YouTu made me watch an ad for the new Hunger Games movie which made me remember the Hunger Games so I rewatched The Hunger Games spoilers by the way for The Hunger Games it's okay it's fine it's it's a seven out of 10 more importantly it has some weird game design choices the entire story revolves around a game it's the title The Hunger Games are what we are here to see and the game design of The Hunger Games is fine it does what it needs to do but not very well the Hunger Games are a tool of subjugation disguised as a game and it's strange that they do so little to sell that disguise look the point I am making here is that the design choices made in The Hunger Games are interesting to think about and then there's no other funny business after that my biggest criticism of The Hunger Games is that the game makers consistently interfere with the games to make players encounter each other particularly towards the end a number of these diversion tactics risk injuring or eliminating players this is sloppy game design direct interference with The Hunger Games in progress should be an absolute lack Last Resort something we only use when control of the games is being threatened it should not be the go-to tool of the game makers this interference is necessary though because of the current approach to Resource distribution with the exception of sponsor gifts every valuable resource is placed right next to every player at the start of the game this is done in an attempt to create a dramatic opening and it's an interesting setup in the moment but in terms of game design tools this choice is extremely wasteful the ultimate goal is to get these players to fight not just in the opening minutes but over days or even weeks that's much harder to achieve if grabbing a backpack and hiding away is a viable strategy and that brings us to the strange relationship between the Hunger Games and Wilderness Survival on some level the need for food water and shelter is an effective tool to force players out of hiding and encourage Encounters this is extremely easy to overdo though and it risks undermining the point of the game as a television show I understand the morbid fascination with watching players eliminate each other but a player slowly dying of heat exhaustion does not make captivating TV and senica crane knows this because he regularly takes measures to drive players closer together he wants them to fight that's the whole point of The Hunger Games in conclusion the 74th Hunger Games does the job but there are some interesting consequences to the game design choices these elements might seem trivial to the Casual viewer but they risk undermining the entire point of the game but it begs the question how would you make it better I mean if you got promoted to head game maker for The Hunger Games in panm a lifetime of glory on the line a capital to please and 12 districts to intimidate how would you design The Hunger Games I know what I would do I would change the Cornucopia it's a bold move I know but I promise it's worth it as it is now the Cornucopia sacrifices a ton of helpful game design tools to promote a starting blood bath that will happen regardless of the Arena's design to understand why this blood bath is guaranteed we need to talk about exponential decay exponential decay is the big scary math term for a concept that isn't actually very big or very scary it describes something that starts large then decreases rapidly it's a bit challenging to grasp in the abstract though so let's get a concrete example let's say I am the head game maker and I want the game G to last 7 days to keep things interesting every day I want each player to find and fight another player so that one of those two players will be eliminated in other words each player remaining will encounter one other player every day this seems like a reasonable goal at first glance but a question naturally follows how many players do I need to start with to achieve this rate of encounters for 7 days to figure this out we can start at the end and work backwards the seventh day will need to begin with two two remaining players a Victor and another player for that Victor to triumph over the sixth day then needs to add two more players one for the Victor to fight and another for the runner up to fight the total needs to be doubled again for the fifth day and you might see a dangerous pattern emerging the fourth day doubles our eight players to 16 the third doubles that 16 to 32 and the second day doubles it again to 64 in order to sustain each player having just one encounter per day for 7 days I need to start with 128 players playing this example back in the correct order reveals the exponential decay notice that players are eliminated extremely quickly at the start of the game then eliminations slow down significantly as the days progress this is an unavoidable mathematical truth sure this example doesn't account for players having non-lethal encounters or even teaming up but those scenarios will only create slight deviations to this over overwhelming Trend even altering the rate of encounters only changes how this exponential decay manifests no matter what we do the vast majority of eliminations will happen very early and the longer the game goes the harder it is to invoke a fight between players the Cornucopia is the opposite of a solution to this problem by starting players close together we artificially accelerate the exponential decay of our player count and drive players away from one another if anything the fact that we only have have 24 players to work with means that we need to design away from the initial Slaughter and subtly Drive players closer together over time rather than pulling all of the resources into a big pile and starting every player right next to it I want to spread those resources out as if the entire Arena was a massive Cornucopia low value resources are spread around the outer edges of the arena and the better the item is the closer it will be to the center I'll then start the players along the outer Edge this design implicitly encourages players to travel towards the center of the arena in search of better resources as they do so they'll have to converge into a smaller area and it will become increasingly likely that they encounter one another this new design poses the exact same question do they risk an encounter for a chance at better resources rather than a split-second decision amidst the chaos though venturing further into the arena becomes a potentially hours or days long gamble making the drama of these decisions much more televis we're no longer at the mercy of an unintelligible bloodbath we can track players as they get closer to one another predict fights before they occur and prepare the audience for a show there's also a couple tertiary benefits worth noting about this design the biggest of which is how it curbs the problem of weapons in the arena weapons are a lot harder to design around in the Hunger Games than they are in say a video game getting hit with a dagger in a video game is a minor setback a reason to duck for cover and scarf down a medical cat getting hit with a dagger in The Hunger Games is getting stabbed in real life a player might win the fight but they've sustained a huge disadvantage for the foreseeable future this issue is significantly intensified by the Cornucopia ensuring that the deadliest players are armed often with their preferred weapon only minutes after the start of the games my redesign takes the exact opposite approach it entices those players into dangerous territory through the promise of more effective weapons another important benefit of this new design is the effect on teaming forming a team is Far and Away the dominant strategy at the start of The Hunger Games teaming up will significantly increase a player's odds of surviving the initial Slaughter this new design responds to the overwhelming effectiveness of teams without sacrificing the television opportunities that come with inevitable betrayals placing potential team members a few podium apart means that there's a chance each team member will encounter a hostile player before they can rely on their teammates for backup this design is malleable as well if a certain pairing has become an audience favorite we can place them on adjacent podiums to make their team more likely to occur if we need to prevent a potential team we can place them on opposite ends of the Arena both of these factors also synergize well with the exponential decay most fights will occur in the opening stages of The Hunger Games by withholding the best weapons until players can make it to the center of the Arena we ensure that these early fights have to happen without weapons that can easily inflict a crippling long-term injury we're also throwing a wrench into the teaming strategy at precisely the time in which teaming would be the most useful by the time a team finds each other the games may have already progressed to a point where betrayal has become a tempting course of action we've also happened to pun another beneficial Quirk which is the sudden importance of where a player Begins the games in the traditional Cornucopia players mix fast enough that their starting positions cease to matter almost immediately in this design players will generally encounter one another in the order they've been placed this might seem like a drawback and if done without intention I would agree but there's a way to harness this pattern for greater effect and it's already implemented into the Hunger Games it's strange that effort is placed into scoring How likely player are to win when those training scores are never addressed by the design of the games these scores aren't useless in the current design they help players win over sponsors and intimidate competitors but the ability to categorize players based on their likelihood to win is a powerful tool that isn't being used ideally we want to see players with higher scores last longer a final showdown between a 2 and a six probably won't be as thrilling as one between 11s this is likely to happen happen on its own assuming the scoring system is accurate but we can tip the odds in favor of a better show we already know that players are most likely to encounter other players who spawn adjacent to them so we can first spread our higher scoring players out notice how far they are away Prime to avoid each other until the mid game we then fill in the gaps with lower scoring players in doing so we're encouraging players who are expected to be very effective to rack up a few easier eliminations at the start of the game showcasing their individual deadliness before they get into fights with one another this is also malleable to the specifics of Any Given Hunger Games if we have a problem player who we need to eliminate we could concentrate higher scorers around them to worsen their odds of making it to the end while simultaneously giving the rest of the players better odds if we have two opposing teams of two we could bring the teammates closer to one another while putting each team on opposite sides of the Arena bettering the chances says that the final fight will be a two versus two brawl we wouldn't be able to accommodate every situation and each of these possibilities are still subject to the chaos of The Hunger Games but we could intentionally set the odds to favor specific game outcomes in conclusion the design choices of the 74th Hunger Games do the job but there's some surprisingly feasible changes that could make them capable of producing a much better paste in more entertaining show while subtly giving us a greater degree of control over how the games play out if I were the head game maker this is how I would design the arena with greater intention to the design choices being made I'm optimistic enough to even say that this design could revolutionize The Hunger Games as a tool for Capital control and then head game maker senica crane was so bad at his job that he got assassinated by the capital I I can be so wrong that I die suddenly m my ideas don't seem so airtight I I never even took into account the opinion my audience has of the Cornucopia how would they react to the removal of such an iconic piece of The Hunger Games this could kill the entire media ecosystem that revolves around unpacking what happens in the initial blood bath even worse the ideas aren't as good as I think they are I mean what what about resource density resources would have to cover a much larger area so we'd have to flood the arena with useful items how does that change it on me in an instant this entire time the Hunger Games have been a place where Legendary game makers were forged an opportunity to show my talents to the nation of panm I've been dreaming of The Hunger Game so long that they've lost all context I've been calling these children players their deaths are eliminations to be staged for optimal game performance and while I tweaked my little hypothesis I didn't even notice that the capital had placed me into an arena of my own the stakes have never been higher Sena crane is dead a new head game maker must be selected I look back to my redesign no no no this won't do throwing conjectures into the void and hoping that I'm right will only get me killed I need something that puts me ahead of the other game maker something that lets me Bend The Hunger Games to my will I need data I pull up record of previous games and comb through them Page by Page there are only three recorded games it'll have to do between the three games there are 54 known player eliminations I catalog each elimination by day method and weapon I also mark down each player's number of eliminations the place they achieved in the games and their training score with this data I can search for information that can help me prove my Design's potential I start simple eliminations per day the graph follows the exponential decay as expected what about the method of these eliminations there's a lot of ways for a player to be eliminated so I have to narrow it down to three broad categories the first is player versus player the second is Game Maker intervention this ranges from mations to traps to Arena gimmicks and the third is natural causes succumbing to the elements starving to death or otherwise dying due to the environment categorizing eliminations by Method reveals that in the current design there is an overwhelming trend of players eliminating each other still it's interesting that eight players died to gain maker intervention while only three died to environmental causes while this suggests that game maker intervention is relatively effective at encouraging fights between players I still think that 15% of players being eliminated by the game makers is unacceptable this should be the last line of defense against a game that is out of control not something we expect to eliminate two to three players per game regardless there might be interesting information hidden in those 41 player versus player eliminations broken down by weapon I find I find nothing what knives are a popular choice bees are weirdly more effective than swords there's no Trend to draw conclusions about here what about player scores do players with higher scores get more eliminations or last longer in the games I can divide each player's elimination count by their score and average all of the results to get a rough measure of how many eliminations are expected per Point scored in training it's 0.134 eliminations per point so we'd only expect players to reliably get an elimination if they score an eight or above grouping players by their scores and averaging their placement in the games reveals that the scores aren't actually even that reliable of a predictor of a player's longevity it's not like these conclusions are even reliable either I only have 15 confirmed scores and there's a heavy bias towards high-scoring players it's it's not enough data not even close worse it's not the right kind of data I need to know where these players are when they're eliminated what do they have on them what injuries have they sustained how long since they ate and slept how big was the arena what kind of terrain I need dozens of data points when I only have a few to work with and even then I'd need hundreds of games to establish reliable Trends I mean what am I thinking that I could plot a few graphs and annihilate the competition so this is it I've lost they'll play my silly game and my stupid arena the people will bemoan the Cornucopia absence the media will stumble as they try try to keep up with the changes the audience will lose interest and then I'll follow senica crane to the Grave unless I could make more data create sophisticated fa similes of players within a simulated Hunger Games every thought in their heads recorded every action filed into a spread sheet for future analysis a tool to fine-tune the very nature of The Hunger Games a tool tool to create infinite Hunger Games the simulator will require two broad components the arena and the players testing the players without an arena would be difficult so I start with creating the arena first I need terrain which can be thought of as a two-dimensional grid of numbers this is similar to a topological map where each number in the grid indicates how high the terrain is at that location if all the numbers are the same the terrain will be flat if the numbers are too random though the terrain becomes Jagged I can find a happy medium between these two extremes by using a noise function to create semi- random terrain that follows clear pattern I still need to interpret these Heights though and I'll do that by creating four types of terrain water sand grass and rock these terrain types each occur at specific height ranges and just giving an equal share of the height to each produces decent results but there's a lot of room for improvement to make this better I'm going to do something a little weird rather than creating a set Threshold at which each terrain type occurs I'm going to Define each threshold in relation to the one that precedes it so I can keep the water Threshold at .25 instead of the sand threshold being .5 I'll set it to the water threshold plus a sand size of of 0.25 I can do the same thing to create the grass threshold setting it equal to the sand threshold plus a grass size of 0.25 and then whatever is left over gets interpreted as rock because the colors are now dependent on these sand and grass sizes I can change how the arena looks by changing these sizes the default arena is too Sandy so I can fix that by lowering the sand size the same now goes for rocks so I can increase the grass size to balance things back out if I want to remove a Terrain type alog together I can do that by setting its size to zero which lets me make a more Environ mentally extreme Arenas like islands and mountains now that I have terrain I need to fill the arena with resources for the players to find these resources are simulating items placed in the arena by the game makers before the start of The Hunger Games things like weapons medical supplies food items water bottles rather than tracking every individual resource I can make the simulator far more manageable by keeping track of resource quality quantity and type at each location in the arena this is the same concept as the terrain I'm just measuring the resources at each spot in the grid instead of the height I can recreate the original design of the Cornucopia by placing a high quantity of resources in the center instead I'll test the new design by creating a ring of high quantity lowquality resources around the outer edge of the Arena as we get closer to the center I can turn the quantity down and the quality up and start sprinkling in some weapons to help bait players into dangerous territory finally at the center of the Arena I place a bunch of high quality weapons now that I have an arena I can create the players the players will consist of two logical layers which can be thought of as the body in the brain the body is their ability to perform actions things like moving and eating and fighting creating the body comes down to defining what each of these actions mean in the context of the simulation for example each player has a position in the arena their x and y coordinate if I add to the x coordinate I can simulate them moving to the right the same goes for y to simulate moving up by subtracting I can include left and down movement and I've now mathematically defined what it means for the player to move within the simulation if I repeat this process of mathematically defining each action I can effectively give them a body with which they can act the brain will be harder to create this is their ability to decide which action they want to perform for now I only have one real decision to throw at the Players where if anywhere do they want to move they don't have much to base this decision on yet so I'll just have them seek Water by traveling downhill until they find it once they get to the water they have nothing to do so they do nothing since they've gotten to water it makes sense to Define what it means for them to drink I'll start by giving them a thirst score which can be measured with any range of numbers so long as it's consistent throughout the simulation because the rest of the simulator has been using a range of 0 to 1 I'll measure thirst from 0 to 1 as well to make connecting things easier later on their thirst score will steadily fall as time progresses if it reaches zero they die when players detect that they're in water they can take the drink action instead of moving which just raises their thirst score by a set of Mount they can then repeatedly take the drink action effectively spending more time drinking to fill their thirst score to test this I will spawn 24 players and give them each a random starting thirst [Music] score the players with low thirst scores and unlucky starting positions get eliminated and the rest keep themselves hydrated this rate of dehydration is useful for testing but it's unrealistically fast so we'll turn it down now that I know that the system is working I can do the exact same process to simulate hunger create a hunger score and decrease it over time to force players to eat to check that it's working I can give each player a food item which they can eat to add to their hunger score once they run out of food their hunger drops to zero and they starve so I need to let them hunt for more hunting presents A new challenge though because players aren't guaranteed to be successful how do I determine whether or not a player gets food when they attempt to hunt once again I'll give each player player a new score between 0 and 1 this time it's to measure their aptitude for survival unlike thirst or hunger their survival score stays the same the higher their survival score the more food the simulator should give them when they hunt to do that I can calculate a number to measure how difficult hunting is at the player's location I'll keep it simple for now and just say that hunting has a difficulty of0 2 in grassy terrain and 6 in water now when a player takes the hunt action I can compare their survival score to the difficulty of the terrain that they're in if their survival score is lower than the difficulty they fail and get nothing if their survival score is equal they'll get one food and I'll give them an additional food for every interval of 0.1 their score is above the difficulty I don't want hunting to always be a guaranteed outcome though so I'll add some random variation to their survival score when they attempt to hunt in making it easier to hunt in Grass than it is in water I've accidentally created a decision the players have to make once they are hungry do they stay near the water despite the higher hunting difficulty or do they leave the water in search of easier hunting and in the grasslands they can make this decision by checking their own Survival scores against the difficulty of hunting in the water if they know they probably can't catch a fish they'll look for grasslands but what if their hunting score is exactly the point 6 needed to succeed they'd have a 50/50 shot of getting food if they stayed but would probably get a much bigger payout if they risked leaving the water of course how much food they'll get is only one element they should consider when deciding what to do if they're also thirsty they probably shouldn't leave a source of water in search of a food Sur and the same goes for if they're hungry they might want to risk a lower food payout so they have something to eat immediately so the player is being pulled in multiple directions depending on which Instinct they want to follow I can allow them to decide on a direction by thinking of their thirst hunger and survival scores as voters if thirst is low it'll cast a vote towards drinking the same goes for Hunger voting to eat but if hunger is high it might instead vote to hunt their survival score would work similarly adding a vote to whichever action it thinks is more critical I'm also not limited to one vote per score it makes sense that a player would insist on drinking if they're about to die of dehydration I can simulate this Behavior by giving more votes to scores that are lower this makes players more likely to address their critical issues before attempting a longer term strategy the inner workings of the brain balloons out as I add more actions like looking for resources and fighting and hiding but the important part is that all of these actions are passing through this core set of logic I first Define what it means for the player to perform in action then I tie it to one or more scores that the player is keeping track of and finally I allow the player to internally vote on an action depending on those scores and with the players complete the simulator is ready to [Music] [Music] test here it is my salvation no my legacy every possible game tried and tested a thousand times every emergent action plotted every second of Slaughter categorized for comparison I have created the greatest tool a game maker could wield and with this tool I can forge the perfect weapon with this tool I can design the perfect Hunger Games in conclusion The Hunger Games is fine