Hello All,

I am currently working on a game project for Nintendo (as an indie developper) and I am currently working on the level difficulty balancing.

I wonder if there is any existing methodology to do it, except from playing the game for many hours.

I know that game balancing mainly depends on the game developped, and that there is not "true science" when it comes to balancing, but some of you may have already worked on this aspect of game development and may have hints or methodologies.

Thank you for your help.

Regards,

Baptiste.

In general, games become more difficult as you progress further, but this is not necessarily a linear progression.

If difficulty progression is slow and levels stay easy for too long, players get bored and stop playing.

If difficulty progression is fast and levels get difficult early on, some players will give up and claim the game is too hard.

The reason that players stop playing, no matter whether the game is too difficult or too easy, is because they have not been able to keep an interest in the game long enough to become invested emotionally. They've got nothing to lose by quitting early.

This means that getting the difficulty progression right in the early stages of the game is, commercially speaking, more important than getting it right later on.

When considering difficulty progression, take into account the following:

1. Player skill level

All players come to the game with a differing skill level. Good progression would cater to lesser skilled players, whilst still creating interesting challenges for more advanced players. It would be good to create challenges that would improve the lesser skilled player's abilities so that at a certain point in the game, the majority of players would all be at a similar level.

2. Difficulty spikes

I believe that difficulty should come in stages, with lulls in between. If we measure difficulty as a number from 0 to 9, where 0 is intuitive to all, and 9 is attainable for the most skilled players, instead of doing this:

game--difficulty--timeline------------>

0-1-1-1-2-2-2-3-3-3-4-4-4-5-5-5-6-6-6-7-7-7-8-8-8-9-END

I would recommend something like this:

game--difficulty--timeline------------>

0-1-0-1-2-2-1-3-2-4-2-3-3-5-4-6-3-4-5-7-6-6-4-7-8-6-8-7-9-7-6-8-END

Notice how the difficulty constantly drops down after a new challenge, allowing players to get a feel for their new skill/ability and further develop in into how they play the game.

Also notice that the most difficult part is around 90% of the way through the game. I think it is good to let players play on after the most difficult challenge and boast a little, show off their skills in the rest of the game.

All of this just my opinion of course, and I'm sure others will offer theirs too.

As MooseDev wrote, it is highly dependant on your game.



In multiplayer balancing, competition time is key to balance.

Generally early in the competition you want a stable game, make it easier to catch up from behind with an unlucky situation, and also find ways to make it more difficult for the most advanced player to leap ahead. Slow growth is useful here as small corrections have big effect. Slow growth is useful at the beginning as it reduces the effect of random chance, yet still allows an extremely advanced player to win over a novice. Early game you want the game to self-stabilize.

Generally mid-competition you want to enable powerups and strategy. Nothing that is too powerful, but plenty of power in whatever balancing factor you are using. Combos, powerups, whatever, should encourage a wide range of strategy. Details are unique to each game. Often you want to build a "perfect imbalance", where every strategy has both strengths and weaknesses, where every strategy can be overcome by another strategy.

Late in the competition you want the game to destabilize, one player to surge ahead as the others quickly drop off. This generally means high power and major events, big enough to destabilize the game. Players who are behind will fight with all their might to keep it stable, but you want the game to throw itself out of balance rapidly.

In games that rely on cards/crafting/building/training units it also depends on the combination of results a player could have like a platoon of units against other combinations of that resource and how single resources effect certain situation.

For example in a tower defence game you need to balance the towers, enemy waves and side effects so that more pwerfull towers appear later in the timeline but also more difficult enemy waves with e.g. more health so that there will never be a missbalance between the tower damage and the enemys that may appear at the time you would first achieve that towers.

Same for some other scenario like in a build up strategy game where the result of combats depends on player units against enemy units, strength, armor, health e.g. upgrades that may been have made to the one or other side. Here a single value of one single unit could change the game as seen in the early days of Warcraft 3 (Blizzard released certain balancing patches for example adjusting Riffleman's damage)

The third example from my job is ctiy building genre games. Here you need to balance the gain of resources, growth of your citizens and environment so that the fischerman for example dosent feed the whole village where half the people die when using a farmer to feed them. How fast are childs born in relation to the food someone could gain and so on.

I'm using tools for such balancing tasks that are almost self wirtten depending on the game I want to balance. Such tools calculate every possiblilty in combinations of resources at certain time with other combinations on the same time in the timeline (for example the time when both player and AI have setup there first platoon of melee units or how long it take to get to this timeline with certain race played) and then compares a middle between the worst and the best case to see if something needs to be adjusted

A second balancing option is to let testers play the game and gather there feedback after you rough balanced. While each player may have its own style there will be a wide range of feedback to work with.

Hello guys,

Thank you for your fast and very interesting answers.

The reason that players stop playing, no matter whether the game is too difficult or too easy, is because they have not been able to keep an interest in the game long enough to become invested emotionally. They've got nothing to lose by quitting early.

I totally agree. One of the most important thing for a game is to keep the player interested in the game, especially at the beginning. This is why I choose to let the user the choice between several levels every time he finish one (the level selection menu is a grid where each level is represented as a square ; When a level is completed, the adjacent squares can be selected and played). This can allows the user to keep interest in the game by giving him different choices (I hope so haha), and this can be linked to your "1. Player skill level" point, allowing differently skilled player to find different challenges, even at the beginning of the game.

I would recommend something like this: game--difficulty--timeline------------> 0-1-0-1-2-2-1-3-2-4-2-3-3-5-4-6-3-4-5-7-6-6-4-7-8-6-8-7-9-7-6-8-END

I think you hit a good point here. Even if the level selection will allows the user to be kind of free, I want to add levels which can only be completed when having specific objects unlocked. In this case, I can use this kind of Level difficulty pattern, allowing the user to access easier levels after completing harder ones. This can give the player an "almighty feeling", which is very good. Thanks for this very interesting point !

Generally early in the competition you want a stable game, make it easier to catch up from behind with an unlucky situation, and also find ways to make it more difficult for the most advanced player to leap ahead. Slow growth is useful here as small corrections have big effect. Slow growth is useful at the beginning as it reduces the effect of random chance, yet still allows an extremely advanced player to win over a novice. Early game you want the game to self-stabilize. Generally mid-competition you want to enable powerups and strategy. Nothing that is too powerful, but plenty of power in whatever balancing factor you are using. Combos, powerups, whatever, should encourage a wide range of strategy. Details are unique to each game. Often you want to build a "perfect imbalance", where every strategy has both strengths and weaknesses, where every strategy can be overcome by another strategy. Late in the competition you want the game to destabilize, one player to surge ahead as the others quickly drop off. This generally means high power and major events, big enough to destabilize the game. Players who are behind will fight with all their might to keep it stable, but you want the game to throw itself out of balance rapidly.

This three step method view is also pretty interested. I hope I understand what you mean : The player starts to play the game without many startegies, in order to understand the main mechanics. Then, there is the phase where he needs to improve his skills. He will have to really developp startegies and fine tuning to beat these mid-levels. And to finish, the game will add new mechanics or change current mechanics enough to force the player to find new strategies.

I like this point of view too, and I think I can use it in the game I am working on. Aside from the main level difficulty ramp up, I want to add scoring elements. I think I will use the last of these three steps in the scoring aspect of the game. The player will still be able to finish the game without any destabilization, but if he wants to masterise the game (by going into the scoring thing), he will need to break his habbits and find new strategies.

Thank you for your answer guys. It helps me a lot !

Thank you for your answer. This is my first post on this forum and I wasn't expecting such interesting answers.

Everyone has views about balancing, and it is interesting to get the point of view of people who are in the video game industry since a long time.

I'm using tools for such balancing tasks that are almost self wirtten depending on the game I want to balance. Such tools calculate every possiblilty in combinations of resources at certain time with other combinations on the same time in the timeline (for example the time when both player and AI have setup there first platoon of melee units or how long it take to get to this timeline with certain race played) and then compares a middle between the worst and the best case to see if something needs to be adjusted

I think I will have to use my own tool ,as you do. The game I am working on isn't directly related to the three gameplay example you have used, but It has many aspect of these three gameplay type.

To be more precise, I am working on a shooter with RPG and Social relations dimensions. RPG aspect (like player level improvment, specific objects and power up use, and a big range of different enemies) has game mechanics related to tower defense or strategy games, while Social relations aspect (going into in game pubs, clubs and many other places to recruit new members) will be closer to the city builder gameplay (where ressource growth of relationships can be linked to growth of citizen).

But I will definetly have to use some home made documents, allowing to calculate "easily" the power of each specific objects, at different skill levels.

A second balancing option is to let testers play the game and gather there feedback after you rough balanced. While each player may have its own style there will be a wide range of feedback to work with.

This balancing option is indeed used by many developpers. I will probably do it at a small scale with familly and friends, but it will be like fine tuning for me, as I would like people to play the game like I have imagined it (creating emotions where I want emotions, and so on....) .

P.S.: I hope that my English isn't too bad for this forum. English isn't my mother tongue so I hope that you understand what I mean :)

Thanks for your help anyway !

I see you said you're working on a shooter rpg. The main thing that comes to my mind in terms of balancing is the weapons. What if one particular gun is completely overpowered? Your other guns become poor choices and your overall diversity of weapons suffers because of it.

But unfortunately I'm not familiar with balancing shooters. I have studied a bit about fighting game balance (balance is hugely important and very complicated for those games). I'll say one thing. In fighting games, one particular type of character tends to be overpowered the most. This is the character that literally can do everything, but has low life to compensate for it. People make the error of thinking that giving a character low life will balance out all the powerful tools it has. But this often tends not to be the case, and their powerful options more than make up for their low life, and they often become overpowered.

Maybe that relates to something in your game or maybe not. Balance is always an interesting topic :)

I wonder if there is any existing methodology to do it, except from playing the game for many hours.

All you have to do is test game balance at each level of progression through the game - whatever "level of progression " means for that game.

A formula based design for difficulty progression can work well. Ideas like introduce one new type of opponent or terrain feature per level. Or maybe that's imbalanced - too fast. So introduce a new one every other level. Or x amount of total enemy hit points per level. Of course this sort of requires a level based game to begin with, but it can be applied to RPGs as well. Skyrim introduces high level magic items spells, and opponents based on player level (IE experience level), as opposed to arcade/shooter level (IE dungeon level).

Hello guys, thanks for your answers.

The main thing that comes to my mind in terms of balancing is the weapons

You are right, weapons is one of the aspect balancing will have an impact on, as long as ennemies, level layout, and many other things. Balancing everything is going to take a lot of time. Moreover, as level selection will be free, I need to be sure that the player will not be blocked if he can't unlock another weapons (due to lack of money or something else) but need these locked weapons to continue to play.

I have studied a bit about fighting game balance

When it come to fighting game balancing, Rise of the Robots (on Megadrive) is clearly a bad example. The Red Robot is really really too strong compared to all other caracters.

Of course this sort of requires a level based game to begin with

There will not be a level progression (at least for the moment). People will have to modify their spaceship caracteristics to be able to beat stronger ennemies. But you are right, getting new gameplay mechanics can still be related to the skill of the player, as they will be implemented in later level/worlds, only reachable after beating some specific ennemies.

Thanks guys :)