Player Traits (or Player Preferred Moves – PPM’s as they used to be) can have a huge impact on your Football Manager experience. Just like in the real world, from the Premier League to non-league, in FM20 what a player prefers to do can change the way your tactics actually work. You might be telling your player to do one things, and they might be trying something completely different. Perhaps more importantly though the right moves or traits can change an average player into a world beater. You don’t always need wonderkids, Ronaldo’s or Messi’s.
In this article we are going to take a look at four moves that can change the way your strikers play in FM20. Placing shots, shooting with power, rounding keepers and lobbing keepers. Using our FM Lab experiment league we pit the different player traits against each other. In an attempt to see which is best, and how much attributes play a role.
There are a huge range of Player Traits in FM20. Some govern movement like ‘Cuts inside’ or ‘Gets Forward Whenever Possible’. Others cover technique or specific actions like ‘Tries Tricks’ or ‘Tries Killer Balls’. These all give indications of things that players like to do, and will try to do. Note that doesn’t mean they will always do it, or even do it well.
Some traits can contradict each other or even the player instructions for your tactics. You can actually see where a trait contradicts your instructions in the tactics screen.
Player Traits for Strikers
Depending on the player role there are some moves or traits that are more appropriate than others. For example strikers might benefit from moves such as: likes to place shots, shoots with power, likes to round the keeper, and tries to lob the keeper.
The situation you find the striker in might have a knock on effect when you consider these potential moves. As can the type or attributes of the striker. One common train of thought is that shooting with power is more beneficial for strikers who aren’t actually that good at finishing. The accuracy of the shot is lowered by the power makes it harder to stop. Whereas a striker that is a little more skillful will benefit from placing shots in difficult to reach places.
Another viewpoint comes from the late great SFraser all the way back in 2011. You can find the thread here. Their view point was that the position the striker was in would have an impact on which PPM’s (as they were then) were better. If your striker was more central then power would be better. Out wide or from angles means there’s less goal to aim for, so placing shots becomes the better option.
Rounding keepers is another option for fast skillful composed strikers. They might not be the greatest finishers but they might often find themselves one on one with the keeper. Likewise those with flair, technique and good dose of decision making might be well suited for the occasional lob over the keeper.
In the right circumstances all of these moves could help your strikers get more goals. But are any of these moves just better than the others? And how much impact does actual skill/ability have on the moves? This is where we head back to the FM Statistics Lab.
FM Lab Recap
The FM Statistics Lab we had in FM19 was a league with 6 equal teams. The same number and type of players, with the same set up and facillities. More details can be found here. Basically as many variables as can be controlled for have been in this set up. Giving us an experimental league where we can edit our variables of interest (in this case striker Traits) and see what impact they have over several repeated seasons. This database was updated for FM20.
Player Trait Set-Up
For this experiment we created two versions of the database. In the first version all the players including the strikers had a current ability of around 100, with all the key striker attributes set to 10. The exception to this was heading which was set to 1 as we didn’t want headed goals to skew our figures. Our Player Traits don’t involve heading at all.
In the second version the current ability or CA was upped so all the stirker attibutes were set at 20. Again heading was left at 1.
This gave us two databases, or experimental leagues to run. One with moderate ability of 10 across the board, and another of high striker ability set at 20. Everything else between the two versions was the same, including the non-striker ability which was set to 10’s. In all cases the strikers could use either foot so it didn’t matter which side of the pitch they played on.
Tactically in both cases all teams used the vanilla 442 that is created when you select create a tactic from scratch.
The next change we made was to both versions. Each striker was given a trait.
- Shoots with Power (Striker 1)
- Places Shots (Striker 2)
- Likes to Round the Keeper (Striker 3)
- Tries to Lob the Keeper (Striker 4)
What we are left with are two variables to look at: ability (moderate or high), and Trait type (power, places, round and lobs). Any difference in the amount of goals scored we can potentially attribute to the role of ability, the trait or an interaction of the two variables.
To get a good sized data set each version (moderate and high) was run for five individual season (reset to run from 2019-2020 in each case).
For the statistics nerd’s amoungst us an ANOVA was run with a series of t-tests to break down any differences found. If you don’t know what thats means don’t worry. Check out the first post in the FM Statistics Lab series or just take my word for it – these are statistical tests that let us know where any differences we find in the total goals for each group are real reliable differences or not.
We are going to focus on goals and goals per 90 but we also considered minutes per goal, shots on targer per 90, shots per 90, shots on target percentage and average rating.
From the graph above we can see that regardless of Player Trait those strikers with high ability are scoring on average 1 goal per 90. Those with 10’s for their ability, the moderate players, are hitting are .2 goals per 90. Again regardless of ability there isn’t much variation in the goals scored per 90 between the different traits.
When we run the actual statistical tests we find a few key things out. First of all we find an effect of ability. When we compare the goals scored by those with moderate skills compared to those with high skill we see a statistically significant difference: those with higher attributes overall score more. Which is boringly obvious. That is exactly what we would expect (better players score more) and it’s not what we are interested in, but it’s a good pre-check. If we found no difference here we’d be worried about the experiment.
Next up we have Traits. Overall, ignoring ability, is there a statistically significant difference in the amount of goals scored by the strikers with different player traits? Is one just better than another?
The answer here is no. There’s some slight variation but overall there is no one trait that leads to more goals than the other.
Traits versus Ability
The next check is the most important one though. Is there an interaction between ability (moderate and high) and the different traits when it comes to goals scored? Are some traits better depending on the ability of your striker? This relates back to the point made about placing or powering shots depending on how good the finishing attibute is.
The answer to here is a resounding no! There’s still no difference between the goals scored or goals scored per 90 between the different traits depending on ability. The hypothesis that there would be between places and power for example doesn’t seem to be supported.
But what about just placement and power?
An issue here though is we have four traits not just places versus power. Rounding and lobbing the keeper might be creating some statistical noise. These moves are arguably rarer than placing or powering in shots.
So, I removed the data from lobbing and rounding and just compared placing shots to powering in. And what happens now?
Not much happens. When we look at average goals per 90 there is no significant difference. You can see from the figure above there for the moderate (low) ability group there’s no real difference between placement and power. For the high ability group there is a slight difference of about .05 goals per game. This isn’t a signficant difference.
What does this actually mean for PPM’s?
This has just been a very long winded way of concluding the following:
- One type of PPM isn’t intrinsically better for goal scoring (or other striker success metrics) than another. It doesn’t matter if you think strikers should be blasting it, or bending it around the keeper. They are all generally the same in this vanilla test.
- Ability doesn’t interact with the trait or PPM. Shooting with power isn’t generally better for poorer strikers. Likewise placement isn’t better for the more skillfull strikers.
- Overall SFraser’s suggestion, all the way back in 2011, still stands. There’s more to it than just the strikers ability. These PPM’s are likely to be very situational. The experiment removed the tactical nuance as all the strikers were playing in the same system. Therefore getting chances in similar positions. We know ability doesn’t have a big impact, nor does the PPM on its own. By elimination it’s not unreasonable to suggest that the position or context of the chance is much more important.
- Spend less time worrying about how good your striker is, and more about what positions they get into. This almost lends itself to the xG way of thinking about chances and good positioning.
In some ways the experiment feels like a bit of a bust. 10 seasons worth of data and the conclusion is that there aren’t really many differences. But a lack of difference is still an important finding. It weakens the old idea that the important factors between which is better, power or placement, is ability. Instead this strengthens the suggestion that context and position are more important.
It might feel like common sense but common sense and gut instincts aren’t always correct. We’ve started to back it up with statistics here.
We can’t really stop here. We have an inkling that position and context of the chance or shot are more important when we look at traits like placement, power, lobs and rounding. But we don’t know for sure.
The next step is to use the same strikers and split databses with moderate and high ability but add more one variable. The type of chances – central or wide/angled. For this we need to alter the tactics used, with half the teams playing a tactic that creates central chances, and half playing a tactic that creates chances from wider positions. If we do this we can work out what impact position and context has when we consider Player Traits.
For this I need you help. If you have tactics that you know produce more of one type of chance (central or wide) then let me know and send me the links to them in the comments. I’ll use them for my next article.