about 4 years ago - Aaron Boulding - Direct link

Getting a Sense of PUBG Pt. 1

When you play a game with guns, they’ve got to feel good

Creating the most authentic game environment possible means engaging with a player’s senses in powerful, distinct ways—the discrete components of the game each appearing, feeling, and sounding just right. In our Getting a Sense of PUBG series, we’re highlighting specific facets of the game in context of the five senses—touch, sight, hearing, smell, and taste. While you can’t—or, at least, shouldn’t—lick your screen when you play PUBG, these interviews should give you a taste of the incredible processes that make PUBG a crisp, true to life experience.


Marek Krasowski was one of the first people to join the PUBG team outside of Korea.

“I was the second or third ‘foreigner’ who joined,” he says with a chuckle. Throwing himself at a little bit of everything—from sound design to animation to physics—when the game was in its infancy, Krasowski has since graduated to Senior Gameplay Programmer on the Action and Gunplay (A&G) team. “I was given full control to rewrite our gunplay system from scratch,” he says.

Krasowski is no stranger to the experience of virtual gunnery. For a period prior to his time at PUBG—a role for which he was scouted thanks to content he posted on YouTube!—he was a competitive FPS player for a small Polish gaming organization. Thus, the bar he set for himself was enormous, but simple: the gunplay in PUBG needed to feel more realistic than any other game. “No other game has really done this on this scale.”

 

Listening to Krasowski is like looking through a fascinating window into the mind of a true designer. He’s a purist, and starts exactly where you’d expect a designer to start: ballistics. “We don’t use that complicated a ballistics system,” he says, simultaneously referring to the massive array of spreadsheets and information he uses. “We look at the weapon, then I go to a ballistics calculator and calculate how it behaves in real life. We use very close data to the real ballistic behavior, tuned up or down a little for balance, but we’d rather look for a weapon that fits the balance and we can maintain realistic ballistics.” Not that complicated!

Pressed for some details on realism, Krasowski digs into the tech a little. “We use gravity and air resistance, but not calculated in real time,” he explains—adding that the game’s performance is improved by simply mapping the ballistic ‘curve’ graph for the bullet’s behavior and applying the outcomes. In practical terms? “Two identical shots will land within a millimeter of each other.” This is extremely important for technical reasons, and why they don’t use things like wind – the additional calculations would complicate cheat detection.

Air drag wasn’t actually a part of the original ballistic physics model in PUBG. Krasowski explains: “In early release, we didn’t have air drag—we only had gravity. This made our huge maps feel really small. If the bullet doesn’t ‘slow down’, a shot at 200 meters might only be twice as hard as a shot at 100 meters, when, in reality, it’s exponentially more difficult.” The post-launch improvement to ballistics was met with what you could describe as a degree of criticism, as all change is, but Krasowski says the eventual feedback was that gameplay became more interesting.

Fundamentally, the goal isn’t necessarily true to life accuracy—it’s fun. The best example, Krasowski says, is how they’ve developed shotguns. Real world shotguns behave very differently than PUBG shotguns, because their pellet spread is much narrower. Multiple generations of games have established shotguns as short range, close quarters weapons, and PUBG leans into that same archetype.

“Shotgun ballistics are a little different,” Krasowski offers. The explanation is a lot, but the short version is that the game treats individual shotgun pellets spreading from the muzzle as ‘enlarging balls’ as they travel, such that some number of pellets ‘snap’ to a target. “Shotguns in early access were useless over distance because the chance of hitting someone was very low,” he muses. “Using this method, we could increase their effective range.”

It’s little tweaks like this that really exemplify what Krasowski is trying to accomplish—finding the correct intersection of realism and enjoyability. “The experience of shooting is all about the player,” he says. While they always start from a place of realism—doing motion capture of real firearms in the Czech Republic, even!—exact realism does not a good game make. “Shooting is more of a visceral experience,” he adds.

The tradeoffs against the realism of gunplay come out in game systems, Krasowski explains. He offers up examples of the inputs on balance considerations—everything from the hit table multipliers for where your bullet strikes a target to latency—and walks through how the team thinks about solving for what’s fair, fun, and still believable. “Our recoil is more game-y than realistic,” he suggests, adding that managing recoil becomes something of a mini-game within the game. “Some games might solve this using bullet spread rather than recoil.” He argues that PUBG’s solution of asking players to think about the pacing of their shots is more interesting.

As if all of that wasn’t enough, the proverbial rubber hits the road when players get their hands on the content. The hard work of balancing all of these weapons against their actual usage in the game keeps Krasowski on his toes. “We usually have a balance meeting once a month,” he says. An entire team pulls data from the live game, as well as comparison data from a couple of previous patches. They’re looking at kill-death ratios (KDR) measured against distance from targets, total kills, what players pick up, and more – a couple dozen metrics altogether, all sorted yet again into buckets by the skill level of the players. “Every time a player picks up a weapon or drops it, that’s logged, so we look at millions of data points,” Krasowski explains. “It’s important, too, not just to look at the data, but to understand the anomalies.”

Data like that led to tons of improvements to the game. In early iterations, PUBG spawned noticeably stronger weapons less frequently, ramping up power gain over the course of a match. Today, Krasowski notes, the team has moved more in the direction of supporting player preference—which types of weapons a player enjoys using most, or feels most effective with—over having specific guns with higher power levels. “A good example is the M16,” he adds. “It used to be the most spawned, most useless weapon, but now it’s a good gateway between an AR (assault rifle) and a DMR (designated marksman rifle).”

Not everything has to be competitive, either. “Some weapons are just for fun,” Krasowski laughs, offering up the crossbow as an example. “It’s not balanced, it’s just there to offer a playstyle.” The stark contrast between the recently named M16 and the generic crossbow or shotguns raises an interesting question—why no names for weapons like the double-barreled shotgun? “For some weapons we use real life names, for some we dodge naming the weapon due to copyright protections.”

In spite of those limitations, Krasowski acknowledges they try as much as possible to utilize iconic guns with interesting heritages—but they’re pushing the boundaries of what’s realistic as is. “We lie a lot about caliber (these bullets shouldn’t work in these guns!). For example, the P90 is a common request,” he says, then noting that nothing approaching the correct caliber of ammunition for the P90 is in the game (re-usability of ammunition being a critical game design consideration).

So where to from here? Krasowski says that more than adding new guns, they’re trying to rethink playstyle ‘niches’ and introduce new ways to use existing weapons. The idea of firing sidearms while driving is a great example, and something that gives the team an opportunity to overhaul pistols. LMGs (light machine guns) dealing extra damage to vehicles is another way that they’ve tried to diversify the available design space they have.

 

A and G team photo
The A&G Team or the happiest battle royale game dev cosplayers you’ve ever seen?

Before he signed off, Krasowski dropped some fun final tidbits on things that they’ve played with. Some of them have made it into the game, like how it sounds to be shot at. “We actually have the full physical implementation of how it sounds to be shot at,” he explains—including how the supersonic behavior of bullet sound behavior is calculated being built into the engine.

That’s it for our first installment of Getting a Sense of PUBG! Join us again next month as we explore game design at PUBG by matching another human sensory perception with an element of battle royale.

 

–Kevin Hovdestad

The post Getting a Sense of PUBG Pt.1 appeared first on PLAYERUNKNOWN'S BATTLEGROUNDS.

about 4 years ago - Aaron Boulding - Direct link

Getting a Sense of PUBG Pt. 1

When you play a game with guns, they’ve got to feel good

Creating the most authentic game environment possible means engaging with a player’s senses in powerful, distinct ways—the discrete components of the game each appearing, feeling, and sounding just right. In our Getting a Sense of PUBG series, we’re highlighting specific facets of the game in context of the five senses—touch, sight, hearing, smell, and taste. While you can’t—or, at least, shouldn’t—lick your screen when you play PUBG, these interviews should give you a taste of the incredible processes that make PUBG a crisp, true to life experience.


Marek Krasowski was one of the first people to join the PUBG team outside of Korea.

“I was the second or third ‘foreigner’ who joined,” he says with a chuckle. Throwing himself at a little bit of everything—from sound design to animation to physics—when the game was in its infancy, Krasowski has since graduated to Senior Gameplay Programmer on the Action and Gunplay (A&G) team. “I was given full control to rewrite our gunplay system from scratch,” he says.

Krasowski is no stranger to the experience of virtual gunnery. For a period prior to his time at PUBG—a role for which he was scouted thanks to content he posted on YouTube!—he was a competitive FPS player for a small Polish gaming organization. Thus, the bar he set for himself was enormous, but simple: the gunplay in PUBG needed to feel more realistic than any other game. “No other game has really done this on this scale.”

 

Listening to Krasowski is like looking through a fascinating window into the mind of a true designer. He’s a purist, and starts exactly where you’d expect a designer to start: ballistics. “We don’t use that complicated a ballistics system,” he says, simultaneously referring to the massive array of spreadsheets and information he uses. “We look at the weapon, then I go to a ballistics calculator and calculate how it behaves in real life. We use very close data to the real ballistic behavior, tuned up or down a little for balance, but we’d rather look for a weapon that fits the balance and we can maintain realistic ballistics.” Not that complicated!

Pressed for some details on realism, Krasowski digs into the tech a little. “We use gravity and air resistance, but not calculated in real time,” he explains—adding that the game’s performance is improved by simply mapping the ballistic ‘curve’ graph for the bullet’s behavior and applying the outcomes. In practical terms? “Two identical shots will land within a millimeter of each other.” This is extremely important for technical reasons, and why they don’t use things like wind – the additional calculations would complicate cheat detection.

Air drag wasn’t actually a part of the original ballistic physics model in PUBG. Krasowski explains: “In early release, we didn’t have air drag—we only had gravity. This made our huge maps feel really small. If the bullet doesn’t ‘slow down’, a shot at 200 meters might only be twice as hard as a shot at 100 meters, when, in reality, it’s exponentially more difficult.” The post-launch improvement to ballistics was met with what you could describe as a degree of criticism, as all change is, but Krasowski says the eventual feedback was that gameplay became more interesting.

Fundamentally, the goal isn’t necessarily true to life accuracy—it’s fun. The best example, Krasowski says, is how they’ve developed shotguns. Real world shotguns behave very differently than PUBG shotguns, because their pellet spread is much narrower. Multiple generations of games have established shotguns as short range, close quarters weapons, and PUBG leans into that same archetype.

“Shotgun ballistics are a little different,” Krasowski offers. The explanation is a lot, but the short version is that the game treats individual shotgun pellets spreading from the muzzle as ‘enlarging balls’ as they travel, such that some number of pellets ‘snap’ to a target. “Shotguns in early access were useless over distance because the chance of hitting someone was very low,” he muses. “Using this method, we could increase their effective range.”

It’s little tweaks like this that really exemplify what Krasowski is trying to accomplish—finding the correct intersection of realism and enjoyability. “The experience of shooting is all about the player,” he says. While they always start from a place of realism—doing motion capture of real firearms in the Czech Republic, even!—exact realism does not a good game make. “Shooting is more of a visceral experience,” he adds.

The tradeoffs against the realism of gunplay come out in game systems, Krasowski explains. He offers up examples of the inputs on balance considerations—everything from the hit table multipliers for where your bullet strikes a target to latency—and walks through how the team thinks about solving for what’s fair, fun, and still believable. “Our recoil is more game-y than realistic,” he suggests, adding that managing recoil becomes something of a mini-game within the game. “Some games might solve this using bullet spread rather than recoil.” He argues that PUBG’s solution of asking players to think about the pacing of their shots is more interesting.

As if all of that wasn’t enough, the proverbial rubber hits the road when players get their hands on the content. The hard work of balancing all of these weapons against their actual usage in the game keeps Krasowski on his toes. “We usually have a balance meeting once a month,” he says. An entire team pulls data from the live game, as well as comparison data from a couple of previous patches. They’re looking at kill-death ratios (KDR) measured against distance from targets, total kills, what players pick up, and more – a couple dozen metrics altogether, all sorted yet again into buckets by the skill level of the players. “Every time a player picks up a weapon or drops it, that’s logged, so we look at millions of data points,” Krasowski explains. “It’s important, too, not just to look at the data, but to understand the anomalies.”

Data like that led to tons of improvements to the game. In early iterations, PUBG spawned noticeably stronger weapons less frequently, ramping up power gain over the course of a match. Today, Krasowski notes, the team has moved more in the direction of supporting player preference—which types of weapons a player enjoys using most, or feels most effective with—over having specific guns with higher power levels. “A good example is the M16,” he adds. “It used to be the most spawned, most useless weapon, but now it’s a good gateway between an AR (assault rifle) and a DMR (designated marksman rifle).”

Not everything has to be competitive, either. “Some weapons are just for fun,” Krasowski laughs, offering up the crossbow as an example. “It’s not balanced, it’s just there to offer a playstyle.” The stark contrast between the recently named M16 and the generic crossbow or shotguns raises an interesting question—why no names for weapons like the double-barreled shotgun? “For some weapons we use real life names, for some we dodge naming the weapon due to copyright protections.”

In spite of those limitations, Krasowski acknowledges they try as much as possible to utilize iconic guns with interesting heritages—but they’re pushing the boundaries of what’s realistic as is. “We lie a lot about caliber (these bullets shouldn’t work in these guns!). For example, the P90 is a common request,” he says, then noting that nothing approaching the correct caliber of ammunition for the P90 is in the game (re-usability of ammunition being a critical game design consideration).

So where to from here? Krasowski says that more than adding new guns, they’re trying to rethink playstyle ‘niches’ and introduce new ways to use existing weapons. The idea of firing sidearms while driving is a great example, and something that gives the team an opportunity to overhaul pistols. LMGs (light machine guns) dealing extra damage to vehicles is another way that they’ve tried to diversify the available design space they have.

 

A and G team photo
The A&G Team or the happiest battle royale game dev cosplayers you’ve ever seen?

Before he signed off, Krasowski dropped some fun final tidbits on things that they’ve played with. Some of them have made it into the game, like how it sounds to be shot at. “We actually have the full physical implementation of how it sounds to be shot at,” he explains—including how the supersonic behavior of bullet sound behavior is calculated being built into the engine.

That’s it for our first installment of Getting a Sense of PUBG! Join us again next month as we explore game design at PUBG by matching another human sensory perception with an element of battle royale.

 

–Kevin Hovdestad

The post Getting a Sense of PUBG Pt.1 appeared first on PLAYERUNKNOWN'S BATTLEGROUNDS.