FOOTBALL MANAGER 2019: TUTORIAL

SUMMARY

I was part of a 5-person team who designed and engineered the new tutorial system for Football Manager 2019. Creating a tutorial for an extraordinarily expansive and largely text-based simulation within a 15 year-old engine brought many interesting challenges!

KEY DISCUSSION POINTS

Considerations for new player experiences when text is the primary medium
Striking a balance between simplicity and broad coverage in a gigantic pre-existing game
Building a reusable system for future editions and features

ROLE

Gameplay Engineering, Design

TYPE OF WORK

Feature on shipped game: FM19 (2018)

SELECT A VIEWING MODE

Text highlights with option to expand

All text shown with no need to click

BACKGROUND

Football Manager is an almost 20 year-old football simulation franchise created by Sports Interactive, in which players take on the role of manager at a football club of their choosing. The series features a full 3D simulation of the matches themselves, in-between which players are presented with large amounts of data and make decisions - both big and small - affecting their club.

The minute-to-minute core gameplay loop sees the player being presented with these scenarios to handle, digging into the data if they choose, and making a decision; the hour-to-hour loop is generally assessing and setting up for the next match, then playing out that match and seeing the result.

The series is well-known within footballing circles for its accuracy and comprehensive nature, and its extensive database has had many real-world applications. This incredible level of detail can be a double-edged sword, however, in that new players are often overwhelmed with the sheer amount of data and navigation options.

In late 2017 and early 2018, a team of 5 people - including myself - began work on a revamped new player experience. I handled gameplay engineering, and worked with a UI engineer, mockup artist, producer and writer, with design duties shared between us. We ended up creating something which had applications for all sorts of players - new, lapsed and regulars - and there were plenty of interesting decisions along the way!

Here's a brief rundown of the important gameplay and UI elements you'll need to know for the remainder:

The page where the average player typically spends the most time in FM is their inbox. This is where they receive messages, just as they would at a real football club - for example, information from their scouts or reminders of board meetings.
The sections on the left - like Transfers, Squad, and Training - allow players to dig into specific areas. Each area has what we would refer to as a landing page, which serves as an overview of that section. Intermediate players tend to have specific sections they visit more than others.
Specific functionality - like making a transfer offer or changing an individual's training workload - is then buried a little deeper in separate tabs or drop-down menus within a section. Highly-experienced players know their way around most of these extra tabs.
Players aren't alone in their managerial journey - almost every club has a selection of backroom staff and an assistant manager to whom you can delegate work.
Delegating in this way used a system called staff responsibilities, but before FM19 it was fairly hidden away and it was largely up to the user to seek it out.

DESIGN MOTIVATION

MITIGATE COMPLEXITY FOR NEW AND LAPSED PLAYERS

Football Manager has yearly iterations and, as such, each new game has to demonstrate its value to players by building on top of what came before. This philosophy lends itself well to experienced players, but means that if someone doesn't have a history of playing the game or hasn't played for a while then it can be difficult to get into it.

Early on, we latched onto a target archetype to represent the new player experience - our dads. They represented the kind of player we wanted to get involved - people who were football fans, and technically adept enough to play a slow-paced simulation, but who had previously been put off by the complexity of the series.

Our goal was to shepherd them from the start of the game through to at least the initial friendly matches in a club's season, and to do so in a way which would reduce the complexity of their experience as much as possible. This period of time would effectively represent one or two iterations of the hour-to-hour core gameplay loop.

KEEP THE PLAYER IMMERSED

One of FM's greatest strengths is its ability to keep players immersed for long periods of time, and it achieves this in large part by presenting information through a real-world lens rather than breaking the fourth wall.

The displaying and sending of data to the user is done as close to how it would occur at an actual club as possible - for example, a report sent to you by a data analyst, or your assistant manager on the touchline telling you what they can see.

We wanted to retain this feeling for the tutorial - that the club was taking the new manager under their wing, and helping them understand the day-to-day workings through conversations and questions rather than random text prompts without context.

GIVE PLAYERS FREEDOM TO CONTROL WHAT THEY WANT

One of the early constraints we were given when designing the feature - and the one which played the biggest role in shaping it - was that even a new player should be able to do anything they want from the moment they start the game. This is a particularly important concept to understand because so many game-shaping decisions and philosophies are set when a user takes charge of the club for the first time, and it's important that we don't affect their experience negatively.

The implications of this were wide-reaching. For example, we couldn't hide any UI elements in the game and introduce them to the player gradually, or railroad the user into specific decisions. It also meant that we had to pivot to gradually surfacing the most relevant information to the player and giving them the choice about how much complexity they wanted to take on.

This theme of freedom to control what the player cared about and hiding complexity through other in-game mechanisms became key to the design, and ended up making it applicable to all players - not just those who were new to the series.

KEY DESIGN DECISIONS

EARLY DESIGNS AND DISCUSSION

Before the tutorial was approved as a feature for FM19, two of the five on the tutorial team (independently, by chance!) had already written up feature requests for a revamped new player experience - each with different ideas!

My proposal involved removing almost the entire UI to introduce sections to the player as they became important. The idea was to delegate everything that wasn't visible yet to the player's assistant manager and other backroom staff, letting the game run essentially as normal in the background. Looking back, unwinding ~15 years of UI would have probably been technically too difficult, and it also risked the possibility of the assistant making poor decisions; however, the concept of delegation is one we kept in mind.

Our mockup artist came up with a separate idea, that of simplified landing pages for each section with the option to "delve deeper" if the player decided they wanted to open themselves up to more complexity. This wasn't included in the first iteration of the tutorial design, but we all agreed it was a great idea and I believe it ended up being added in the following edition, FM20 - design on new features like this would often be iterated on over 2-3 years as there simply wasn't enough time to get everything done in one cycle.

There was also a third idea floated (I forget by who!) to have a searchable "how-to" system, sort of like an encyclopedia of how to do everything in the game, where you could get short tailored tutorials for individual things like "I want to sign a player" or similar. A simplified version of this also made it into FM20, and I imagine with generative AI it might be more possible to achieve nowadays.

THE INDUCTION SYSTEM

In the end, we settled on an induction system - similar to the kind you'd receive when starting a new job. This ticked all our boxes, and also allowed us to grow the system beyond a new player experience by providing tours of a new version's back-of-the-box features for long-term players. Inductions were almost entirely data-driven so new ones could be added even by those without a technical background, which helped immensely with reusability.

The system reduced complexity by guiding players through the different sections available to them from the safety of their inbox, without them having to explore. We made it clear from the outset that these inductions would arrive and when - so the player didn't feel the need to leave their inbox to learn. If they did, we gave them the induction straight away.

On the immersion front, it certainly felt like a real-world setup - your assistant manager was guiding you around the club!

Finally, we managed to tie it neatly into the staff responsibilities system by ending each induction with a set of choices, allowing the player to either choose to handle things themselves or delegate them to their backroom staff. As they went through the inductions, this gave new players complete freedom to choose how much complexity and responsibility they wanted to take on.

RESPONDING TO FEEDBACK

When I worked there, a new edition of Football Manager was generally fully playable internally - with some bugs and omissions - by about July of the year it was released. This gave us a couple of months to playtest, polish and iterate before that edition's beta was released to the public. There were several small changes to the induction system during that time.

By far the biggest area of debate was the inclusion of a Skip button. Early builds of the game didn't have this option, and a debug option to permanently skip all inductions was quickly added in response to developer feedback. When it came to giving this option to players, some - myself included - felt that being able to skip inductions would cheapen them, or cause people to accidentally miss out on important information; others felt that not being able to skip inductions would be a source of great frustration to experienced players, or those who didn't care about a particular area.

In the end, we settled on adding a Skip button to inductions and I'm happy to say I was wrong - the second we began watching people play the beta we knew it was the correct decision!

We had already made a couple of supporting changes in anticipation of potential problems - on the inbox items, the "take induction" button was made considerably larger, and if a user skipped an induction they could go back to inbox item and redo it. The Skip button itself during an induction was also made small on purpose, to minimise the risk of misclicks. However, experienced players playing FM today will almost instantaneously skip any induction that isn't a brand new feature.

IMPLEMENTATION AND RESULTS

TECHNICAL HURDLES

The biggest technical and design hurdle on the gameplay engineering side was creating a system that didn't outstay its welcome in future playthroughs, while still catering for the possibility that people might share the game. This took considerable iteration to get correct.

Once a player has done an induction and played the game for some time, or decided they want to manage a different team, chances are they don't need reminding - so we don't want to show already-seen inductions on a second new career game.

Furthermore, we want this to be automatic - we don't want to rely on a player remembering to set themselves as Experienced; they might not even want to, if there are still some inductions they haven't seen. On top of that, there are some inductions we do want to show to an Experienced player - like the back-of-the-box new features.

However, what if someone shares the game with a family member who's new to it? That person definitely does want the inductions.

We went through several iterations before landing on a compromise. It's possible to create different manager profiles in FM, which can be reused across different career games; per-profile information is typically rare, but we decided to attach data on completed inductions to the manager profile. This way, when someone lends the game to a friend or family member, chances are they'll make a new manager and the inductions will be available once again (if desired). We considered this the least disruptive option.

WHAT WENT WELL

The induction system was generally well-received by reviewers and the public, and had a few standout achievements:

The system was engineered well enough that it was successfully - and fairly quickly - ported for work on a mobile game by other engineers. I left before that game was released and I gather in the end it used a different system, but it was helpful for early development.
We created a powerful and lasting system which is still in use on the main FM series today, 5 years later - and was stable enough that we could build on it considerably in FM20 with additional features.
Inductions had applications across all player archetypes, which was beyond what we set out to achieve. They were useful both for new and lapsed players, but also for experienced players - even if their interactions with them were brief.

WHAT WE LEARNED

Although the induction system has stood the test of time, it didn't achieve everything we set out to do:

The game remained complicated, and many were still put off within the opening hour. Those belonging to the original player archetype at which we were aiming didn't tend to stick with the game for longer periods of time.
It's possible that more drastic solutions would be required to truly get new players invested. We had anecdotal data that once new players got past their initial matches they were much more likely to continue; one interesting idea which was floated in a playtest but never implemented was starting the new player experience in an actual match to get them straight into the action, or at least have them watch from the crowd - as a very recently-appointed manager might do.
One thing all of us wanted to do was get actual new players in to test the game, and watch them play/ask them questions. This wasn't something we were able to make happen, and if I was designing a new player experience nowadays I'd immediately look to set up representative playtests.

BURNING UP