A flash of steel, some fast slashes, and the enemy is dead. This is an illusion of how the samurai fight. Deadly, accurate, fast … but how does it translate into gameplay? Respecting this illusion, the so-called “lethal contract,” was a major challenge in creating the Ghost of Tsushima.
Keeping things deadly was a pain. Assassinations and standoffs nailed it, but what about standard combat? Lethality had to last throughout the game. That meant it had to work for:
- Armor, charms, weapon upgrades, techniques, Ghost of Tsushima: Legends gear scaling for over 30 hours of player progress
- Easy to Lethal difficulty, Ghost of Tsushima: Legends Bronze to Gold difficulty, a wide range of player groups from Nightmare / Raid challenges
- Combat experience with different enemies to keep players interested
- Responsive and fast-paced gameplay
The number of moving parts that had to be pinned to do all this work (and still make it fun) was staggering. Welcome to my job! I’m excited to be able to elaborate on how I attacked these issues.
So tighten and break your sword, we go back in time.
“Sword sponge” problem
Challenges from enemies had to grow in parallel as players explored Tsushima and upgraded their weapons and skills. Weapon upgrades increase player damage, so we’ve adjusted enemy HP and damage accordingly. This was to increase the difficulty and show significant growth in the upgrade. All good ideas … but we have found a difficult way that there is a subtle line between showing growth in power and breaking the “lethal covenant.”
During the early playtests, I received very negative feedback that the enemy felt like a “sword sponge”. My favorite quote from the player was “I felt like I was attacking an enemy with a foam bat.” This was … not what we wanted to hear. Players clearly told us that their expectation for the sword was a “deadly contract”, without which the realism we strived for would have been shattered. Unexpectedly (at least initially) I tried many different versions of “hit points” with different wrappers like “armor points” but eventually failed.
We were hitting our head against a brick wall, and it was painfully clear that we had to change our approach. As a result, we forced all enemies in the game to have the maximum “hits”, even if they upgraded their weapons. This includes all upgrades to all enemies in the game and a large excel sheet of armor. I won’t bore you (although some of you may enjoy it).
This allowed the upgrade to be meaningful, taking care not to overwhelm the enemy’s HP. However, to show this power growth by investing in character art, we had to justify the increase in HP. This also allowed us to visually represent a wider range of enemy types.
But that wasn’t the only problem. The experience felt flat as we implemented more new types of enemies and reduced their HP values, as they died so quickly. We had to increase the difficulty in other ways without breaking the “lethal contract” with the player.
Increased difficulty and enemy behavior
Respecting “lethal contracts” has become the core of the combat experience and has become difficult to maintain, making it even more difficult to play Easy, Medium, Hard, or Lethal, as the enemy’s HP will increase further. I decided early on that the degree would not increase. In some respects, this facilitated other decisions by having a fixed point to work on, but it also made the difficulty starting with the enemies trying to protect their low HP pool. It meant that I had to take advantage of other factors to raise it.
Blocking and parrying players was the main way to protect HP, but the first prototype was a bit overkill.
It was beneficial to have the enemy launch an attack, but it was too frequent. I felt the enemy couldn’t invade and it wasn’t fun to fight. The fighting became overly reactionary, just waiting for parry and strikes … this wasn’t what we wanted either. We wanted to give players an offensive opportunity and not always be passive as the most effective strategy. To achieve this, we added staggering damage to the heavy attack to break through the defense and allow the player to become aggressive.
As a result, the speed at which the enemy staggers has become an important factor in standard encounters. The time it took to stagger an enemy was the time that other enemies could attack, and was a central aspect of how to inject enemy diversity. This was especially true when adding high cadence attackers (such as dual wield swordfighters) to give them a more challenging experience. We’ve also changed the attack patterns of other enemies, the amount of time they shift for each type of enemy, and how much they can defend to further change their combat experience.
I’ve tried many ideas for new enemy attack patterns and defensive actions. Not everything was successful, but in retrospect it’s great.
Too defensive, looks ridiculous
Immediately added a stance as a way for players to express themselves more often and as another way to be more aggressive in combat. It didn’t take long to achieve the four core enemy archetypes that were directly mapped to the four stances we were building. Changed stance attack to stagger enemies twice as fast. This allowed us to encourage the correct use of our stance. This was a double victory. It fits perfectly with the fantasy of being a samurai, flowing between each stance and quickly astonishing and killing enemies only strengthened the “lethal covenant” we were trying to achieve.
Lethal Contract Bending: Duel and Legend
We knew that having a 1v1 duel had to be included in the game. It was a requirement that the two samurai face each other in a spectacular way. After firmly establishing the idea of a “lethal contract”, I started working on duels. Therefore, we kept the HP pool low. Not only was it sick, it was too cheerful and fast.
The Samurai Cinema metaphor is that two swordsmen face each other and are killed in the first attack, but we knew that the players’ expectations were different. These were boss battles with high challenges and fierceness. We had to bend the contract and respect this moment.
Increased HP, attack speed / damage, various movements, and added evasive actions. Players enjoyed it, some felt the drag was too long during the test, but was told that lowering the HP was too short. We were successful when we divided the duel into “phases” by a sword clash. There, each chunk of the fight had the right amount of HP, but nevertheless continued the fight for as long as we wanted.
Like Duel, Ghost of Tsushima: Legends demanded to change the maximum “hits to kill”. This was considered a mini-boss and was especially true for the mythical enemy, the demon. Like Duel, we increased our HP to meet those expectations. In addition, Ghost of Tsushima: Legends has the potential for 2-4 players to attack a single target with a wide range of abilities (many ignoring defensive actions) and needs to support gear Ki scaling. Therefore, it was necessary to increase the HP pool.
Connect everything: player skills, progress, and become a “ghost”
So far, we’ve mentioned a small subset of what was done to maintain a “lethal contract,” but there were many aspects that supported this concept with increasing difficulty. This is all of them, including what I have discussed so far:
- Enemy defense action (stagger meter, block, avoidance)
- Enemy attack speed, group aggression, combo length
- Timing required to block, parry, and avoid players
- A new set of enemy movements to learn
- Players resolve the benefits of skilled behavior
- Damage output from the enemy
- Increase the enemy’s HP (last resort)
All these items were also important in supporting different types of players and more axes for player growth. Experienced players will learn the combat created by enhancing all of the above aspects (such as parry timing), and upgrades (and armor and charms) will ease these restrictions and give them a sense of growth. I wanted to show you. Combining this with different difficulty levels allowed a wider range of players to enjoy the ghost, and there were multiple ways to achieve victory.
In the midst of deadly difficulties we pushed them all to the extreme, but interestingly, to respect the “lethal covenant” we had to make. To increase Damage from the player. Not only did it feel more fair, but whether it was a victory or your own demise, the battle could end in an instant.
I wanted to increase my power, but before moving on to the more difficult story, I didn’t completely anticipate how wide a range of players the players would play. Initially, we predicted a narrower playthrough of content based on how players participated in the game during playtesting, but in real life it was very different. As a result, many players “overwhelmed” themselves before encountering a more challenging story. Although not ideal, we prioritized this over the option of breaking a “lethal contract” when the player felt ineffective.
In addition, I wanted to empower the player to embark on the same journey as Jin and become a “ghost.” Assassinations, kunais, smoke grenades, etc. had to be very deadly, not only to be more effective than samurai, but to reinforce the reason for using them. It was acceptable to feel “overwhelmed” with these weapons, which fit better with the story we were trying to tell. I admit freely, but those kunai are a little too easy to get from the body;)
Huh! This was just the tip of the iceberg about how we worked on one of the pillars of Ghost of Tsushima’s combat balance, the “lethal contract.” I want to continue, but this must be done someday! Hopefully the next time you switch stances and destroy the spearmen, you’ll get a little understanding of what happened to make you feel like a samurai.