Lifting Wales #3: analysing to stop losing

You know the feeling, you are smashing through results, finish second in the league and get further in Europe than ever then all of a sudden the AI finds you out and you start to crash like a 5 year old the morning after being locked in to Woolworths all night with the pick and mix.

No? Then lucky you, but if you are one of the 1 billion people who have experienced a sudden unexplained halt to a string of gorgeous, unadulterated wins in FM then read on.

How we got here

As I said, it was all going swimmingly in season three, we’d managed to finish as Cymru Premier runners up for the second year on the bounce to TNS and then progressed further than ever in the Europa Conference league by reaching the dizzy heights of the second qualifying round with the sweet £488k that came with it. So far so good. Then all of a sudden, with not a single change to my previously dominant tactic that won us 15 games on the bounce at the end of last season, this happened:

Now you may not think this little stutter is a big deal, especially if you are managing a club who is mid table or lower when you expect such a mixed run of results but two things raised the alarm for me immediately when this happened, as follows:

1. We had had that unbeaten run going into the new season
2. This is the strange world called TNS-land, a team who conceded only 13 goals and lost no games in an invincible run last season, meaning that those dropped 8pts are like absolute gold dust that may have already cost us the league 😦

So that said, what to do about it? First off, I decided to delve a bit deeper.

Why are we losing?

I thought it made sense to split the analysis into two areas, which of course are attack and defence, or more specifically, chance creation/conversion and opponent chance prevention. As we know, this year’s wonderful data hub now provides easy access to a range of metrics to help me analyse that so here goes, first by looking at our passing:

Immediately this tells me that our previous team strength of lots of accurate passing has now dropped dramatically. By comparison, at the end of last season we were approximately where I have put the yellow cross.

In addition, though I knew this already the xG table tells me that we should be scoring more but just aren’t and while the difference is only two goals, a goal in each of those last three results would have given us an additional 4pts (also reflected in the -3 xPTs score):

We are also way too wasteful in shooting (again, which I knew already), with lots of attempts but poor conversion:

As for defence, the fact that we had only let 5 goals in so far, which would have put us 4th in the table instead of 8th, didn’t cause much concern and our defending stats are not too bad at all, though clearly we could work on interceptions and attempted tackles:

Ok, so I learned from all that a) we need to improve our passing and shooting and b) to (not panic) but keep an eye on our defending, especially our interception and tackles attempted rate.

Reviewing my (no instructions) 4411

Now before you all start shouting “no wonder you are losing because you don’t have any instructions”, let me tell you why. The reason is simply because I’ve found as the save has gone on that it has been much easier to rely on the role PIs and tweaking them for particular effects and how I want the players to specifically move and play. I then add select team instructions (but only when necessary) to react to different situations, which has served us extremely well so far, taking us to two second league places in consecutive years with a team that is mid table at best.

For example, in this latest version below I have the “play for set pieces” instruction as at the moment we have a cracking centre half, Trystan Jones, who is getting a lot of headed corner goals.

To give you an example of those custom PIs, this is what I mean, with my left FB set to mark their AMR and crossing early and deep, a pretty basic change but I find it effective:

On to the changes then based on all that analysis above, with tweaks to try to address the passing deficiency, interception and tackling problem (we’ll come to the shooting later), adding shorter passing, shorter GK distribution and more pressing, which I’m hoping won’t backfire if players close down too soon and open up passing lanes:

As for that chance creation plus shooting accuracy. I noticed that all our recent goals were coming from the left side from a player called Charley Edge on Winger support but nothing at all was coming from the right from a player called Wealth Da Silva who was on IW attack, so I changed him as well to Winger support to try to create more opportunities.

Training to win

Now with only two days of part time training a week this is a hard one, though this of course can be supplemented with the individual training regimes.

First off, I made a bunch of new general training schedules that always included a lot of passing, ball retention and of course the essential match prep required when you only have a few slots to work with:

As for the individual training I added additional focuses of defensive positioning for all defenders, passing for AMs and MCs and shooting for the strikers, which may all sound obvious but can easily get forgotten as a save gets more complex. I also added individual trait training to match those aims, such as “shoots with power” for my main striker who clearly needed a shove in that direction as he was constantly hitting the ball weakly into the keeper.

Pressing the button (game 1: Newtown, home)

After a few hours of all this analysis it was time to get started with a fitting challenge, home to Newtown who were in 3rd to our 8th and on the up with a DWW in their last 3 games. My opinion going into it was that if I hadn’t made any changes it would likely be a 1 or 2 goal loss or at best a draw, so it would be interesting to see what happened. I would be watching the game on comprehensive and as well as the result I was also looking out for stats improvements in passes completed, possession, defending and shooting accuracy.

Immediately I noticed that our passing and ball retention was much better with what appeared to be a calmer approach overall which was more considered in pass choices rather than hoofing and losing possession needlessly. Interceptions and tackles were also much better as in this cracking saving tackle here:

In a much more composed performance by half time we were 1 nil up with much better passing and George Hughes, my DLP who had been running at an avg of 57.86 completed passes per game had already surpassed that.

By the game end it finished 2 nil and this is how it compared with our last loss to Barry, also at home and a worse team:

As you can see, possession was up as was pass completion, but also less fouls (showing we were less last minute and more composed maybe?). Shots however, although on the up, were still sitting at around 30% on target which I was hoping would improve with training.

Meanwhile my good buddy DLP George Hughes finished on 91 passes completed! Again though, the quality needs a vast improvement as while he is contributing a lot to the distribution his key passes are still very low so I have added “tries killer balls” to his training.

Game 2: Cardiff Met, home

And the momentum continued and even improved 🙂 I am pleased to say that in another 2 nil win we maintained our passing rate and increased our SoT! Hallebloodylujah to the FM Gods 🙂 Up to 3rd from 8th!

Game 3: TNbloodyS, away

But then it happened, the first real test against the dreaded TNS in our 3rd game, a team we had never even drawn with, never mind beat.

The team who hadn’t lost for 37 matches were standing directly above us in the league, with one slight glimmer of hope being they had been pegged back to draws in their last two games. With 6 goals conceded already in just 5 games compared to just 13 in the whole of the last campaign could cracks finally be appearing in their steely resolve? Time to find out with me thinking that I would be very happy indeed with even a draw just to add to their woes and prevent them from taking another 3pts off us.

With a slightly more defensive tactic with extra marking up for their key threats, the game started well enough with us holding them back and even matching them for passes, possession and defending. However our joy was short lived as they gained the upper hand on 12 minutes with a screaming thunderblaster from 30 yards from their Irish WB Shane Flynn whom I hadn’t even assessed as a threat!

With the stats looking neck and neck I didn’t see a reason to panic and we stayed composed to equalise on 34 minutes with this gorgeous effort from Jack Wilson who had clearly started to benefit from his new finishing training:

FM knows u know

What then followed was what can only be described as a pitiful attempt by FM to make sense of the whole situation that was unfolding before it, with us both having two players injured a piece and a player each sent off, all within the last 15 minutes, meaning it ended in a glorious 1-1 draw which was our best result against them in 4 years so far. It wasn’t the prettiest game but I’ll take it!

The proof of the pudding, 9 games on

So there you go, it worked, not only in the short term but 9 games on we find ourselves still on an unbeaten streak:

Sure, we’re not lighting any fires in terms of goal output but it is a hell of an improvement and to top it all we also managed to peg back TNbloodyS for the first time in the save.

As for what happens next, who knows? But for now, Rock n roll because as we all know when FM is good life is good! 🙂

Thanks for reading,

Daz aka @fmheathen everywhere