PLAYERUNKNOWN'S BATTLEGROUNDS
About the Game
Playerunknown's Battlegrounds (PUBG) is a third-person battle royale shooter which entered Early Access in March 2017. In each match, 100 players parachute onto an island, where the last combatant standing is the winner. Players must explore the island's many buildings to find weapons, ammo, armor, medical supplies, vehicles, and other tools of guerrilla combat. In order to force players to engage their opponents, the play area gradually shrinks, and players outside the designated area rapidly lose health. Less than two months after its release, PUBG has already sold more than 2 million copies, likely aided by being featured on several popular Twitch streams.
Playerunknown's Battlegrounds (PUBG) is a third-person battle royale shooter which entered Early Access in March 2017. In each match, 100 players parachute onto an island, where the last combatant standing is the winner. Players must explore the island's many buildings to find weapons, ammo, armor, medical supplies, vehicles, and other tools of guerrilla combat. In order to force players to engage their opponents, the play area gradually shrinks, and players outside the designated area rapidly lose health. Less than two months after its release, PUBG has already sold more than 2 million copies, likely aided by being featured on several popular Twitch streams.
Busted: 100 Matches with No Bullets
Conventional strategy in PUBG places a high value on acquiring weapons that can kill quickly (assault rifles, sniper rifles, and shotguns) as well as various attachments that improve their performance (scopes, silencers, extended magazines, and grips). Because loot spawn locations are randomized, players who want the best gear must take on the added risk of exploring multiple areas and fighting opponents to take their equipment. As the number of players dwindles, the best gear gradually accumulates among the most skilled combatants, who are then forced to face off in the final play area.
In an attempt to turn this model on its head, I decided to play 100 matches of solo queue on the North America (NA) server without using bullets, limiting my offensive capabilities to melee weapons, grenades, and vehicles. My general approach was to rush for a vehicle as quickly as possible, then drive away from populated areas to find a safe place to loot. After acquiring a reasonable supply of healing items and grenades, I would usually move to a small building in the center of the play area, hiding from the majority of players who would hug the perimeter. The following video shows some highlights from my valiant but wildly unsuccessful efforts.
Conventional strategy in PUBG places a high value on acquiring weapons that can kill quickly (assault rifles, sniper rifles, and shotguns) as well as various attachments that improve their performance (scopes, silencers, extended magazines, and grips). Because loot spawn locations are randomized, players who want the best gear must take on the added risk of exploring multiple areas and fighting opponents to take their equipment. As the number of players dwindles, the best gear gradually accumulates among the most skilled combatants, who are then forced to face off in the final play area.
In an attempt to turn this model on its head, I decided to play 100 matches of solo queue on the North America (NA) server without using bullets, limiting my offensive capabilities to melee weapons, grenades, and vehicles. My general approach was to rush for a vehicle as quickly as possible, then drive away from populated areas to find a safe place to loot. After acquiring a reasonable supply of healing items and grenades, I would usually move to a small building in the center of the play area, hiding from the majority of players who would hug the perimeter. The following video shows some highlights from my valiant but wildly unsuccessful efforts.
In the course of these 100 matches, I reached the final 10 players 20 times, and my best result was #3. I scored a total of 10 kills: 1 by frag grenade and 9 by running enemies over with a vehicle. The game ranks players on 3 different ladders: Rating, Win Rating, and Kill Rating. My peak rank on these ladders was 1159, 502, and roughly 24000 respectively.
Design Analysis
While I was never able to win a match bullet-free, I did gain some useful insight into the balancing of the PUBG system. As the game is currently in Early Access, it is safe assume that bug fixes, performance improvements and new features will continuously enhance the experience for gamers at all skill levels. I have identified four changes that I believe will improve the game's balance, listed in ascending order of how controversial they are likely to be:
While I was never able to win a match bullet-free, I did gain some useful insight into the balancing of the PUBG system. As the game is currently in Early Access, it is safe assume that bug fixes, performance improvements and new features will continuously enhance the experience for gamers at all skill levels. I have identified four changes that I believe will improve the game's balance, listed in ascending order of how controversial they are likely to be:
|
- Rating Systems Do Not Accurately Measure Skill: Checking the leaderboards after my first match (in which I died almost immediately) showed that my Kill rank was at 695,663. Assuming I started at the very bottom, and assuming that the number of players on the NA server has grown since I started playing, it is reasonable to infer that there are approximately 700,000 players on the NA server. Based on this assumption and my peak ranks listed earlier, my performance placed me among the top 0.17% for Rating, top 0.07% for Win Rating, and top 3.4% for Kill Rating.
As shown in the highlight video above, years of singleplayer turn-based strategy games have provided me with no skills whatsoever that transfer to a multiplayer real-time shooter. I am unashamed to admit that I am awful at this game, and yet by memorizing a few car spawns and spending most of my time hiding in shacks doing nothing, my rankings place me at a percentile that would be considered elite in any other discipline. Perhaps most confusing is that my Kill Rating continued to increase even after dozens of consecutive matches without scoring a kill.
While I do not know what algorithms are used to compute the three ratings, it is clear to me that they do not effectively measure player skill. In addition to kills and survival time, the ratings could consider two additional factors. First is the 3-way ratio of shots taken to shots hit to shots received. The ability to fire accurately while consistently avoiding enemy fire should increase a player's ratings. Second is looting. Given two players with comparable match outcomes, the player who has explored more loot spawn locations is likely the player with greater skill, for it is much easier to hide in a shack than it is to clear 10 apartment complexes.
This is not skillful play.
- Equipment Trade-Offs: Currently when a player finds a weapon, backpack, armor, or helmet, there is only one question that player needs to ask: "Is this thing better than the thing that I already have?" If yes, take it. If not, leave it. Similarly, when encountering supplies that live in a player's backpack, there are two questions: "Do I have enough space to carry this thing? If not, should I drop something else to make space for it?"
Game design is all about giving the player interesting decisions to make, and none of the questions above are interesting decisions. With the exception of some weapons with situational advantages, all looting decisions boil down to "get as much of the best stuff as possible." A simple method of introducing more complexity and realism would be for the weight of equipment to affect the player's movement speed. Why should it be the case that a player carrying an assault rifle, a sniper rifle, 200 bullets, a military vest and helmet, 10 first aid kits, and 3 grenades can move as quickly as a player carrying nothing but a frying pan?
Tying movement speed to gear could help to solve three common complaints: the initial scramble for gear is too heavily influenced by RNG, pistols and melee weapons are useless after this phase of the game, and the speed at which the play area shrinks in the late game is too punishing for players who get caught in unfavorable situations. In the early game, a player who can only find a pistol may not be able to win in a standing firefight with a better armed opponent, but their increased movement speed may give them enough of an edge to escape unharmed.
Similarly, players who find themselves attempting to outrun the shrinking play area would now have a real decision to make: do I keep all of my gear to improve my tactical options once I reach the safe zone? Or do I drop something non-essential to increase the likelihood that I will reach the safe zone? These are the types of questions that enable interesting strategic play. - Vehicles and Fuel: The single most reliable predictor of my final rank in each match was whether or not I secured a vehicle within the first minute of play. Driving away from popular drop sites essentially guarantees that a player will outlast those who die while scrambling for gear on foot. It is certainly possible to drop into a crowded city and fight it out, but the randomness of loot spawns means that even the best players can be caught in an unwinnable situation if they can't find a weapon quickly.
There is a simple change that can be made to encourage more diverse strategies in the early game, though it would likely be harshly criticized by the majority of players: vehicles should start with empty fuel tanks. A player would need to find a gas canister before they can use a vehicle. This follows the model of the "care packages" that drop high level loot into open fields: access to powerful options should be rewarded to those players who are willing to take risks.
Increasing the difficulty of acquiring a working vehicle also increases its value, which in turn can lead to more interesting decisions for players. Is it a viable play to try to steal an opponent's vehicle while they step out to loot a building? Conversely, can a working vehicle be left in plain sight as a trap for overeager players? Is it worth picking up a heavy fuel canister in case a vehicle is found later, or even to simply deny the fuel for other players? Again, these questions lead to interesting play, and currently none of them are being asked because it is too easy to acquire a working vehicle.