Girona FC – S2,4: Did the Data Hub save our season?

FM22 Analytics, tactics and soccernomics

In the last update we’d used the Data Hub to figure out why we were still failing to create ample goalscoring opportunities. Using our findings we make tweaks to (hopefully) allow us to:

• Win the ball more often, as well as slightly higher up the pitch,
• Hold on to possession more to let us progress the ball better through the phases
• Create better chances and score more goals.

To achieve our season goal of beating Pablo Machin’s 10th placed La Liga finish we’d need these changes to work as we entered a “month of doom” as we faced Barcelona, Sevilla, Valencia and Atletico Madrid, before finishing off with an easier game against Elche.

Fresh off the confidence building win over Athletic Bilbao, we travelled to the Camp Nou to take on 4th placed Barcelona. After a pretty even start we managed to seize the initiative after 38 minutes when we caught them on the break and Kalimeundo skipped clear of the Barca defenders and gave us the lead.

The home team looked rather toothless in attack and despite their dominance possession-wise, it was us who created the better chances, with both Kalimuendo and Bustos hitting the woodwork on 1v1’s as well as Bernardo twice hitting the crossbar from corners.

As the clock ticked on we reached the 90th minute with an epic victory in sight.

To make matters worse, our back-up AMC, Baena, was recalled by Villareal just after the game.

With the hurt of the Barca game fresh in our minds, we welcomed Sevilla to the Montilivi Stadium. Previously, they’d beaten us 3-0 so I made sure the players were well up for this one. Kalimuendo was taking to his Pressing Forward role well, and twice set up Bustos to give us an impressive 2-0 win.

We lost Junca for 2 weeks due to injury, so Benfica loanee Araujo stepped in for the game against 2nd placed Valencia. We were dominated for the first 25 minutes and didn’t have 1 highlight so swapped to a Cautious mentality. After that we looked much better and took control of the game, playing some lovely football in the process. Try as we might, we couldn’t beat Domenech, who was in fine form and were headed for a hard-fought 0-0.

We sat in 11th place, 5 points behind Sevilla who occupied the 10th spot. It was going to take some catching up and we needed to stop the ridiculous late goals.

What better game next then, than Atletico Madrid? The team who beat us 5-0 earlier in the season!

After going down 1-0 thanks to a Suarez goal after 7 minutes we took control of the game and pulled one back – and deservedly so – from a Kalimuendo header thanks to a beautifully delivered Saiz free kick.

We continued to create some good chanced before another free kick, this time from Filipe Cruz, was smashed home by central defender, Bueno to give us a 2-1 win.

After managing to secure the loan services of Nahuel Bustos once again, and then him getting injured for a week, we finished off the month with a 2-0 victory over Elche. We should have scored 3 in the first half but instead left it late, with Thuram scoring in the 84th minute and then assisted Stuani who sealed the win.

Considering the teams we played, a return of 9 points is very pleasing. We are playing nice football, creating decent chances and getting crucial wins.

The xG and xGA totals are satisfying, especially considering Barcelona alone notched up an xG of 2.86.

Things are looking up with the new system we have in place, and what’s more pleasing is that we haven’t sacrificed our way of playing.

May

Heading into May, the final month of the season we’re greeted with injuries to two of our starting centre backs, Gnagnon and Bernardo who’ll both miss the first game of the month, against Villareal.

Valencia are a possession-hungry team and upon noticing their three CBs have a lot of touches, I decided to open the game tightly marking them in an attempt to make it hard to Valencia to build up from the back. To do this, I had my two strikers, Bustos and Kalimuendo, mark the outside CBs and the AMC, Essahel, mark the central CB.

Knowing this would leave us 2v2 in midfield I chose to start two more physical midfielders, in Thuram and Fuentes. The trade off of having a 3v3 against the CBs was worth it though….

You’ll see from the below it worked nicely. Thuram headed clear and Bustos collected the ball and played it over the top for Kalimuendo to steal the ball off the defender and put us in the lead.

I took these instructions off as Valencia were then dominating the game as we went back to having a 3v2 in midfield just before Chukweze equalised in the 39th minute for a 1-1 draw.

After remaining relatively quite in terms of being a goal threat, Samu Saiz banged 2 in against Celta Vigo, both assisted by Kalimuendo to put us 2-0 before the home team scored 2 goals in 2 minutes to go into half time level.

We continued to spurn chance after chance before finally getting the go-ahead goal on 80 mins after a freekick lead to a goal mouth scramble which Bueno poked home. Final score 3-2. (xG 4.23 vs xGA 1.06!)

We were now up to 10th place but still 6 points off European places, with 4 games to go. So whilst it wasn’t impossible to have the dream first season, it was going to be very difficult.

We went into the game against Real Valladolid without in-form striker Kalimuendo, who had 2 goals and 2 assists in his last 5 games, as he’d picked up an injury in training. He wasn’t needed, however, as we took a 3-0 lead after just 16 minutes in a game which ended 3-1.

The game against Catalan rivals, 7th placed, Espanyol was a special occasion as it would be the last EVER game we played at our home, the Montilivi Stadium, before moving to pastures new next season.

Obviously, we wanted to say goodbye in style and go out with a win.

We lined up our old 352 to match up with their 4-2-3-1, as their AMC was in a rich vein of form. Despite Espanyol dominating possession, we went into half time 3-0 up thanks to goals from Juanpe and a brace from Kalimuendo.

They pulled two goals back, in the 53rd and 72nd minutes to set up a very nervy ending but we held firm and left Montilivi with the 3 points.

The Espanyol win put us up to 8th place but we were still 6 point off Europe and had 2 tough games remaining. A hard fought 0-0 against Real Sociedad was marred by Samu Saiz breaking his leg and we fell 2-0 to an excellent Real Betis side who managed to clinch 4th place, ahead of Barcelona.

A tough run in to end the season but I’m delighted with the 11 points we managed to secure.

It’s also the best month in terms of xG generated, with a massive 11.63. If we can keep up this level of chance creation we should be have a much steadier ship next season.

Beating Machin

One of our main aims when starting this save was to beat the 10th place finish of Pablo Machin… well we did it! Just!

A 9th place finish with the lowest budgets in the league is amazing and I’m considering it a massive success and testament to the tactical set up which Pablo instilled into the team.

Another title for Real Madrid, with Barcelona slipping to 5th place. It was a tough season for Real Sociedad, Villareal and Athletic Bilbao.

Real Valladolid, who were promoted with us last season managed to avoid the drop, whilst unfortunately for Huesca its another relegation.

It’ll come a no surprise that Real Madrid players did well over the course of the season:

Did the Data Hub save our season?

In the last update, we took a deep dive into the Data Hub to see where improvements could be made to help us create better chances and score more goals.

Did those changes work? Let’s have a look.

General Performance

This is where we took a high level view of our team from. We quickly highlighted that we were under the league average for shooting, shooting accuracy, scoring and xG. These were the basis of the changes we made, so have we improved post-changes?

I’m happy to report that yes, we’ve improved in every aspect. Some only slight improvements but as previously mentioned: step by step, game by game.

Not only have the attacking metrics all improved but it’s also impacted the other end too as our goals conceded and xG against per 90 have decreased.

Possession

Another area we wanted to improve upon were our possession stats. Not in terms of dominating the game with lots of pretty passes, but how effective we are at WINNING the ball and WHERE we win it. We wanted to win the ball slightly higher up the pitch to give us less distance to go before we could get a shot on goal.

Our stats seem to be a mixed bag since the team have been asked to press more often and more aggressively. It seems like we’ve gone from winning the ball back around 74 times per game to around 73, and our possession lost has decreased from 86 times per 90 to 85. So very small changes.

In the possession gained chart we wanted to get higher percentages further to the right. We’ve actually seen very little change, other than possession won just inside the opponents half has increased by 1%. As that 1% seems to have caused a knock on effect of all the areas in the general performance chart to improve I’m happy with it but eager to push it up a little further.

The best way to see if our changes made us win the ball more is by looking at the tackling stats:

We definitely have improvements here, without having to click on the “get stuck in” instruction. We improved from 16 tackles attempted per game at a 79% ratio to 17 at 80%. Again, small improvements and still below the league average for attempts, so we still need more aggression in midfield, but we’re clearly getting there.

Passing

Before the changes, we made the bulk of our passes in our own half and set out to improve on the 14% of passes completed in the final third.

Now this is a strange one… the data hub charts show that our passing stats have decreased in the final third from 14% to 13%, however, from watching games I certainly thought we were playing better in that area. Maybe the sample size is too small as we’ve played some ball-dominant teams, some of which regularly attempt over 700 passes per game:

In our last two months we’ve played against (from the top downwards) Real Sociedad, Real Betis, Espanyol, Barcelona, Real Valladolid, Villareal and Celta Vigo. All of whom are above average of passing stats. Therefore I’m eager to see how this works out of the start of next season where we should get a better spread of teams.

Are we creating better chances?

The whole aim of all the above is to get to this: are we creating better chances for our strikers?

Clearly, yes. Yes we are. As you can see there’s been a shift from the bottom left quadrant (BAD) to the bottom right quadrant (BETTER). This means, that whilst we’re still now mostly below the league average in terms of shots per game, we’re creating an ABOVE average xG per shot. Woo hoo!

So, has the Data Hub saved our season?

Its safe to say having a play around with the Data Hub and using what it told us to make amends has without doubt worked.

Here’s the before and after stats:

More points per game, more goals per game and less goals conceded per game!

That’s an impressive turn around.

If we’d carried on at 1.28 points per game we’d have finished off in 12th place. Instead, we avoided a bottom half finished and in fact ended up being 6 points off the European places in our first season in La Liga, with the lowest budgets in the league.

I hope you all enjoyed this update. I’ll venture into the main post-season review in the next blog.