to always concede, but not all goals (and xG) are created equally. Conceding 1 reasonably big chance a game is enough to blow a clean sheet, and it seems like that’s what this team did. Their average xG per shot against was actually low (0.08). It just seems that 1-2 concentration lapses per game let opponents in, rather than a big systemic issue.
Without the ball, Crewe are a passive team, which may seem at odds with their high possession style. You can show this through measuring their Passes Per Defensive Action (PPDA). This stat shows how many passes opponents made in their own half and the middle of the pitch before a Crewe player stopped the possession through a tackle, interception, header, foul etc. Crewe’s PPDA was 11.76, which was the most in League Two (the League average was 9.13). Though their front 5 players did press (especially the central midfielders), they only did so once the team settled into a low-medium block, and in their own half.
In keeping with their ‘nice’ reputation, the team gave away the least fouls in League Two by a long way (297). They also had the least yellow cards (38).
The second half of the season brought personnel changes, which in turn tweaked the defensive formula. The use of Michael Nottingham and Luke Offord at centre back allowed the team to use a slightly higher defensive line, moving the team into a clearer medium block. This didn’t turn Crewe into a high line pressing machine by any means, but it did help nudge their defensive actions slightly higher up the pitch.
Defending is a team effort, so highlighting individuals seems wrong. Who am I kidding? Here’s Perry Ng’s output:
He’s a special talent. This Total Football Analysis graph shows that Ng progressed the ball forward better than any other League Two fullback. But you can’t ignore his defensive output. He was among league leaders in interceptions and defensive duels. He led Crewe’s side in tackles per 90 and was 3rd in interceptions per 90. He gave away the most fouls and was the most fouled player in Crewe’s side. When the game gets feisty, Ng is in his element.
6. They have attacking threat all over the pitch (and on the bench).
In open play, Crewe created 1.06 xG per 90, which ranked 3rd best in League Two.
The goalkeeper and defenders were often their first form of attack. Their incisive passing, and the sheer technical talent of the full backs, meant that Crewe having the ball was always a headache for opponents, even in ‘innocuous’ areas.
In Perry Ng and Harry Pickering, Crewe (whisper it) have one of the best pairs of fullbacks in the Football League, never mind League Two.
Pickering stood out so much that The Athletic recently highlighted him as the ideal left back signing for Liverpool. Their analytics writer Tom Worville used this graphic to display the enormous output that the left back generated:
This chart shows Pickering’s performance compared to other full backs in a variety of skills. He is one of the best tacklers (an area of his game he’s vastly improved), but it’s his attacking output that sets him apart. Pickering created at an elite level for his teammates, he got forward effectively, and even showed a great eye for goal himself. Most notably, Pickering provided variety to Crewe’s attack. His skill and willingness to cross (222 crosses) was a contrast to his teammates. Only one other Crewe player (Lowery, 119) hit more than 100 crosses.
Holding midfielder Ryan Wintle was the most reliable form of ball progression for the side. He won both both the Player’s Player of the Year, and the Club’s Player of the Year awards for 2019/2020, which shows the value of this contribution. He made things tick.
His main strength is the range of his passing. Far from simply retaining and recycling possession, he used the ball bravely and often broke opponent’s defensive shape with killer passes. Wintle led all midfielders in League Two with around 11 progressive passes per 90 minutes.
Left winger Charlie Kirk was main beneficiary of these passes. He in turn passed the ball onto a teammate, and then invariably it would end up in the net. Kirk led League Two with 14 assists. I’ve mapped them below:
The key thing to note is how Kirk put chances on a plate for Crewe’s attackers. Most of these assists were low passes to players within 10 yards of goal. Kirk’s assist record was far from a fluke. He makes attackers lives’ so much easier.
Kirk was also critical to Crewe adding another string to their attacking bow – counter attacking. The Alex led League Two with 5 goals from counter attacks. Artell left Kirk upfield when games stretched, and opponents left bigger spaces in behind. His incisive pace and dribbling ability maximised Crewe’s counter attacking threat, most notably against Grimsby.
While top assister Kirk and top scorer Porter were more likely to grab headlines, the true beauty of Crewe’s attacking was their variety of threats. Anene is an example of this from the bench, though the most notably ‘unsung’ attacking threats for the side were Oliver Finney, Callum Ainley and Daniel Powell. They all provided a healthy attacking threat, as shown in this Expermental361 chart:
Finney especially excites me. Though we need to be careful with a small sample size, and he usually appeared against weaker opposition, his stats just pop, just like Tom Lowery’s did a year ago. He could be a breakout star, as this chart from Total Football Analysis shows:
He led League Two midfielders in non-penalty goals per 90 minutes, also excelling in deep completions (passes in the final third), dribbles, and expected assists (xA per 90).
Far from just an attacking threat, Finney also shined ‘off the ball’. Watching him closely shows he is happy to do dirty work, providing the box-to-box energy the role vacated by Paul Green demands. His interception (0.76 per 90 minutes, 2nd among Crewe midfielders) and tackle (1.24 per 90, 1st among Crewe midfielders) numbers show this. Finney is left footed which is a further bonus. He adds balance to the midfield, especially in quick build up play.
What became of Tom Lowery, whose stats popped from limited outings in 18/19? Well, it was a season of two halves for ‘Miniesta’. In early games, such as against Bradford , he ran the show. At his best, he managed to fill 2 creative roles for the side, sometimes drifting into the number 10 space between the midfield and defensive lines, other times pulling wide to the right and delivering crosses. He also scored the goal of the season against Salford. In the second half of the season, he struggled with injury.
Finally, Chris Porter. Occasionally calamitous, often offside, he was always reliable last term. He has a simple job – hold the ball up, win headers, score goals. His League Two stats speak for themselves:
· Goals: 12 (led the team, 7th in League Two)
· Goals per 90: 0.52 (led the team, 3rd in League Two)
· Aerials won: 6 per game (led the team, 14th in League Two)
· Offsides: 1.2 per game (led the team, 1st in League Two)
7. They are good at set pieces, but could still gain an extra edge.
Crewe generated 0.56 xG per 90 through set pieces (corners, free kicks and penalties). This ranked 5th most in League Two. They conceded 0.41 xG per 90 through set pieces (again corners, free kicks and penalties). This ranked 7th least in League Two.
Crewe created 9 from corners and free kicks (17th best record in League Two); they also conceded 9 goals from corners and free kicks (8th best record in League Two). This is a bit of weird quirk of set pieces; League Two teams had a clear bell curve – most teams scored and conceded around 9-12 set piece goals during the season.
Scoring from set pieces is clearly an area where they can improve going into next season, though it seems the main issue was the quality of their finishing from free kicks and corners. They certainly have capable set pieces takers – Harry Pickering especially provides an excellent crossing option. Historically they have shown imagination in set piece routines. Alex fans will have wistfully looked at England’s 2018 World Cup success, built on set pieces, as former Crewe coach Steve Holland masterminded these. I’m hoping they can tap into this history next year.
Defending set pieces is hard with a team made up of smaller, more technical players. Considering this, Crewe did a good job defending set pieces overall.
Chris Porter was a reliable penalty taker for The Alex last term. He took 3 and scored 3. Newly recruited Offrande Zanzala, who is most likely to play when Porter isn’t, is also a handy penalty taker.
At the other end, Crewe saved a few penalties this season, which is testament both to their goalkeepers, but also their data analysis. The keepers had valuable information for guessing the right way, shown in the penalty shootout win against Middlesbrough in the Carabao Cup.
8. The manager is exactly what they need.
Dave Artell’s leadership is a superb fit for the club. He had a clear plan for using a young core of homegrown players and wasn’t shy to turn to the academy during the season, most notably with Luke Offord’s emergence.
There was clear continuity in Artell’s selection in the 19/20 season. He rewarded good performances and he kept squad players happy with runs of games when he saw fit. He used a tight knit squad with just 24 players used in League Two, the joint lowest amount in the league. A tight knit cohesive squad tends to be a happier one - reflected in success on the pitch.
Results reflected this continuity. The team’s longest streak without a win was just 3 games. the lowest in League Two. Even when there was a sniff of things going wrong, Artell was unwavering.
The fighting spirit he nurtured paid dividends. No team in English professional football were better at comebacks, with a whopping 26 points gained from losing positions. While this certainly points to sluggish starts, it more starkly points to strong finishes.
This was partly down to the effective use of substitutes. Chuma Anene, Owen Dale, and Callum Ainley were often impactful from the bench. Crewe’s possession heavy, patient style would continue, but fresh legs would be enough to nick points at the end of games against tired opponents. It wasn’t exactly Plan B, more Plan A+.
A small, related nit-pick was that Artell was sometimes slow to change personnel in games when Plan A wasn’t working. Crewe subs got an average of 18 minutes on the pitch, which was second lowest in League Two. Sometimes he didn’t use subs at all. For example, away at Newport when 1-0 down, and with sweltering temperatures, Artell didn’t make a single sub. It was bizarre, but it was a mistake he seemed to learn from.
Artell had a decent record with transfers for the 19/20 season. Daniel Powell was an absolute success and blended perfectly into the attack. Olly Lancashire was not a success, though could atone in the upcoming season. He suffered injury problems but also flattered to deceive when he was on the pitch – struggling in possession especially. Christian Mbulu, who I am so sad to mention due to his tragic recent death, helped fill the central defensive berth during an injury filled winter period.
All the loanees were a success. Chuma Anene was a leftfield find and proved the club’s recruitment is getting smarter. The club identified a specific gap and profile of player they wanted and weren’t afraid to look across Europe to find the right fit. Chuma was certainly that.
Michael Nottingham, despite only joining on a short-term January loan, instantly provided calm and leadership to a team that needed it. I asked Crewe fans on Twitter which addition (transfer or loan) was their favourite in 2019/20. Nottingham was the clear winner with 64% of the vote. If he doesn’t get called “The Blackpool Tower” by Blackpool fans next season, considering he is really tall and is just an iconic colossal structure, then they are really missing a trick.
The other loanee, Stephen Walker, had the least amount of time to impress, but certainly did in his limited outings. Finding young, hungry players from bigger clubs who can complement the existing squad should continue to be a priority for Artell, who makes excellent use of his address book as well as the club’s recruitment analysis.
The team also performed better in cup competitions than they had done for several years. They even had a live game on TV! Making the FA Cup 3rd round, with Artell blending squad players alongside the usual line up, benefited the club financially and is another tick in the review of Artell’s season.
Lastly, and most importantly, Artell has helped bring pride back into the football club. I cannot state this highly enough – Crewe Alexandra have had a horrible, existentially bad few years off the pitch. A clean slate in the ownership, the board room, and on the pitch, has inspired a much better feeling around the club. Artell deserves a huge amount of credit for his stoic leadership during a turbulent period.
9. They made their own luck.
Crewe certainly benefited from elements of luck in their season, though built their success on consistently good performances and underlying numbers.
They scored 67 league goals, when their overall xG figure was 59.94. This aligns – you can certainly allow a small discrepancy between the figures which may indicate above average finishing ability. However, there may have been a small element of luck.
They conceded 43 league goals, when their overall xG against figure was 41.07. Again, these align. Crewe conceded very slightly more than you would expect considering the quality of their opponents’ chances. In basic terms – Crewe played well: scoring, conceding and winning about as much as their underlying performance indicated.
As highlighted earlier, their continuity in selection was key to their success. Therefore, it’s fair to say that the team were fortunate with injuries. They weren’t without issues though. Injuries to central defenders Eddie Nolan, Olly Lancashire and Nicky Hunt timed badly, and left the club in the lurch, until the emergence of the Nottingham and Offord partnership.
If the season ended in November, I think most Crewe fans would highlight Tom Lowery as the player of the season. In the final few months, he struggled with injuries and his creative and goal output suffered. Considering this was his first full season, the sheer amount of games may have taken its toll.
Lastly, the promotion itself had a small element of luck about it. While I think it was the right decision (trying to take bias aside) to promote teams based on Points Per Game, there were other ways of doing it. As fans, we should recognise that the team deserved promotion, but it was far from guaranteed.
10. As things stand, they are going to be okay in League One
History clearly shows that the gap in quality between League Two and League One is small. Until Tranmere’s unfortunate relegation by Points Per Game, no team promoted to League One had gone back down for 9 straight years, a remarkable record. In that time, promoted clubs’ average final position was 11.9– almost exactly mid table. MRKT Insights have done a deeper dive into gaps between divisions in the Football League, which is well worth a read.
Assuming their young team continue to improve, and they retain their cohesion from last season, Crewe’s prospects look good. You can straightforwardly project a starting line-up for next season, made up of players from 19/20, that could compete in League One:
· GK: Will Jaaskelainen
· RB: Perry Ng
· CB: Luke Offord
· CB: Eddie Nolan
· LB: Harry Pickering
· CDM: Ryan Wintle
· CM: Tom Lowery
· CM: Oliver Finney
· RW: Daniel Powell
· LW: Charlie Kirk
· ST: Chris Porter
There is some depth as well, though it is patchy. In attacking positions, Owen Dale (joint 1st for sub appearances with 20 in League Two), Callum Ainley (joint 3rd in League Two with 19) and Offrande Zanzala (3rd in League One with 17 sub appearances) all provide clear options off the bench.
Recruit Offrande Zanzala, though maybe lacking in general build up play, will provide a goal threat from the bench. He scored 0.58 goals per 90 for an average Accrington Stanley side in League One last term, with a healthy xG of around 0.35 per 90. Like Porter, he’s also caught offside a lot (a mammoth 1.55 per 90).
As of July 2020, the squad is thin in a few areas. The lack of depth is clearest at:
· Centre back – Having only 3 senior centre backs, 2 of which had a lot of injury trouble last year, is not enough.
· Central midfield – ideally 1 recruit could provide competition for both the deepest midfield position, and the ‘box to box’ central midfielder role, competing with Finney.
True to form, the academy may provide players to fill these gaps, though Artell seems likely to turn to the free agent transfer market for one or two more players. My next article (once I calm down from this absolute essay) will explore transfer options.
Overall, the team is in a good place during an uncertain time. The increased flux that other clubs may be experiencing due to stretched COVID finances could play into their hands. Crewe are more likely to fight against relegation than for promotion next year, but my gut says mid table is the most probable outcome.