Support NeoGAF

[PotionBleue]

Monster Hunter games have a panic dive that happens when you're running away from a monster with your weapon sheathed. Instead of doing a normal evasive roll, it makes the character do a seemingly useless slow moving dive into the air but in reality actually has the most time of invulnerability of any evasive move which allows "close calls" to happen during moments the player feels unsafe.

It's not really a hidden mechanic, there's even an armor skill (Constitutn) that enhances this specific kind of jump.

 

[Shifty]

This is incredible. What on earth!

Lesson one in the creative arts: Your client does not know what they want.

 

[KDR_11k]

I-frames on dodges in general count as this IMO because very few games document them.

Non-linear health does feel obvious at times, I think I did get a strong impression of the lower parts of the HP pool being tougher in Doom (in addition to the much bigger refills when low). Many games of course bias drops towards what you need, even as far back as e.g. Super Metroid where enemies would drop a lot more health when you got the low health warning (detrimental for speedruns because it makes it hard to regain ammo when at low HP). I also recall Halo 3's shield bar seemingly covering only like 1/3rd of my total health, after it's drained it takes a lot more damage to actually die.

Payday 2's systems are also super weird but I don't think for dramatic purposes: Shotgun pellet damage doesn't stack, only the highest damage one from a shell counts for each enemy (so wide spreads are actually better) and players are momentarily invulnerable to any attack of equal or lower damage after they have taken damage (or dodged via random chance). Enemy attacks are not actually bullet sprays but a random chance to hit you if there's LOS from the enemy's face to yours (reduced by distance) so partial cover does nothing AFAIK. If you want "dramatic" bonuses you need to pick up matching skills or perks.

 

[Dunan]

I really dislike the use of dishonest percentages. I don't know who "feels" that 90% is like 98% or that 20% is like 5%, but I'd prefer honest numbers. Just recently I was playing Deus Ex: Mankind Divided and, in the hacking mini-game, got through maybe ten consecutive nodes where I had an 15% chance of being detected in each. I expected to be detected in one or two of those.

Similarly, to fail at things that supposedly have a 20% chance of success (like inviting enemy soldiers to your side in FF Tactics) is frustrating.

I've watched and played a lot of sports in real life, where the percentages don't lie. A .200 hitter is going to get a hit one time in five, and you have to be ready for it. A .980 fielder is going to bobble or overthrow one ball in twenty; it would be foolish to expect him to never make an error.

That's one thing i love about the Trails games- the roads have magical lampposts that are designed in-universe to keep monsters away, so traders and travellers that stick to the path are generally safe, and adventurers looking to grind just need to step off the path.

I remember being pleasantly surprised to see this in FF8, and Zelda II also did it many years ago.

 

[Fbh]

Some really interesting stuff.

The bit about a lot of games slowing down enemies when you are not facing them is something you never stop to think about but that somehow makes a lot of sense from a design perspective.

 

[Duchess O'Death]

Super interesting but ruined many games Now I can't stop thinking about what's happening behind the scenes while playing!

 

[Lord of Ostia]

I really dislike the use of dishonest percentages. I don't know who "feels" that 90% is like 98% or that 20% is like 5%, but I'd prefer honest numbers. Just recently I was playing Deus Ex: Mankind Divided and, in the hacking mini-game, got through maybe ten consecutive nodes where I had an 15% chance of being detected in each. I expected to be detected in one or two of those.

Similarly, to fail at things that supposedly have a 20% chance of success (like inviting enemy soldiers to your side in FF Tactics) is frustrating.

I've watched and played a lot of sports in real life, where the percentages don't lie. A .200 hitter is going to get a hit one time in five, and you have to be ready for it. A .980 fielder is going to bobble or overthrow one ball in twenty; it would be foolish to expect him to never make an error.




I remember being pleasantly surprised to see this in FF8, and Zelda II also did it many years ago.

Not sure why it's frustrating to fail at something with only a 20% chance, that's a pretty low chance.

 

[Kinyou]

Lesson one in the creative arts: Your client does not know what they want.

Mitchell & Webb have a hilarious skit about that
https://www.youtube.com/watch?v=_LC0JjvAJt8

 

[Plinko]

Still waiting for the Puzzle Quest developer to admit that, yes, they cranked AI up to a ridiculous level at bosses for chains that dropped in from the top of the screen. I'll never forget reading that they 100% denied that. Come on.

 

[Sande]

I think Bioshock did the "survive by the skin of your teeth" thing terribly. Big Daddies deal an insane amount of damage at higher difficulty levels but you always have this buffer that makes it impossible to die unless you run out of health packs. I'd much rather have a Rosie that deals 35% of your health every shot than one who deals 70% but miraculously you're still alive after the second shot that should have easily killed you. The end result is more or less the same but one option is more consistent and less contrived.

Edit: I remember Rogue Galaxy to balance the insane amount of damage enemy do and prevent one hit kill, you always have a little bit of HP left, to prevent 1 hit k.o. all the time. This was getting on my never because it became nearly impossible to die, because it was easy to use instant healing potion.

This is pretty much the exact same thing I was talking about. Just tone down the damage numbers to a reasonable point and get rid of the artificial invulnerability.

 

[RRockman]

Excuse me...


starts writing things down furiously

 

[Skulldead]

I dislike the idea of intentionally displaying wrong percentages for hit accuracy and such because this doesn't just affect the players' experience with your own game but also all similar ones.

When you play a game where 80% hit chance is actually 95% and then play a game where the chance is as shown, you'll perceive the second game to be unfair and worse. So, kinda like beauty magazines have escalated the idea of what a beautiful person looks like, you can no longer display the actual, real hit chances because you'd deviate from the standard number fudging that people have become accustomed to.

And of course, it negatively affects actually strategizing and number crunching when you're shown inaccurate information. One way to alleviate at least some of the perceived unfairness from actually accurate calculations would be to keep stats so that the player can at least look up and see "oh well, I did actually hit 90% of the time at this range" or sth along the lines. That still won't help as much because not hitting at 80%+ will always feel like bullshit in a players mind at first but at least you're being honest with your players and avoid potentially affecting other games negatively as described above.

Of course, most developers won't care as long as their game feels empowering and 'fair' compared to the others and if very few players know about the number fudging the userbase at large can keep feeling better about their aided odds at theoretically no cost.

This is for this simple reason alone i can't stand to play another Fire Emblem game...


Edit: I remember Rogue Galaxy to balance the insane amount of damage enemy do and prevent one hit kill, you always have a little bit of HP left, to prevent 1 hit k.o. all the time. This was getting on my never because it became nearly impossible to die, because it was easy to use instant healing potion.

 

[Syril]

I don't think this has been mentioned yet, but many old games with high scores would add zeroes to the end of the scores to make it seem like you're getting a higher score than you actually are.

Partly that, but scores where the lowest increment is 100 also facilitated the later arcade trend of adding one point for every continue.

 

[HyperZone<3]

Part of the fun from Left For Dead was knowing the AI was screwing with you. Learning how is pretty cool.

 

[Alcotholic]

Son of a

I had a feeling that was the case.

 

[David___]

Not really hidden anymore since everyone abuses the fact in endgame nowadays, but in Borderlands 2, so long as you had at least 50% + 1 health, you couldnt get one-shot by the mobs.

 

[Shifty]

Mitchell & Webb have a hilarious skit about that
https://www.youtube.com/watch?v=_LC0JjvAJt8

lol, that hits close to home.

Just add a "so... just get creative, okay?" to the end of it and it'd be bang-on.

 

[Angelus Errare]

"first shot of every enemy misses on purpouse"

"last bit of your health is actually more health than it is displayed"

"last bullet of your gun does double damage"

"if being flanked by 2+ enemies only 2 of them will interact with you"

WAT

If the Souls series did ANY of this SHIT I am officially DONE with gaming.

You should probably quit gaming, because chances are ridiculously high your favorite game(s) do some under the hood shit to make the gaming experience better.

 

[SLV]

That first multiplayer round for fps games shenanigans mentioned in the thread makes me think of all the CS :GO first rounds I have had after putting the game on a pause for let's say a half a year or so, where I basically devastate the opposition in the first game just to suck utterly on all concurrent matches afterwards....lol,probably just in my head though.

Spoiler

gotta open dem boxes

 

[CloudWolf]

"first shot of every enemy misses on purpouse"

"last bit of your health is actually more health than it is displayed"

"last bullet of your gun does double damage"

"if being flanked by 2+ enemies only 2 of them will interact with you"

WAT

If the Souls series did ANY of this SHIT I am officially DONE with gaming.

The Souls games definitely do this kind of stuff. The amount of times I got hit and survived with a sliver of my health wouldn't make sense otherwise. Also, I'm 99% sure the enemies de-aggro earlier when you're not looking at them.

 

[Daffy Duck]

That shit surrounding online multiplayer mechanics is astounding. I wonder how much of this crazy "under-the-hood" trickery is present in games like Destiny and Overwatch...

What stuff?

It's really hard reading this stuff.

 

[KojiKnight]

Cool article, terrible layout on polygon's part. We don't need to see the initial post with every response. We're not god damn gold fish, we can remember the original question.

 

[Belker]

I've watched and played a lot of sports in real life, where the percentages don't lie. A .200 hitter is going to get a hit one time in five, and you have to be ready for it. A .980 fielder is going to bobble or overthrow one ball in twenty; it would be foolish to expect him to never make an error.

There's a book called 'The Numbers Game', about football (UK), where a group of statisticians looked at the outcomes of various sports.

In all the sports they looked at - basketball, American football, hockey etc. - the team that was predicted to win would win something like 80%of the time. The clear favourite was the favourite for a reason.

In football, the favourite would win something like 53% of the time. It's a much more unpredictable sport. Giants can be taken down by giant killers.

 

[HueyFreeman]

Hmm - I kind of wish creative assembly hadn't done that. You can tell that the xenomorph does have knowledge of you. Having said that, I guess it acts as its animal instincts as it were. Still my GOTY 2014

Neat article/thread

 

[gypsygib]

What's thumbstick correction?

Mike Bithell&#8207; mentions it but also says it was already explained. I couldn't find the explanation.

 

[snarge]

What's thumbstick correction?

Mike Bithell&#8207; mentions it but also says it was already explained. I couldn't find the explanation.

I'm still not clear on it, but here's more of his quote:

Third person game thumbstick correction is a favourite. Pretty standard in AAA ... game detects collision blocks and steers player around them, ignoring direction of input. Pioneered (I think) by Insomniac but popularised by Ubi.. One of the things I'm proudest of in Volume's controls."

 

[Belker]

I'm still not clear on it, but here's more of his quote:

Sounds something like that if you're moving forward, but are actually accidentally, subtly pushing into an in-game object that will stop you, the game will ignore that input so you keep going forward.

That's my guess.

 

[MaLDo]

It reminds me of how people believe FEAR AI is super good, even if they explained how it worked years ago. Basically the key were the maps themselves, they were pretty linear, usually they were only 2 paths to traverse. The AI sent a pair of of soldiers in each path until they bumped into the player, and naturally it would seem one of the groups was being smart flanking the player.

And that is exactly an example of clever implementation of AI.

 

[Sande]

The Souls games definitely do this kind of stuff. The amount of times I got hit and survived with a sliver of my health wouldn't make sense otherwise.

Sounds like confirmation bias. What about the countless times you were hit and the damage was enough to kill you? If a game was really doing this, I'd expect it to be pretty obvious. I've played all Souls games multiple times and I've never once felt like I was taking less damage than I should have.

What makes me even more skeptical about any mercy invincibility or de-aggro is how completely they would contradict the core design philosophy. Souls games are nothing if not unforgiving. The AI doesn't hang back when you heal, it sees an opening. "We want the player to feel comfortable here and survive even if the screw up" is not a sentence I can imagine uttered in the meeting room at From Software. And this is not to say that From is better than other devs for it. It's just a different focus.

What's thumbstick correction?

Mike Bithell&#8207; mentions it but also says it was already explained. I couldn't find the explanation.

I was wondering that too, couldn't find it anywhere. Christ Twitter is garbage for stuff like this.

 

[Lijik]

I feel like the lack of "coyote time" style tricks is why Crash Bandicoot is so hard. There ain't no pity.

the original crash games have coyote time, but its very slight

But this brings us to the gameplay. We were forging new ground here, causing a lot of growing pains. I started fairly programming the control of the main character early. This is the single most important thing in a CAG, and while intellectually I knew this from Way of the Warrior, it was really Mark who drove the message home. I did all the programming, but Mark helped a lot with the complaining. For example, “he doesn’t stop fast enough,” or “he needs to be able to jump for a frame or two AFTER he’s run off a cliff or it will be frustrating.”

http://all-things-andy-gavin.com/2011/02/05/making-crash-bandicoot-part-4/

 

[Servbot24]

is there a hidden trick to keep gamers from turning any cool reveal of previously-unknown information (most if not all of which were designed to benefit the player experience) into a negative "devs lied" conversation

"Devs lied!" is the most hilarious thing on this entire forum to me

Fibbing is a sin!!! I'll tell my mom!

XD

 

[Candescence]

Game developers are magicians, not politicians. The illusory stuff is for the enjoyment of the player, after all, and reading about this sort of thing is pretty neat.

 

[LurkerPrime]

Man this thread is the best. It's amazing how much smoke and mirrors makes up game design--as someone who strictly plays games, it's difficult to grasp just how much developers go through to make games "fun."


Now to add to this thread's greatness: Here's an explanation of XCOM 2's RNG, by a redditor. As mentioned earlier in the thread--it cheats! (in your favour) It's much more elaborate than one might think, tuned at giving players an exciting battles above all else. Higher difficulties offer increasingly "pure"/unaltered RNG tables. That's XCOM, baby.


The thing I haven't seen mentioned is RNG seeds. They're basically giant tables of pre-determined die rolls that probability-based games use to determine results. Some games have set tables that never change between a session, while others randomise every reload. What this means is, in the former group, a specific set of actions always has the same result, and in the latter, the results always change. Particularly for strategy games, the benefit to random seeds is that they basically allow players to brute force their way through a situation via saving and reloading, until a favourable roll occurs.

Firaxis uses them to great effect in both XCOM and Civilization, allowing players to play either way.

 

[More_Badass]

Man this thread is the best. It's amazing how much smoke and mirrors makes up game design--as someone who strictly plays games, it's difficult to grasp just how much developers go through to make games "fun."

One of the reasons I enjoy reading devlogs. Seeing how it's all made and all the behind-the-scene effort and decisions makes the final seamless cohesive project even more incredible

Game design is as much psychological as it is technical. Using lightning and color and camera angles and enemy placement and other elements to subtly guide players, prime their emotions and expectations, and all that

 

[PMS341]

Lesson one in the creative arts: Your client does not know what they want.

.

 

[UpUpDownDown]

Lesson one in the creative arts: Your client does not know what they want.

This x1000

Great post.

 

[The Illiterati]

I'd like to read the behind the scenes magic of The Evil Within. That game more than any other always had me low on ammo and health but never out. And I was NOT perfect with what they were giving me. So whatever they gave, they gave me the right amount to waste and yet make it through by the skin of my teeth every time. And I still barely had anything at the start of every encounter. The balancing act was incredible to me.

 

[Servbot24]

Lesson one in the creative arts: Your client does not know what they want.

I mean lesson one is to not suck at making art
Lesson 2 is that your client gives you money
So lesson 3 and I agree

 

[gingerbeardman]

Son of a

Excite Truck

 

[Crossing Eden]

Lesson one in the creative arts: Your client does not know what they want.

Can confirm. >>

 

[Erheller]

Dark Souls bosses usually do some percentage of health instead of flat damage. When I was doing my SL2 run, very few bosses would actually one-shot me. Mobs are a different story, though.

 

[Sleight of Jam]

That's actually a really cool arcticle, thanks for sharing.

 

[Carnby]

The Doom confession literally ruined the game for me.

 

[Silvergun-Blue]

This is probably my favorite development secret. I'll have to give that video a watch.

it's crazy that Capcom never went public about it. Sounds like something they could advertise pre-release as having "The perfect difficulty for all gamers"

Yeah, it must have been hard to resist (I wouldn't be surprised if Mikami just didn't tell the marketing team) but as the video details, it's so defective precisely because the player doesn't know.

Its same in real life if you floor it in a car on an open road it dont feel fast like if you do it in a built up area with everything flying past

Yup, which is why the Monaco Grand Prix looks so ridiculously fast from the car camera view despite having the slowest average speed on the F1 calander.

Man... watching this just makes me want to buy RE4 for PS4..even if I already own it twice :x.

Ha, one of my alltime fave's that I replay regularly. It's an absolute masterclass in game design.

 

[Speely]

Lesson one in the creative arts: Your client does not know what they want.

This is so true.

Fake Edit: this thread is fascinating!

 

[FyreWulff]

oh, one of my favorites:

in games that have gun pickups on the ground, the model that sits on the ground is much, much larger to make it more visible to the player to be picked up. The actual gun you are holding is smaller in the in-game world.

 

[webrunner]

i ran some small online games with random chances and stuff before, and here's two things I noticed:

1. two seperate groups of people get completely convinced entirely different things are 'overpowerded' or 'too weak'

2. actual randomness doesn;'t feel random to most people.

 

[Morrigan Stark]

Pretty sure enemies in Souls games will attack less frequently and lose aggro pretty fast when they're off-camera, you wouldn't be able to sprint from bonfire to bonfire so easily otherwise.

This is wrong. They lose aggro by increasing the distance with the player, not by being off-camera. This isn't a "hidden under the hood" thing, it's an obvious thing you see right away.

I'm 90% sure they do this for both you and the AI. Consider how many times you've seen an enemy miraculously survive an attack and scrape by with a completely empty health bar. If you're like me and a friend who noticed this independently, it has happened often enough that you might conclude that there must be some sort of overkill threshold under which a character will always survive a killing blow with 1 hp.

I'm pretty confident that there's an overkill threshold in the game. The AI not doing anything off camera is super obvious tho.

The Souls games definitely do this kind of stuff. The amount of times I got hit and survived with a sliver of my health wouldn't make sense otherwise. Also, I'm 99% sure the enemies de-aggro earlier when you're not looking at them.

It's adorable how wrong y'all are. xD

Sometimes, you will survive with 1 or 2hp left... by pure chance. You had 322hp left, the enemy hit for 320 damage.

That's it. No mystery. It rarely happens, because, well, it's just statistically unlikely to get this specific combination, but there's nothing mysterious about it.

And mobs do attack off camera in Souls games all the time and gank you constantly. I mean the games are kinda known for this. With funny gifs and all.

I don't think I've ever noticed this. I've played trough all Souls games multiple times and what you're describing is very rare. I'm pretty sure enemies have fixed HP while weapons always do the same amount of damage if nothing else changes.

Yup.

Dark Souls bosses usually do some percentage of health instead of flat damage. When I was doing my SL2 run, very few bosses would actually one-shot me. Mobs are a different story, though.

That's probably just a result of the defense gains through levelling up not being a significant stat though. But I wonder, did you upgrade your armour?

 

[Graphics Horse]

I'm still not clear on it, but here's more of his quote:

When Mike Bithell says pioneered by Insomniac, which game would that be? Cos MGS1 does this to an extent.. Obviously pushing into the wall made you lean, but if you're outside of that threshold Snake will run along the wall's direction, ignoring the exact joystick direction.

 

[Amibguous Cad]

I really dislike the use of dishonest percentages. I don't know who "feels" that 90% is like 98% or that 20% is like 5%, but I'd prefer honest numbers. Just recently I was playing Deus Ex: Mankind Divided and, in the hacking mini-game, got through maybe ten consecutive nodes where I had an 15% chance of being detected in each. I expected to be detected in one or two of those.

Similarly, to fail at things that supposedly have a 20% chance of success (like inviting enemy soldiers to your side in FF Tactics) is frustrating.

I've watched and played a lot of sports in real life, where the percentages don't lie. A .200 hitter is going to get a hit one time in five, and you have to be ready for it. A .980 fielder is going to bobble or overthrow one ball in twenty; it would be foolish to expect him to never make an error.




I remember being pleasantly surprised to see this in FF8, and Zelda II also did it many years ago.

Excellent post. I'm always really confused by people who think the RNG is stacked against them - what possible reason could a developer have for biasing results that way?