FOR HONOR TRAINING MODES

FOR HONOR TRAINING MODES

OVERVIEW

For Honor is an action fighting game with a host of innovative, deep and complex systems. Studio Gobo was hired to revamp the format of the existing onboarding to better aid the transition of new players to online competitive play, as well as develop advanced practice features to help seasoned players improve their game. My role involved supporting the team with the new onboarding, as well as leading the design of the Training Arena feature.

DATE

February '17 - July '18

ROLE

UX Designer, Feature Lead

TOOLS

TOOLS

Axure, Photoshop, Illustrator, InDesign, Anvil (Ubisoft Engine), Jira

TEAM MEMBERS

TEAM MEMBERS

Game Designers (x3), Technical Designer, UI Artist, Programmers (x4), QA (x2)

The challenge

Improving the Training tools for players

At launch, For Honor was celebrated for it’s unique melee combat system with its significant depth and high skill ceiling. But being a competitive player-versus-player (PvP) game with a host of novel mechanics to contend with, this created a punishingly steep learning curve for new players.

Prior to launch the team at Ubisoft Montreal had already uncovered issues with the current format of the onboarding. Usability testing and frequent live periods showed that players were often not retaining the skills they were taught, while other key skills were not covered at all. This meant players simply weren't prepared for the sharp increase in difficulty in a PvP environment, leading to a significant number of players not sticking with the game.

The For Honor team partnered with Studio Gobo to overhaul the onboarding, as well as create a dedicated freeform training mode where players could practice at their own pace. Both of these mandates challenged us to understand the pain points experience by players, and deliver updates that truly added value to the game

"What's best about the new training mode, though, is the arena. Here, you can choose any hero and adjust the game settings to play whichever way you want. You can modify health, stamina and revenge, customise the AI enemies to practice new moves and prepare for specific scenarios... It looks like a pretty decent effort!"

"What's best about the new training mode, though, is the arena. Here, you can choose any hero and adjust the game settings to play whichever way you want. You can modify health, stamina and revenge, customise the AI enemies to practice new moves and prepare for specific scenarios... It looks like a pretty decent effort!"

"What's best about the new training mode, though, is the arena. Here, you can choose any hero and adjust the game settings to play whichever way you want. You can modify health, stamina and revenge, customise the AI enemies to practice new moves and prepare for specific scenarios... It looks like a pretty decent effort!"

- Eurogamer

Discovery

From Reddit to Requirements

Studio Gobo’s work on For Honor began the day the game launched. By this point the game’s training tools had gone through multiple rounds of internal user testing, and while there were issues with the onboarding (more on that later), there was also a growing list of player needs that were not currently supported by the game. These were echoed by a growing community of dedicated players, across both the For Honor forums and Reddit.

For the Training Arena, we kicked off the project meeting with stakeholders including Game Directors, Community Managers, User Research leads, and producers, to understand the problems to solve. After gathering all the different strands of feedback both players and stakeholders, I created a list of user stories to present back to the team. These would form the main feature set of the Arena.

Training Arena pillars

Unlike the onboarding, the Training Arena would need to cater to a wide range of players - from relative beginners to the most advanced - and provide a number of features that could scale with them. As a team we also worked to define pillars to guide the design of the Arena, focusing on big ticket items that maximised the value they would deliver to players.

  • Moveset Focused: The goal of the Arena is to allow the player to learn and master the abilities of specific Heroes in a safe environment, receiving feedback to help them improve.
  • Customisable: The player is given the freedom to control and customise how they learn. AI in the Arena should support the player in learning their Hero's moveset, as well as helping the player understand and get better against specific Heroes.
  • Flexible and Scalable: The Arena should be as flexible as possible, allowing players to quickly and easily modify settings without needing to exit. It should also be designed in a way that allows for the addition of new features as the functionality of the Arena grows.

Unlike the onboarding, the Training Arena would need to cater to a wide range of players - from relative beginners to the most advanced - and provide a number of features that could scale with them. As a team we also worked to define pillars to guide the design of the Arena, focusing on big ticket items that maximised the value they would deliver to players.

  • Moveset Focused: The goal of the Arena is to allow the player to learn and master the abilities of specific Heroes in a safe environment, receiving feedback to help them improve.
  • Customisable: The player is given the freedom to control and customise how they learn. AI in the Arena should support the player in learning their Hero's moveset, as well as helping the player understand and get better against specific Heroes.
  • Flexible and Scalable: The Arena should be as flexible as possible, allowing players to quickly and easily modify settings without needing to exit. It should also be designed in a way that allows for the addition of new features as the functionality of the Arena grows.

Information Architecture and Flow

Information Architecture and Flow

Information Architecture and Flow

Organising the settings

Once in the Training Arena, the menus needed to support a number of features. It was important to organise these in a way that made sense for players in order to ensure the information was easy to find and use.

Working with stakeholders, we separated and prioritised the features based on our discovery and research, which is summarised in the Red Routes diagram below. This enabled us to identify what actions players would most frequently perform, and ensure they were available with the fewest number of interactions.

Once in the Training Arena, the menus needed to support a number of features. It was important to organise these in a way that made sense for players in order to ensure the information was easy to find and use.

Working with stakeholders, we separated and prioritised the features based on our discovery and research, which is summarised in the Red Routes diagram below. This enabled us to identify what actions players would most frequently perform, and ensure they were available with the fewest number of interactions.

A UX diagram displaying the anticipated frequency of player actions

User research findings were synthesised into a Red Routes diagram, helping us prioritise actions based on their anticipated frequency.

We also worked to group similar settings together, with these falling into either general Arena Settings, or more specific Opponent Settings. This separation helped reduce the complexities of the menus and organised the settings in a logical way that tested well with players during usability tests.

Enabling quick practice

When considering the existing training tools, there were two major pain points experienced by players. Firstly, the current modes lacked the ability to launch and play quickly as each session had to be manually configured before launching, and settings were not stored between sessions. Secondly, players could not modify settings on the fly - they had to quit the session and relaunch each time. This made the whole experience very cumbersome.

With the Training Arena, I pushed to allow settings to be stored across sessions, meaning it was easy for players to pick up and play. Additionally, the team developed tech to support modifying settings within the session. All of this helped reduce friction and ensured players could focus on training.

To further allow players to access the Arena as quickly as possible, I wanted to limit the amount of configuration required in the front end. Ultimately, the only choice the player has to make is which Hero they want to play as. While they had default values, players could also select the opponent type - practising against a specific opponent was a frequent player action - and map.

Arena front end flow

The screen flow diagram for the Training Arena shows the player's flow through the various screens. Emphasis was placed on reducing decisions in the front end to provide the quick play desired by players.

The screen flow diagram for the Training Arena shows the player's flow through the various screens. Emphasis was placed on reducing decisions in the front end to provide the quick play desired by players.

Design

Adding new features to a live game

Working on a live game which had been in production for several years has it's challenges. Any new features had to fit seamlessly into the game, using existing patterns and tech. As the team had admittedly spent lots of time working on building the game, and not much documenting it, there was little in the way of UX guidelines or patterns libraries.

When I first started on the project, I went through the game and identified different patterns and UI elements that were in place. As the team used Axure for UX documents, I created a set of Masters (design templates) that I could reference and reuse to speed up wireframing and prototyping. These were useful as a way of documenting what was used already, and referencing them along the way.

An example file from Axure displaying the master elements created

Some of the Masters I created to support quick and consistent wireframes. These were useful as I always had a set of templates I could reference and use when designing UI.

Communicating design internally and externally

Working with the team internally at Gobo, as well as external stakeholders at Ubisoft Montreal, it was important that everyone was kept up-to-date on the UX design for the project.

At Gobo, we printed out all screens and stuck them to a whiteboard in the team space. Alongside Jira, this was our primary way of tracking the progress of UI tasks and helped everyone working on the feature to maintain a shared design vision. We also collated stakeholder and usability feedback here too, keeping everything in one place.

A whiteboard with wireframes attached

The wall used to communicate the design vision to the team, as well as track tasks and feedback

To communicate designs with the team at Montreal, I had weekly catchups with the Presentation Director to go over the current state of the design and gather feedback. In between these weekly meetings, I kept the latest wireframes and prototypes available in a shared folder so anyone could access at any time. We also held frequent reviews with the game director and feature leads.

Prototyping

As I was using Axure for all design documents, this allowed me to quickly turn static wireframes into clickable prototypes which could be tested and iterated with players, as well as forming the key deliverable for feature sign off. These detailed prototypes allowed us to quickly iterate and nail the optimal experience, long before engineers even typed a line of code. This meant that once the UX was signed off, any subsequent changes would be small updates based on usability feedback, ensuring we kept churn to a minimum.

With Axure, I was able to create robust and representative prototypes. Using global variables and basic logic, it was straightforward to simulate state changes and flow, enabling us to really validate the experience outside of the engine. The team at Ubisoft also developed gamepad support, meaning we could even prototype and assess the controls at this stage.

My typical Axure file contained all user flows and wireframes, along with the prototype. This made it easy to collaborate on the file, even with the distributed team.

Axure Prototype setup

An example of an Axure workfile for the Training Arena. Global Variables (the popup) and OnPageLoad events (right-handle panel) allowed me to script basic logic and produce representative prototypes that helped make decisions before implementation took place.

An interactive Axure prototype demonstrating the final Training Arena experience, the detail helping us refine the final product and iterate rapidly with stakeholders and players.

Key Features

Overview

With the player needs clearly defined, and various pain points in mind, we came up with a feature set for the Training Arena that attempted to address these. My role was making sure we delivered these features in a way that felt intuitive, were consistent with the wider game, and removed as much friction as possible.

Below are the key features, and the UX considerations that went into them.

Practice Bots

Allowing players to practice against specific Heroes was one of the most important needs to address with the Arena. We often referred to this as the Raider's Unblockable Zone Attack problem as this was a move that had a specific timing and unique properties that gave the community a lot of trouble.

We went through many different designs, and even crude prototypes. In the end, we addressed this problem by letting players select moves from the opponent's moveset. This was chosen as it was the most accessible approach - it didn't rely on players knowing precisely which moves they struggled with, nor did they need to manually perform them (which wasn't always the easiest thing to do!). We listed the move name and input to help players have a better idea of which move it was they wanted to practice against.

Of course, with enough practice of a single move on repeat, anyone could get the timings right. To help players really get comfortable dealing with tricky moves, we expanded this to allow them to enable up to 4 moves for the AI to perform. This introduced more randomness into the scenario, and really helped players get better, without adding too much complexity to the menu. Players could also turn on the full AI, and set the difficulty, all with a few quick changes.

opponent_settings

From the Opponent Settings menu, players are able to customise the AI they're practising against.

Opponent settings and comfort features

With the inclusion of each Opponent Hero’s full Moveset alongside a number of common moves, the list was becoming larger and larger. To help ease the difficulty navigating this, I broke moves up into categories to allow players to filter and locate these more efficiently. Categories were selected based on the types of moves players most frequently wanted to practice against.

Additionally, to help players manage the moves selected, we added a persistent panel which displayed all selected moves. This not only provided an overview of which moves the AI was performing, but also included functionality to allow players to deselect moves directly from this panel, rather than needing to locate them again within the moveset.

opponent_moveset_

The Filter is anchored to the bottom of the Moveset and allows players to more efficiently find moves. The Selected Moves Panel is on the left, showing all currently enabled Moves and allowing players to deselect these individually.

Move Tagging

To help new players learn how to perform a Hero’s moves, we added a move tagging feature, something seen in many fighting games. Players can tag moves directly from their Moveset, and these are then displayed in the HUD.

We also added simple feedback to augment this and aid players in learning. In addition to flashing each input as it was made (to help the player track where they were in a sequence), we also had a final flash to signal when the full move had been performed. This helped expose some of the trickier timings, and confirmed to players they had succeeded in performing the move.

As this new Move Tagging feature made use of the existing Hero Moveset page, we needed to make sure this new functionality was discoverable to players. To solve this, by default, 3 out of a possible 4 moves are already tagged when players launch the Arena. This helped better expose and introduce the feature to new players.

An example of moves tagged to the HUD, and flow for updating these.

An example of moves tagged to the HUD, and flow for updating these.

Changing Hero

One hard tech restriction we had for the Arena was that players could not change their Hero without ending the session. In the initial design, for players to change their Hero they needed to quit to the Arena menu manually, and then relaunch. While this achieved the basic function, the lack of a clear call to action meant that players were not always immediately sure how to complete this fundamental task.

Working with engineers, we added an explicit "Change Hero" option to the menu. Additionally, rather than simply returning players to the Training Menu, we developed additional tech to load directly into the lobby, removing more steps from the flow.

The final Change Hero flow in game, demonstrating the minimal steps required. Though new tech was needed to support it, it was a massive win for players.

Play any Hero

As players work through For Honor, they can invest currency to purchase and unlock new Heroes. While many Heroes were available to play even if not unlocked, Heroes released as part of the DLC could not be used unless purchased.

With the Training Arena, we saw an opportunity to let players play as any Hero in the game. This allowed them to test drive the Heroes before committing, creating a better experience for players, and a stronger drive for them to purchase Heroes they enjoyed practising with.

game_lobby

Players are free to select any Hero in the game, whether they own them or not.

Quick Settings

While we had designed the menus to be as efficient as possible, I saw an opportunity to make changing the most common of settings (those highlighted in the Red Routes diagram, for instance) even easier. 

Repurposing the existing Quick Chat widget, the most frequently used settings could be modified here without ever needing to access the game menu. This was certainly a feature designed for more advanced players, but offered even more flexibility when practising.

training_arena

The expanded Quick Settings menu is displayed at the bottom left of the screen. Players can toggle between various settings without needing to access the full Arena menu.

Onboarding redesign

Another part of the puzzle

While the original onboarding went some way to prepare players, it simply wasn’t thorough enough to teach the full combat system. Moreover, the structure – perform an action three times and then move on – lacked the context required to really help players not only learn how to play, but understand when and why they should use the different mechanics.

I worked with the Trials team to deliver the new onboarding experience. This involved assisting gameplay designers with how we introduce skills to new players, consulting on the pacing and complexity of what we taught. Additionally, I designed the UI and flow for the mode, including custom HUD elements and new result screens that served to encourage players to engage with the tutorial.

The final design, called Skill Trials, included contextual scenarios which challenged players to not only learn skills, but apply these to reach the end of the trial. The team also designed a scoring system to provide additional feedback to players on how well they had done, giving an extrinsic reason to want to replay and improve.

A low barrier to entry

Based on the design of the trials, we wanted to have a really tight loop to prevent players dropping off due to frustration, as well as supporting replaying. Selecting a trial in the front end menu loaded players directly into a scenario, bypassing Hero selection.

For first time players, they would be launched directly into the Apprentice Trials once the game was booted, without navigating the front end menu. This removed all barriers, and gave the best chance that players would complete the content. As successfully completing the Trials was so important for retention, the option to quit was not advertised from the grading card for first time players, and was only available as part of the pause menu.

Trial flow

The user flow for the Apprentice Trials, including the First-time User Experience, minimising steps and allowing players to quickly replay to improve their rating.

Encouraging replayability

To emphasise the importance of the skills being taught, the designers came up with a scoring system. Players were awarded points for landing attacks, and bonus points for performing specific skills in each trial. This required a new HUD elements and feedback.

This went through a number of iterations before setting on a progression wheel which fills during the Trial. This could also communicate how well the player was doing, and how close they were to their goal. 

trial_gameplay

The HUD widget shows the player optional objectives and their score (top right), with the radial progress bar showing how close they are to achieving an A grade.

At the end of each trial the player is presented with their grade representing how well they'd done. From here they can continue to the next trial, or replay the current one - supporting the ability to very quickly try and improve their score. This was optional, next gating progress for players,

A wireframe showing the Grading Card for For Honor's trials

The design for the grading card emphasised the player's Hero to create a celebratory moment, while prominently displaying their grade and options to quickly retry or continue.

To further encourage players to replay and improve, we included their progression on the Training Menu at a high level, meaning they could see how well they had done without needing to dig deeper into the architecture.

training__1_

The Training Menu clearly shows the player how close they are to "completing" a set of trials. This extra progression helped enough players to replay in order to reach 100%, and learn the skills along the way.

Layering on additional feedback

One of the biggest challenges when it came to teaching some of the deeper mechanics of For Honor was exposing the specific timing required. Skills such as Dodging, Parrying and Counter Guardbreaking all have small input windows which required dozens hours of play to become comfortable with. For new players we wanted to expose some of this so they could more quickly build familiarity.

We felt that augmenting the existing For Honor HUD would be the best bet. In this approach we would highlight the existing HUD elements with additional feedback to draw the player's attention toward them, or new widgets that expose information that isn't otherwise there. This would act as training wheels that could be removed gradually without impairing the player's ability to understand and learn the actual HUD elements.

Highlighting HUD elements was straightforward enough, but actually exposing timing windows proved trickier. We went through several iterations on paper, before settling on a timeline-based approach I proposed (and can be seen in the video below). This felt good in game, though we switched to a semi-circle (as attack windows were often counted in milliseconds and progress moved too fast), and then finally to a needle to provide more feedback to players.

The original storyboard for the Opponent Typing Feedback, followed by the final implementation.

"For Honor would have significantly benefited from these features at launch. These new modes are a far more successful attempt at bestowing a player with the required knowledge."

"For Honor would have significantly benefited from these features at launch. These new modes are a far more successful attempt at bestowing a player with the required knowledge."

"For Honor would have significantly benefited from these features at launch. These new modes are a far more successful attempt at bestowing a player with the required knowledge."

"For Honor would have significantly benefited from these features at launch. These new modes are a far more successful attempt at bestowing a player with the required knowledge."

- PC Games News

- PC Games News

- PC Games News

- PC Games News

Reflection

Conclusion

At launch, the new training tools were well received by the community and critics alike. It may have been the first time I've seen players get excited for a tutorial! Praise was leveled at the depth of the trials, as well as the extra feedback added to help expose the deeper mechanics. New and expert players alike also loved the utility of the Training Arena, which they saw as providing a range of tools that supported them from mastering the basics, through to perfecting advanced playstyles at strategies at the highest levels of play.

The new training features were released to coincide with a free weekend. Of players that played through the Trials and spent time in the Arena, 79% ended up converting and purchasing the full game, a large improvement from previous free weekends and a testament to the success of these new features.

Internally at Ubisoft the format of the modes were seen as a benchmark for the quality and depth of new training features across all of their games.

Designed with somewhere between Brighton and Vancouver