Game difficulty

[X3M]

Levels of difficulty

There are many different levels of difficulty. An example would be:

Easy/Normal/Hard

And it is used for many games.

 

Of course some games have like 8 different levels; Unreal Tournament:

Novice/Average/Experienced/Skilled/Adept/Masterful/Inhuman/Godlike

Of course Godlike from that still means notching if you are used to games of these days. Unless you know how to tweak the bots. :)

 

But are 8 levels useful for a game?

How much do you prefer?

Does it depend on the game(play)?

 

Changes in difficulty

 

Well, I am going to list some that I am aware of. Of course you might add some too if I missed one.

 

These are mainly based on RTS.

1- AI, the one where the AI actually plans ahead further and further.

2- AI, the one where they hate you more then other targets.

3- Game speed

4- Unit costs

5- Unit health

6- Unit damage

7- Specific unit targeting, type targeting or almost a dead unit.

8- Extra forces on the map

9- Build speeds

10 - Extra money

 

What do you prefer?

I like the 1, 7, 8 and 10.

1 and 7 are fair play.

And 8 and 10 are good map difficulties for a single player game.

 

Personally, I hate the difference in unit costs. You get prices like this:

90% out of 950 is 855.

But it can be worse. Once a game start rounding numbers, it really gets annoying.

 

 

[MrFlibble]

These are mainly based on RTS.

1- AI, the one where the AI actually plans ahead further and further.

2- AI, the one where they hate you more then other targets.

3- Game speed

4- Unit costs

5- Unit health

6- Unit damage

7- Specific unit targeting, type targeting or almost a dead unit.

8- Extra forces on the map

9- Build speeds

10 - Extra money

Come to think of it, this sounds a lot like handicapping in the traditional, non-videogame sense of the word.

I think the whole concept of difficulty very much depends on the actual playing mechanics of a particular game, so you could expand your list even further, adding more aspects of a game (even of you still focus on the RTS genre) that could be tweaked to affect the difficulty.

It should be also noted that in the list above, the changes might apply to either the player or the AI only, or to all players. For example, if you take C-evo, lower difficulty levels have such tweaks (like faster growth of cities because they have to accumulate less food to increase population), however these tweaks apply equally to the human player and the AI players. So the real challenge is in how well each player handles the situation in given circumstances, not in the handicap of one of the players.

On the other hand, the question of difficulty also depends on what goals the designers have in mind when they implement various difficulty levels. First off, of course, the difficulty level scale should contribute to the replay value of a game, but this is not always the only purpose of game difficulty. For example, I've read somewhere that some Japanese games are much harder in their home release as compared to the foreign releases, which is done for the following reason: renting video games is widely practised in Japan, and of course it would be more beneficial for such kind of business if the players were to spend more time playing the game they have rented - and the playing difficulty is increased appropriately :)

That said, I appreciate it if a game has several difficulty levels to pick from, so that one can replay the game on a higher difficulty. However, I'm not a fan of the Nintendo hard kind of difficulty when beating a level becomes plain frustrating.

I also find it interesting if different difficulty levels that are implemented in a game actually affect the game world on a vast scale. Realms of Chaos is supposedly an example to this (I haven't played this game a lot though, but it says so in the readme :)).

[X3M]

Nin-Ten-Do?

Sound like, (level) nine, ten, DOH!!

Yeah, they have some pretty hard difficulty settings out there. But during the game, you level up.

 

 

With RTS you also get harder and harder levels.

But I am more talking of the kind you choose at the very start. And you have to complete the entire game on that setting.

 

---

 

Some games have a good looking apply of difficulty settings.

Warcraft 3 has one of those:

Where missions are actually altered by replacing units with stronger ones.

Or the AI sends 6 units instead of 4.

In that case, less choices are left for a player, but new ones might open. With a little thinking, I got 4 examples.

 

Positive things on higher difficulties:

- Mind controlling a certain unit actually becomes more useful since the opponent is harder. (Devastator instead of a Rocket Launcher)

- A splash weapon that hits 6 units, now actually is able to hit 6 units. Or else it is just overkill and not worth your money. (Scarabs from the Reaver, from the game Starcraft) (Of course this only happens in custom maps where you can choose difficulty)

- If units become stronger by experience. This goes faster if you have to beat more units. Then some missions are actually done faster. (Hero levelling Warcraft 3) (Or C&C3)

- The enemy units have a higher rank. Beating them gives an extra bonus to your units. (C&C3)

 

---

 

Of course there are settings that make a game "ugly":

A fine example would be Dune2000 with 3 difficulty settings:

Easy with 75% costs

Medium with 100% costs

Hard with 125% costs

According to combat math, a 1v1 fight between 2 units. 125% costs is actually 2,78 times harder than 75% costs.

But that ain't the problem. I actually think adjusting prices would be one of the best ways.

 

But programmers need to apply it in a good way. For example: the Light Infantry costs 50 on normal. But 37,5 on easy and 62,5 on hard. They round the costs. Don't do that. Why didn't they use 80% and 120%? It gives better numbers.

Besides, I loose track on what I am doing. If I see that I have 3500 in credits. Then I can plan ahead with the units with round numbers. But if I have to spend the money on numbers like 875, 67, 1312 (1313 would have been cooler). Then I need a calculator.

[MrFlibble]

Besides, I loose track on what I am doing. If I see that I have 3500 in credits. Then I can plan ahead with the units with round numbers. But if I have to spend the money on numbers like 875, 67, 1312 (1313 would have been cooler). Then I need a calculator.

You're right about this. I don't know why Westwood decided to go for this kind of difficulty setting approach in their games. I think in RA2 already (maybe even in TS this was changed to the more reasonable ways of adjusting difficulty, but I'm not sure about that.

I remember I've tried replaying some of the Warcraft III and The Frozen Throne campaigns on hard difficulty, and found it frustratingly hard. Maybe I'm bad at micro, or whatever, but already in the original Alliance campaign (which is supposed to be the easiest one) I had trouble in the later levels. That wasn't fun at all. [i guess that this "hard" setting is probably a piece of cake for pro or even average MP players of WC3; but I never play MP games.]

You're also right that in games where the player can choose from a variety of weapons and strategies, various difficulty levels may imply making different choices about what to use. I especially felt this while playing The Catacomb Abyss, where on the "Novice" difficulty the enemies usually go down after a few shots of your default fireball weapon, leaving little or no use for the limited powered-up versions of the fireball attack.

One more thing I remembered about differences between difficulty levels: in Will Rock, on easy and normal settings monsters with projectile attacks (like the Minotaurs) will aim at the player's current location, regardless of whether the player is moving or standing, which makes dodging projectiles rather easy. On hard, they will take into account the direction of the player's movement and aim their shots accordingly, so you'll get hit unless you constantly change your direction of movement. I guess other games also have this feature (e.g. Serious Sam).

[X3M]

Ah yes. The AI of shooting in FPS.

Unreal (Return to Na Pali)

The enemy did the same by shooting where you stand. And later on trying to hit you where you walked.

On the level Unreal, some Skaarj actually spawned really a lot of energy balls. See it like rapid fire. Because that way you would die in the long run.

 

Tiberian Sun had only modifications in health, damage, build speed and number of attacking units.

 

Warcraft 3 on Hard.

You simply need to know what to do. Going for low upkeep instead of no upkeep.

Ehm, let me guess, that survival mission with the humans against the undead. Well, that is indeed the hardest mission in the entire game if you ask me. Even on normal difficulty I find that one harder than "hard" mission.

 

Try to have 1 defence point instead of 2 to 3 walls. Some towers with cannons, some towers with arrows, some mortar teams behind it. Have the mortar teams in 1 group and aim for his hero's. Then the small fry. For the other forces.Well, your forces need to be mixed up. Don't use Rifle men, they are only good against the Abominations in fair fights. Better yet, you have your towers with arrows for that. Kill the small fry first, they are there as support, so when they are gone, everything focusses on the big target that does little damage.

 

Further more, I don't know anyone who managed to get the extra quests on hard mode in that map.

[MrFlibble]

Haha, yes, the human survival mission gave me a lot of trouble on hard difficulty :)

I didn't figure out riflemen were bad (I guess they are in this case!), but I managed to beat it by "almost" cheating - instead of trying to save the base I built a few farms in the starting corner of the map (where you get three extra footmen if you go there), and then I think I also managed to build a town hall and some farms near the other gold mine.

Then I focused on using Arthas and some troops I could gather to break through the northern exit from the base that is only guarded by two spirit towers, and went for the orange caravans. More than that, I was also able to get the Ogre side-quest that yielded the Gauntlets of Ogre Strength or whatever its called :)

I barely managed to keep Arthas alive untilt he time ran out, while the undead were mopping up the main base.

Lousy strategy I guess but it worked :)

Then I beat the other missions with much less trouble, but gave up on the last mission after all.

As for The Frozen Throne, I couldn't get past mission 2 for the Night Elves :-/ Well, NEs have never been my favourites, but I enjoyed this campaign on normal mode a lot more than the original NE campaign in the vanilla game.

Come to think of it, I actually like TFT more than RoC on the whole, with the possible exception of the demo campaign :)

[X3M]

Human survival.

Of course, being a nomad on that mission also is useful. Indeed, that is how I, one time succeeded the quests. I had farms in each corner of the map. Enough money for buildings in each corner with 2 peasants moving around. While the undead moved as one trying to destroy my "bases" :D.

Don't attack the undead though, or they hunt down your army. And once they find your army. Let the damaged units, take another route :D.

 

TFT NE mission 2,

is actually a very easy one. Once again, if you play on hard. Just go for low upkeep. 70/70. In all the missions.

The trick with night elves is having a lot, really a lot of moon wells for extra health. Right at your defence. Front line units in the front. Support units behind the moon wells. Defeat the enemy half way. Then complete additional quests. I like the uneasy orcs in that mission ^^. Night elves are the best in the starting missions. Once you need to do micro with each attack, I hate them :D. O well, eventually I only use roar en rejuvenation.

 

RoC versus TFT

Even though TFT has more differences in units. The tactics actually plummeted. The need for medium and long ranged support units. And magic users actually become obsolete in TFT. In RoC, they all had plenty of durability. Especially the effects of Catapults/Mortar teams/Balista's/Meat Weagons on the "non armoured" in TFT is ridiculous OP. Just meat and some of those Catapults/etc. is enough. Instead of during the fight, I mostly heal afterwards in multi player games. Since those support units die to quickly.

 

Last mission Humans,

Yes, it's a though one in Hard. But only in the beginning when you don't have your hero there. Use your towers as early warning systems. But keep 1 army. Warcraft 3 is bad in defending multiple corners. Simply because different paths can't act like choke points. This due to only having 90 or 100 food limits. I Feel that Blizzard failed there. While in Warcraft 2, it was easy to create choke points. Yes, I build choke points with farms in WC2 :D.

[MrFlibble]

The trick with night elves is having a lot, really a lot of moon wells for extra health. Right at your defence.

Yeah, I figured that one out pretty quickly. But probably I didn't build enough of them anyway :)

RoC versus TFT

Even though TFT has more differences in units. The tactics actually plummeted. The need for medium and long ranged support units. And magic users actually become obsolete in TFT. In RoC, they all had plenty of durability. Especially the effects of Catapults/Mortar teams/Balista's/Meat Weagons on the "non armoured" in TFT is ridiculous OP.

Hmm, what I meant was the SP campaigns. I don't think that deep into balance :)

RoC campaigns, in my opinion, are mostly rather bland. Only in the undead campaign you have some variety, when playing against the unique high elves instead of the ordinary alliance, or against the "old Horde" in the Alterac mission.

The Chaos orc missions were quite fun a for a change in the orc campaign, and yeah, the dungeon level is pretty good too.

The original NE campaign wanders too much, and too often, into the "stand against overwhelming odds with limited resources" territory. Difficulty's not the problem here, it's just repetitive.

[X3M]

True, it has a lot of repetitive missions. But are there any unique missions lately?

For me, everything is repetitive. The only thing that changes are the units.

 

Which game has a unique mission? The 6 same things I always see:

- Destroy base (everything)

- Destroy specific unit

- Defend base (everything) (or whatever you have to build)

- Defend specific unit

- Take over ... , conquer ...

- Rescue ...

Leaving the tutorials out.

 

Only red alert had some "spy" missions. But basically, that was, defend specific unit. Then "take over" a certain structure.

 

So, who can name something unique?

[MrFlibble]

Well, to be fair, you don't need to come up with something really unique every time you need to break some repetitive patterns. Most of the time, it's not about some genuinely new ideas, but about properly mixing existing ones so that the result doesn't come off as boring.

While you're right that the mission objectives in most RTS games boil down to your list, variety is accomplished by modifying other factors, such as starting conditions (e.g. the "no base" missions introduced in Warcraft: Orcs & Humans), the presence hero units, a time limit and so on.

One major factor in SP campaigns IMO is the races you have to play against (if there are more than two factions in the game that is). If you analyse the campaigns in either Starcraft or Warcraft III (plus respective add-ons), it's easy to notice that the campaigns for different playable races vary a lot in respect to how many different races the player gets to fight.

The problem here is that the SP campaigns have a story to tell, so you can't just balance the game to have an equal number of missions against each race, or a combination thereof. Plot considerations are very restrictive, and in some Warcraft III campaigns this has also resulted in repetitive missions.

[X3M]

Well, if there is one thing that I always hated. It was a time limit in which you needed to achieve a certain condition.

Almost every mission in Warzone2100 has this timer. For me the mission wasn't really that fun. But it was to tackle the infinite income issue (you take money with you to the next mission). And it was to push players forward, keeping the game "real". Your enemy in another map doesn't really wait for you.

 

About starting conditions and during the game, the resources:

Dune 2 has limited money on each map. This is a good thing.

Dune 2000 has unlimited money on each map. This is a bad thing. The regrow is just too fast.

Each game has it's own income rate. And of course spending rate. I find a game more of a challenge if you have only little bits to spend every 30 - 60 seconds. You need to explore and/or defend specifically. Deciding on every penny. Most game these days ask you to build a lot and fast. No fun if you ask me. But it is a personal thing.

 

Warcraft 2 and 3, and of course starcraft have this. Perhaps that's why those games are good.

 

So if it comes to income. There are several classes:

Resources on map:

Limited/Unlimited

Resource income rate:

Linear/Dependable on regrow rate/Exponential

 

Generals has Unlimited with Exponential when you have the maximum tech level.

C&C 3 has Unlimited with Dependable on regrow rate.

Dune 2 has Limited with Linear. After all, you wont build 100 harvesters.

Dune 2000 has Unlimited with Dependable on regrow rate. But this regrow rate is very fast. You could say that your income is Exponential until you reach the entire map. About 20 Harvesters?

 

Of course one of these combinations has to be combined with the spending of resources.

There are several spending classes:

Number of unit limitations:

Unlimited/Supply limit

Build rate:

Linear/Dependable on production structures/Dependable on resource income rate

 

KKND is a funny one, Resources are Limited and the income rate is linear. Just like Dune 2.

But Dune 2 has Supply limit with a build rate of dependable on production structures and dependable on resource income rate.

While KKND has Unlimited supply and while it is also [dependable on production structures and dependable on resource income rate

], it can build every unit at the same time. Meaning, you can build 3 different tanks at the same time, but you can't build 3 of the same tank at the same time.

[MrFlibble]

Rise of Nations has an interesting resource system. While the resources on the map are unlimited, there is an efficiency cap for the collection of each resource, which depends on the player's current technology level. This is also combined with the pretty standard unit limit (in RoN it depends on the military tech level, not on the number of "houses"), and, on the more innovative side, the progressive increase in production cost of each successive unit of a certain type. Meaning that the more units or structures of a given type you have, the more you are going to spend on building even more of those. I find this an interesting alternative to the upkeep thing in Warcraft III, but RoN is generally much more liberal regarding unit numbers, and in fact encourages large-scale battles.

There is yet another intriguing resource feature in some RTS games, namely the "honour points" that are spent on certain key structures, research or special units (or heroes), but these points can only be obtained through victories in combat. This way the player simply can't expand and/or develop without having to engage the enemy on a regular basis. Rising Kingdoms uses this, but I think also other RTS titles have a similar feature.

[X3M]

Ah, yes, Generals has that kind of points as well. But then they count as experience for the commander. It's a grey area.

 

But it does not really apply to difficulty settings. These points make a game easier?

 

---

 

Edit:

I wonder if there are some sort of formula's possible to see differences in difficulty. I am going to try to put something simple here. Some games have their difficulty in % build speed. But is it true?

 

When you have twice as much units, you destroy your enemy twice as fast. So they only do half the damage. A force real strenght is always squared.

 

% difficulty squared = real difficulty.

That means 20% is only 4% and 60% is only 36% difficulty.

Have game designers taken this in account?

 

---

 

There are many more factors to be concerned:

Weapon effectiveness, which one does the AI use?

Will the AI adept to the player in choice of units?

Will the AI scout for this? So fooling the AI is possible?

Will the AI do micro?

Will the AI focus on certain targets?

 

---

 

Ow, here is a fun fact. The first micro AI ever would be the NOD Buggy and NOD Light Tank from C&C Dawn.

They drive away from grenades.

[MrFlibble]

Ow, here is a fun fact. The first micro AI ever would be the NOD Buggy and NOD Light Tank from C&C Dawn.

They drive away from grenades.

You're right, except it's "Nod", not "NOD".

More units doesn't necessary mean you defeat the enemy faster. I understand your example was rather simplified, but nevertheless there are usually many other factors that affect how successful you are in a mission. Same applies to more ammo/health packs etc. in other genres.

I'd rather say that if you have more units, you simply have more chances not to lose compared to the situation where you are more limited in your resources.

Take a very simple RTS example of mission difficulty: starting resources and units. Provided that the AI builds up its army and times its attacks the same way on all difficulty setting, more resources means that the player has more breathing room before the first attack of the AI. Less resources or units means that the player has to be more skilful in quickly establishing their own economy, so that when the AI attacks they will be prepared to successfully counter this attack.

Examples like this can be multiplied of course, and usually there's more than one way to increase difficulty (e.g. less resources + more units in AI's attack teams etc.). However, I believe that the only reliable way to really measure difficulty settings and then compare them across different games can be accomplished by having a number of test players (preferably whose playing skills are more or less equal; or divide them into groups according to their skill levels) play the game(s) on various settings, and then just compare the outcome (victory or defeat in each case for each player, how quickly victory was achieved or the player was defeated; plus the subjective opinion of the players on how easy/hard it was for them to play) to get a statistically valid picture of how difficulty settings as designed by the developers actually differ.

Then for example, if all the testers have no trouble winning on normal but none succeeded in a single hard mission, maybe it's a good idea to adjust some factors that affect difficulty settings after all.

Basically, I suppose this is how games are tested in real life, except maybe there are only rough statistics in many cases. Or maybe not, maybe the statistics are more detailed, IDK :)

[X3M]

Didn't know about NOD, is actually Nod.

I guess those who make the mistake do this because GDI remains GDI. :)

 

---

 

The simplified point of view is depending on the starting resources and enemy attack strength. That is true. But if you play skirmish for example. Only the enemy can be adjusted to some extend. While keeping the game "fair" for all players that join in.

What I see in games is that easy attacks with 2 units, normal goes with 4 and hard gives 6 in an attack. I believe this happened in C&C Tiberian Sun. Let's use this as example.

 

The point is, how do the 3 settings compare? The programmer who thought of these settings clearly misunderstood how a combat flows. Given the fact that most games go chaotic when a player does not pay attention.

As mentioned before, when 2 units attack, they are 2x2=4 strong. When 4 units attack, they are 4x4=16 strong. And 6 gives the number 36.

So the real difficulty is 4, 16 or 36. (If it is a tactical combat flow, we get 3, 10 or 21 as scores)

 

The difference between 2 and 6 attacking units is not a factor 3. But a factor 7 to 9.

Another way to look at this is in % difficulty.

25%, 100%, 225%. (30%, 100%, 210%) That are some big differences in difficulty. Don't you agree?

 

If the programmer really wanted 50%, 100%, 150% as intended. He/She had to calculate back to the original number of units.

For easy I suggest using 3. Since that is chaotic 9 or tactical 6 as score compared to the 16 and 10. The percentage is 56,25% to 60%.

For hard it is obvious that 5 is the number. 150% to 156,25%.

 

---

 

The same goes for the number of waves. I often designed 2 parallel bases in SC1. Where one attacks with 1 big army. And the other one attacks with small waves that follow each up rather fast. Just to make a game more interesting.

 

What I discovered back then was that the smaller waves where 15 times 1 unit. While the big wave was 1 time 15 units. Only the big wave was really dangerous. The smaller waves where not even annoying.

 

Back then I started calculating. And since those maps had really a lot of tactical game-play. I used the sum formula to design proper attack waves. 1 times 15 units had a score of 120. While 15 times 1 unit actually had a total score of 15. This means that the big army was indeed 8 times more effective with the same resources.

 

To balance the 2 armies. I simply had to increase the smaller waves. I got a complete menu of choices now:

120 times 1 unit.

40 times 2 units.

20 times 3 units.

12 times 4 units.

8 times 5 units. etc. Or any combination of these.

 

The only remaining problem until today is the resources spend on those units. That could not be balanced any more.