## Ambition, determination & professionalism

Let’s talk about player development in Football Manager 2022. Which attribute is more important: ambition, determination, or professionalism?

## 1 Background

Much as been said about the effect of different attributes on a player’s development. 3 particular attributes have been the subject of much controversy: **ambition**, **determination**, and **professionalism**. Over the years people (including me!) have wondered which one of these attributes is the most important. Do they all matter the same? Is ambition a negative trait? How do these attributes interact?

If you look on the Sports Interactive (SI) Games forums, you’ll get some mixed information.

This is incorrect. SI have stated that Professionalism, Determination and Ambition are equal in terms of being factors driving development. You can read more here: https://community.sigames.com/forums/topic/405653-determination/

Determination, Ambition and Professionalism all contribute exactly the same amount towards the progression score. However a certain progression score does not guarantee a certain level of progression. A progression score only gives the player a certain chance of achieving a certain level of progression. This chance ranges anywhere from 0% to 70%. As such two players with identical attributes could develop very differently. Similar two players with very different attributes could develop identically. In this way it is not possible to reliably test a single attribute’s impact on progression through simply collating and monitoring a data set.

There is even a video by youtuber Zealand in which he talks about these 3 attributes. The video has a very “clickbaity” title, stating that determination doesn’t matter, but then he says it does?. Anyways, in his video he explains a few experiments that he made with a colleague and he made some key discoveries.

- Professionalism is the most important attribute (player development wise)
- Determination matters, but maybe more for injury recovery than for pure player development.
- Ambition is
**way less**important than the other two attributes, perhaps even being a**negative**attribute.

## 2 My motivation

The experiments done not only by Zealand, but by many others throughout the years, are interesting and insightful. They, however, lack a proper methodology, at least in my opinion. While they all rely on either empirical data or simulations, they all have different drawbacks. Some fail to control for some factors. Some have what appears to be an improper statistical analysis. Neither presents the data in an accessible way. That’s where I come in.

## 3 My experiment

First of all, I wanted to make an experiment that could control for most of the confounding variables. If you can properly control these variables, then it’s possible to isolate the variables of interest (ambition, determination & professionalism). To do that, I set up a fairly simple experiment.

- Created a club that has a whopping 375 players. That is correct,
**375 players**. - Those 375 players were split into different groups, with each group consisting of
**3 identical players**. - Each group has a combination of each of the 3 previously mentioned attributes.
- To make the experiment manageable, I only used
**5 levels**for each group (1, 5, 10, 15 and 20). This means that for example group 1 has 3 players, with all 3 of them having the following attributes: 1 ambition, 1 determination and 1 professionalism.

- To make the experiment manageable, I only used
- In total we ended up with
**125 different combinations**. With each group having 3 replicates (3 identical players), we have our grand total of 375 players (125 * 3).

If that doesn’t make a lot of sense to you, don’t worry. Just check the table in the next section. This table shows all of the experimental design combinations and how we got to 125 different combinations.

### 3.1 Design table

### 3.2 Other variables

#### 3.2.1 Attributes

To make the experiment as accurate as possible, I had to control for all the other attributes. I mean, I don’t really know what’s their effect or how can they help to mould a player. So in this case, I just set **all the other attributes to 10**. This means that every player starts with an **ability of 64** and a **potential ability of 200**.

As a part of the experiment, I decided to freeze some of the player’s attributes. In this case, players had some special conditions that are perhaps a bit atypical.

- Players
**could not**get injured. - Ambition, determination and professionalism were frozen. This means that the original attributes (1, 5, 10, 15 or 20), remained fixed for the duration of the experiment.
- Players
**did not**play any matches. Yes, this is not realistic, but it controls for the effect of matches on a player’s development. This means that players trained for 11 months a year and then had a month off before starting again. - Players had maximum morale and sharpness.

#### 3.2.2 Age

Regarding the age of the player, I decided to go as young as possible since according to SI Games, that’s the age at which players develop the fastest. This means that every player in this experiment started the game at just **14 years of age**.

#### 3.2.3 Facilities and trainers

The game was simulated with a club with exceptional youth and training facilities. All the trainers had an **ability (CA) of 200**, with them having **5 stars in every single training area**.

#### 3.2.3 Length of experiment

I decided to let the game run for **3 years**, taking snapshots of the data every month. By the end of the experiment, I had a total of 36 months of data, spanning more than 1000 days from beginning to end.

Why only 3 years you may ask? Well, to me that seemed to be enough to detect the trend that I wanted to detect. This experiment is not designed to see if a player can reach maximum potential or not, but how 3 specific attributes affect a player’s development. By the end of the experiment, I had 13,875 data points. That seemed to be enough (for now).

## 4 The results

First of all, let me show you an example of how the data looks. The following chart shows the average measurement taken from 8 different groups of players. The x-axis shows the number of days that passed in game time since the beginning of the experiment. The y-axis shows the mean increase in ability for that particular group of players at any particular time that was measured. The groups are not necessarily relevant for this example, so I didn’t even label them, but the trend that the lines follow is indeed very important.

On the left chart, I over imposed the regression lines created from a linear regression model. In the Zealand video, I saw some information about the model that he and his team created, and unless I’m wrong, it looked like a traditional linear model. There is a problem with a model like that though. Do you see how the lines don’t follow the data at all? It makes sense when you think about it. **The data doesn’t follow a linear trend**.

So what can we do about it? In my case, I just fitted another model that allows the lines to be “wavy” or “curvy”. I’ll just call this a modified linear model. This model follows the data a lot better, and while it’s not perfect, it allowed me to get valid results for this analysis.

### 4.1 Importance of attributes (ambition, determination & professionalism)

How can I determine which attribute is the most important? A way to do it is by using our model to generate predictions. Since in our model all attributes interact with each other—meaning that the effect of professionalism for example is not the same when determination is 10 than when determination is 20—we have to hold some values constant.

For this chart, I decided to vary the attribute of interest (ambition, determination & professionalism) on the x-axis, and hold the other two attributes constant. So for example in the ambition plot, I vary the value of ambition on the x-axis and hold both determination and professionalism constant at a value of 10. I decided to use 10 since this is almost the exact average value of the possible values that attributes can take in football manager (1 to 20).

The results are pretty exciting. It seems like professionalism is linked to a higher ability increase compared to the other two attributes.

- A player with 20 professionalism, 10 ambition and 10 determination will increase his ability by about
**46 points**in 3 years. - A player with 10 professionalism, 20 ambition and 10 determination will increase his ability by about
**26 points**in 3 years. - A player with 10 professionalism, 10 ambition and 20 determination will increase his ability by about
**22 points**in 3 years.

We can also see that CA increases over the whole range of professionalism, meaning that more professionalism = more CA points. For both ambition and determination, the trend seems to be different. It appears that after reaching a value of 10, the rate of CA growth slows down or even stops as in the determination chart. However, this is not the end of the story. We know that there are interactions among our three attributes of interest, so we’ll explore them more in-depth in a few moments.

### 4.2 Ambition vs determination vs professionalism

This chart is a bit complex, but first, we’ll describe it and then we’ll talk about how to read it.

- On the x-axis we have determination as a continuous variable. It has a minimum value of 1 and a maximum value of 20.
- On the y-axis we have determination as a continuous variable. It has a minimum value of 1 and a maximum value of 20.
- The colour scale represents the ability increase after 3 years for each combination of the previously mentioned 2 variables.
- Professionalism has been fixed to a value of 10.

Let’s assume this chart is a player. This player has a professionalism value of 10—this is a fixed value—but for unexplained reasons can have different combinations of both ambition and determination. We want to know how much would we expect a player’s ability to increase after 3 years under different combinations of the previously mentioned attributes. With this chart, it’s very easy to do that. All we have to do is see where the lines intersect.

**For example**, if we want to know the expected CA growth for a player with 10 ambition and 10 determination, we just look for the place in which both the determination and ambition lines intersect when they have a value of 10. In this case, I marked this point on the chart with thicker lines to see it better. The result is a CA increase of between 20 and 22 points.

What about other scenarios? Let’s take a look at some of them.

- Player with 15 ambition and 15 determination: Expected CA increase of between 24 and 26 points.
- Player with 5 ambition and 5 determination: Expected CA increase of between 6 and 8 points.
- Player with 20 ambition and 20 determination: Expected CA increase of between 26 and 28 points.
- Player with 15 ambition and 5 determination: Expected CA increase of between 10 and 12 points.
- Player with 5 ambition and 15 determination: Expected CA increase of between 12 and 14 points.

#### 4.2.1 More interactions

We’ve already seen the effect of different combinations of ambition and determination on CA growth, but only while holding professionalism at 10. What if the player had a professionalism of 15, or 20? We can also take a look at that. To do that we have to discretize the attribute of professionalism. This means that now I will treat professionalism as a discrete variable—meaning a variable that has fixed levels—instead of a continuous variable. If that doesn’t make sense don’t worry, just look at the next plot and you’ll see what I mean.

Now we have a plot of plots. The plot in the middle, the one with 10 professionalism is the same as the previous plot. It looks different because the colour scale has changed. The intervals between colours are of 5 points instead of 2 as in the previous chart. We now, however, have created the same plot but for distinct values of professionalism.

So, how can we summarize this new plot of plots? Let me show you a couple of examples.

- Low professionalism (5) is linked to diminished CA growth.
- A player with low ambition and determination will barely get better, with a maximum increase of 5 ability points.
- A player with a high ambition (15) and moderate to high determination (10 to 20) will grow between 10 and 15 ability points.
- A player with very high ambition (18 or more) and moderate to high determination (10 to 20) will grow between 15 and 20 ability points.

- High professionalism (15) is linked to increased CA growth.
- A player with low ambition (5) and determination (5) will substantially get better, with an increase of between 25 and 30 ability points.
- A player with high ambition (15) and low determination (5) will grow between 30 and 35 ability points.
- A player with high ambition (15) and moderate determination (10) will grow between 35 and 40 ability points.
- A player with high ambition (15) and moderately high determination (13 or more) will grow between 45 and 50 ability points.

The same data can be presented as a line plot. The main advantage over the 2-dimension surface plot that we just saw is that now the ability increase is not binned into 5-point intervals. The main disadvantage is that now ambition is discretized into 5 different bins(1, 5, 10, 15 and 20). The results show the exact same information, so I decided to show both versions because some people may understand better the surface plot and some the more traditional line plot.

#### 4.2.2 Interesting findings

Let’s take the mini-plot that’s on the right side of the previous chart. This is the plot that contains the information for a player with 20 professionalism and varying levels of ambition and determination. There is a very clear pattern of vertical and horizontal zones.

What do I mean by this? **Take a look at the ambition value of 10**. I’ve highlighted this value in the chart with a thicker horizontal line.
You can see that if determination increases from for example 5 to 6, there is an increase in ability points. The same thing happens with an increase of determination from 7 to 8. However, once determination reaches 10, increasing more determination is not associated with an increase in ability. At least not in a meaningful way. Let’s try to summarize that.

- A player with 20 professionalism and 10 ambition can go from an increase of fewer than 30 points—when determination is less than 1.5—to an increase of over 45 points when determination is 10. This is a difference of 15+ ability points.
- A player with 20 professionalism, 10 ambition and 10 determination will have an ability increase of between 45 and 50 points.
- If determination changes to 15, the ability increase will remain almost the same (between 45 and 50 points).
- If determination is increased to the maximum possible level (20), the ability increase will remain almost the same (between 45 and 50 points).
- If both ambition and determination are increased to the max, then there is a slight extra increase in ability, but it’s quite marginal—at most 5 or 6 points.

Another way of looking at the same data is by breaking the attributes into two separate plots. Instead of having a 2-dimension plot, we can instead create two 1-dimension plots. Both of these plots show more or less the same information, but it may be easier to see the patterns that I was talking about.

Both of these plots are very similar. They show the ability increase on the y-axis, with the only difference that the plot on the left has determination on the x-axis and ambition as coloured lines, while the plot on the right has ambition on the x-axis and determination as coloured lines.

If you look at the left chart, you can see how ambition (10, 15 and 20) cluster together, meaning that after reaching 10 ambition, adding more of it makes little difference. You can also see another pattern, which is the way the lines flatten after reaching 10-12 determination. This means that after reaching this point, increasing determination even more makes little difference.

If you look at the right chart you’ll see a very similar pattern, which shows the exact same information that I just mentioned but viewed from a different point of view. Now determination (10, 15 and 20) cluster together, meaning that after reaching 10 determination, adding more of it makes little difference. Same with the way the lines flatten. After reaching 10-12 ambition, increasing ambition even more makes little difference. These findings are the exact same thing we saw on the left chart, which makes sense since it’s the same data presented in a different way.

### 4.3 Ability (CA) growth throughout time

It is quite clear by now that professionalism is the most important hidden attribute that impacts player development. We’ve seen the effect after 3 years of training, but what is the impact after 3, 6, 9 or 12 months?

This chart is fairly dense, but I will try to explain it as best as I can. The chart contains:

- 3 rows, with each row representing a level of professionalism (either 1, 10 or 20).
- 12 columns, with each column representing the results in intervals of 3 months.
- The x-axis represents a level of determination, with values from 1 to 20.
- The y-axis represents a level of ambition, with values from 1 to 20.
- The colour gradient represents the ability points increase based on all the possible combinations of ambition, determination, professionalism and days since the beginning of the experiment.

It is clear to see that when professionalism is low, a player will barely develop at all, even with high ambition and determination. After approximately 27 months (2 years and 3 months), a player’s progress will almost stall.

The results for players with high professionalism are very interesting. **A player with very high professionalism (20), ambition and determination will develop more than 15 points in just 3 months. This is more or less equivalent to what a player with moderate professionalism (10), medium/high ambition and medium/high determination will develop after 15 months.**

### 4.4 Player personality

I know what you’re thinking. “It’s all good and dandy, but I play without the editor so I **can’t** see these hidden attributes”. And to be fair, you would be right. I don’t use the editor while playing. It feels like it takes something away from the game. So while we can’t see these attributes, we **can** see the personality of each player.

It’s important to mention that personalities are based on a range of attributes and not necessarily on a specific value. Take for example a player with a balanced personality. This player can have a wide variety of hidden attributes. His professionalism, ambition and determination can take any value from 1 to 14. So, how can we predict his growth?

In this case, I decided to create 3 groups:

- The first group was set to have the
**average**attribute values for each category. In the case of a balanced player, he would have professionalism, determination and ambition set to 7.5. Since the game doesn’t allow us to see decimal points, I decided to round up to a value of 8. - The second group would have the
**maximum**possible value for each group. In the case of a balanced player, he would have professionalism, determination and ambition set to 14. - The third group would have the
**minimum**possible value for each group. In the case of a balanced player, he would have professionalism, determination and ambition set to 1.

We can see that 3 of the top 6 personalities contain the word “professional”. This makes sense since we already know that highly professional players are the ones with the highest development rate. A personality such as model citizen, with set to the minimum values it could take (15 professionalism, 14 determination and 12 ambition) will develop almost as well as the same personality but with values set to the max (20 of each of the 3 attributes). **A model citizen is a guarantee to develop to the max as long as he’s fit.**

On the other side of the scale, we have players with personalities such as slack, casual, jovial and unambitious. Casual and slack players are basically doomed. Both of them have very low professionalism levels (maximum of 4 and 1 professionalism respectively), so as we’ve seen, they just won’t develop. An unambitious player is different though. While an unambitious player will have, well, low ambition (maximum of 5), he can still have high professionalism and determination. This means that an unambitious player can still develop fairly well—growing around 63% of what a model citizen would in the same timespan.

## 5 Final remarks

That was a lot of information. I know some people finished reading this post and are thinking wtf. Once again, that is fair. So for the people that enjoy the TLDR, let me do a final summary of the findings.

- Professionalism is the most important hidden attribute for player development
- Ambition and determination are fairly equal in terms of importance, but way behind professionalism.
- When ambition is low (around 10 or less), increasing determination only has a meaningful impact from the 1 to 10 range. After reaching 10 determination, the increase in ability is minor.
- When ambition is high (10 or more), increasing determination has a major impact from the 1 to 10 range. After reaching 10 determination, the increase in ability is minor.
- When determination is low (around 10 or less), increasing ambition only has a meaningful impact from the 1 to 10 range. After reaching 10 ambition, the increase in ability is minor.
- When determination is high (10 or more), increasing ambition has a major impact from the 1 to 10 range. After reaching 10 ambition, the increase in ability is minor.
- In other words, a player with 10 professionalism, 10 ambition and 10 determination (gain of 20.5 ability points after 3 years) will develop a bit less than a player with 10 professionalism, 20 ambition and 20 determination (gain of 26.6 points after 3 years).
- However, a player with 10 professionalism, 10 ambition and 10 determination will develop a lot more than a player with 10 professionalism, 1 ambition and 1 determination (gain of 5.75 ability points after 3 years).

- Players with personalities that have high professionalism will develop at a very high rate. These include personalities such as professional, perfectionist, model citizen, and resolute.
- Players with personalities with low professionalism such as slack and casual will not develop properly, and unless something drastic happens to increase their professionalism, they will never reach their full potential.