 All right. Today, we are going to talk to you about where is the finish line designing game endings, right? Every game, every single game has an ending, right? If it doesn't end, it's not a game. It's a trap, right? Games are like a little virtual, you know, world. They're not the real world, right? You can leave them very, very easily, right? But when games don't end officially, right? Saying, ah, this is the end of the game officially, according to the rules. Then there's sort of this psychological trap they put the player in where the player feels like they're still in the game world. They haven't officially been released and leaving a game when you're not supposed to, according to the rules, can sometimes be burdensome or sort of coercive in different ways. I mean, many of you might be in one of these chats right now playing a game with some people at PAX and you might want to bail on that game, but you don't because you're a good person. You feel that pressure. You are, in fact, trapped with those people. Yep. So if you're making a game, right, you have to write a rule that says the game ends when such and such happens, right? And if you write that rule poorly, if you do a bad job of writing that one rule, when does the game end, right? Then that could actually ruin an otherwise good game. There are plenty of good or even great games out there that because that one rule is designed poorly, the game ends up being a disappointment. Even though the rest of the game, it's like I was having so much fun, but the ending rule of the game just ruined the whole thing, right? That can happen. Yep. Now the opposite rarely happens. It is almost impossible for a game to be terrible, but for the end to somehow save it like the up good end to Monopoly is not going to make Monopoly fun. But it will make it significantly better, but it won't. I think just ending sooner would fix that problem. Sure. All right. So what we're going to do today is we're going to talk to you all about the different ending rules for games, right? And mostly if we're designing a game, how do we come up with the best possible ending rule so that we don't ruin our otherwise well designed game with a bad ending rule? And if you're not a game designer, right, you're not making games. This is still very, very helpful information for you because you'll be able to recognize in games you're playing. Ah, this is a great game, but the ending rule is bad or wow, this game could really be improved if we ended it such and such. Right. So yeah, let's let's go ahead. So first, right, we talked about what is the best ending? Well, best, best for who, right? There's all kinds of people involved in any in any game, right? Games have spectators, right? Some ending rule is might be better or worse for spectators. Games have players, right? Players are involved. They might say, oh, this ending rule is better, right? Some ending rules might make a game more or less fair, right? Ah, it would be more fair to end the game now, right? It would be more fair if the blue shell couldn't prevent the end of the game and delay the end and the victory and then cause me to lose. Sure. And some ending rules would make the game, the game will be more fun, right? It's like, ah, if we end the game later, we'll be more fun. If we end the game sooner, it'll be less fun, right? So you have to take all these things into account when you're writing your ending rule, right? Now, you really have to take all these things into account when writing any game rule. But we're in luck in that when it comes to endings, most of the time a good ending rule will be good for everybody, right? Other rules in a game, you know, tend to have to favor one over the other. Ah, this crazy random explosion makes the game a lot more fun and it better for spectators. But it wakes it way less fair and worse for players, the players have a worse experience. But when it comes to ending rules, usually, not always, but usually a good one is good for everybody and a bad one is not as good for everybody. So we're in luck there, but we do have to pay attention to this because it's not always, always true. Okay, another disclaimer here. We are only talking about designing ending rules for ortho games, right? A game where there's winning and losing is the only thing we can talk about today. If we're talking about some like D&D where no one wins and no one loses, there's no right or wrong time to end it. There's no method you can use to figure out when's the best time to end my D&D session. Same with single player. We're not talking about single player. You can just stop playing Skyrim whatever. No one's forcing you, at least I hope no one's forcing you to keep playing Skyrim. Competitive games are the games where we can actually mathematically somewhat figure out a right or objectively-ish right or wrong time to end. Okay, so let's talk about all the different ways in which games end. This is not a 100% exhaustive list, right? But this list of game endings covers the vast majority of all games. It is actually quite difficult but not impossible to come up with a game that doesn't have one of these types of endings. If you think of one, please tell me because those are usually really interesting games people like to learn about. But yeah, it will be slightly challenging to think of a game that doesn't have one of these types. So let's go through them. Okay, the most easiest type of ending ever. Just let the players decide. Hey players, I'm a lazy game designer. End it when you want to end it. Yup, when all players agree. Now some games will bake that in but it's not actually that common in like modern games that people, especially tabletop games, very few games have a rule that says when the players all agree that someone has won and the game is over, just stop. Except weird games that come in weird boxes. A lot of war games have stuff like that in them. Right, so the positives of this are that, you know, game designer doesn't have to do a lot of work. Right, players will end. They won't know what players won't really be dissatisfied with when the game ends. The negative is that it only works well when players are very high skill, for example, in Go, right? Newbies at Go will often keep playing well after the game has already been decided for no reason because they don't have enough skill to recognize that the game is already over. Or imagine the two horrible ways that can go. You're playing Go, you're an expert, you're playing with someone who is still learning and they don't think the game is over and you can't convince them otherwise. Yep, or worse, an unskilled player ends the game early, right, when there was still more to do. So this can be good but used sparingly and wisely, usually in games where you're going to have high skilled play. Right, and lazy designers. Okay, the most basic kind of game ending you've ever seen. Everyone knows this one, first pass the post. There is some event, right, usually a player's score, something like that, and as soon as someone reaches that point, we have concluded, right? So a race, you know Usain Bolt, he gets past the finish line, that's the end of the race. Yeah, they count second and third place, whatever, but it's basically over as soon as someone crosses the finish line. But not just races, also something like this, right? You play Pong until someone has 10. As long as someone has nine or less, you keep playing. It could be one to nine and Pong keeps going. As soon as someone gets 10, immediate game over, right? Very simple. The positives of first pass the post is that you keep playing as long as the game is meaningful, right? It's like, oh, as long as everyone has a chance to win, we're going to keep going. And we're going to end immediately as soon as someone does win, right? Yep, you avoid a lot of the problems of the game ending being in the wrong place or the game ending deciding something different from what you intended because it's objective. Either you won or you didn't. And until the game ends, you can literally, like, you could win. All right, this is really popular game ending lately. Last player standing, right? So we're going to eliminate players over the course of the game. This is going to start with some number of players. It could even be two, right? But that's similar to first pass, the post is only two. But and players will lose. And after everyone has lost except for one player or one team, then that team would win and the game will end, right? So the positives of this is that, yeah, people who can't win stop playing. They leave, right? They could go play another game or something. The negatives people who don't do well enough can't keep playing the game. If they keep losing early, they don't get to experience anything. This can even cause a weird cycle. Not so much about the end in the game, but because the ending for poor players is so early, poor players never get a chance to play the game long enough to be able to get better. It causes like a meta snowball. All right, but yeah, this obviously is the problem with Monopoly. All right, this is of course going to come up a whole. We're going to try to bring it up as little as we can, but you can't ignore it in a panel like this, right? And that is a real version of the game, too. Like, don't that is not a joke. Yeah, no, they publish this. You can buy it if you want. Don't play it. It's just a joke. But yeah, Monopoly isn't a great game by any stretch, right? We always make fun of it. But the ending rule of Monopoly is its worst aspect. If you fix that rule and make the game end sooner and perhaps by a different means, then its fun level will go way, way up. And actually, it might become like instead of awful, it might become like, okay, or mediocre at least. You know, it might, you know, most people when they do play Monopoly, they play until they start to feel the pain and then they stop. If the game officially stopped then people wouldn't be so upset with it, right? And it's because Monopoly almost turns into players decide. The players all agree. We're bored. You've won. Just let us leave. Yep, because the last player standing ending condition for Monopoly is the wrong choice. Okay. Game clock, really basic one. You have a literal clock, right? A timer and you're going to keep playing this game until that time, literal were real world clock runs out of time. You might stop it and started again at different points, right? But it's always going down and never it goes back up only when there's like something you have to redo very rarely, right? Positives of the game clock. You know exactly how long you're going to be playing, right? How many times does someone come over and is like, how long is this game going to be? We don't know. It might be 20 minutes. It might be an hour. Depends, right? This game of hockey is going to be an hour on the game clock, at least. We know, you know, if you don't have commercial breaks, if you're playing hockey locally, right? Three periods of 20 minutes, we know exactly how long we're going to have to play for, right? This is really important to like spectators, really important to advertisers, to the people running the games. Yep, to people who are running the arena, right? They don't have many hours. They have to work, right? They know how many beers they can sell, right? There's so much going on with clocks. They're super, super important. Even in Street Fighter, you could play with a timed clock if you want. That's, you know, super important. Same thing. All right, but in addition to game clocks, literal clocks, there's also resource clocks. A resource clock is very similar to a game clock in that there is some number that is counting downwards. And when it hits zero, the game or runs out or hits a certain number, then the game will end, right? So in Race for the Galaxy, which we're going to use an example a few times in this panel, in Race for the Galaxy, one of the ways the game can end is there is a pile of victory point chips and over the course of the game, players are obviously taking those chips as they score points. If that pile of chips runs out and there are no more chips remaining, the game will end, right? So that's very similar to a real clock, right? It's the thing that is ticking down. It's simply that the rate at which it ticks is somewhat controlled by the players, right? So that's really good because it makes the game very interesting, right? It's like, ooh, not only, right? Am I trying to win? I'm also controlling in some small way when the game ends, right? But it could be bad because now players are controlling when the game ends and that might make the ending happen at the bad time. Yep. This is really common in tabletop games. Like, think about most of the tabletop games you play. You'll see a lot of rules like when the 12th Baphmedad is built or when four towers have collapsed. There's always like specific conditions that players can interact with that cause those games to end. It's gone really out of fashion to have games end after a set number of turns recently. Mm-hmm. All right. Until you can't play anymore. So this is pretty rare because you have to design a game where this is even a good fit, right? But Jenga, you play Jenga until you literally cannot play anymore because the tower fell apart, right? And Blockus, nothing falls apart. But what happens in Blockus is eventually the board is so full that no player has any legal moves remaining. So if you usually this happens a lot in abstract games like Jenga, Blockus, whatever, right? Where there are, you know, a strict set of moves that players can make and there are simply no more or the playing space has filled in Tetris, right? You play until the board is full and you can't put any more pieces, you know, similar shapes to Blockus, right? Tetromino's, whatever. So this is a great way to end a game because it means that players can completely exhaust the play space. It's annoying because you have to design a game to get this kind of ending and that's quite difficult. Yep, you often see it not so much in designed games, but in historic games like stalemate rules in chess. Some abstract games have rules like if the number, like if the players make the same actions too many times in the row the game ends. Those happen because while the pieces haven't run out, the things players can do have run out and anything players do will cause a big circular loop to get them right back where they started and no one can force the game to actually end. So detecting that condition ends up being the thing that ends the game. You also see this sometimes in games like Warhammer or Magic the Gathering where you play and so you just don't have any money anymore, right? And you just literally can't play anymore. Okay, escape patch. So escape patch isn't necessarily a type of ending. It's usually one of the other types of endings, but it's notable in that a game will have a default way of ending and then another different possible way of ending. So it's like, okay, most of the time the game ends when this happens, but in a once in a blue moon, the game ends when this other thing happens, right? So it's different, some games have multiple ending conditions like play until someone has a hundred points or end the game after 10 minutes, right? Whichever happens first. That's two sort of side-by-side endings. Escape patch is when there's a main ending that happens like 90% of the time, like that is the way the game ends, but in rare instances, there's another way to end, right? So this is really a good thing to do in a lot of games because if a player doesn't think that they have a chance to win and the real ending is far, far away, they can stay involved in the game by aiming for that escape patch, you know, once in a blue moon kind of ending situation. Yep, and ideally by aiming for that escape patch, it's well designed to the point that by interacting with that, the losing player will not be messing up the board for all the other players who could still win. We'll talk about that more later on. Yep. Okay. So another thing, multiple finish lines, right? So we talked about, you know, two different finish lines. It's like, oh, you could end with A or B, but what I'm talking about is a finish line followed by another finish line most famously implemented in the Elam ending. You may have heard of this, right? So this is my fellow bald friend, Professor Elam over there on the side. Professor Elam said, man, you know, sometimes basketball just the ending is not exciting, but when the ending of basketball is exciting, that's awesome, right? How can we make the ending of basketball more exciting, more often and better for everybody, right? So he came up with the Elam ending and the way the Elam ending works is that, yeah, you play basketball for a set number of minutes, just like normal, but when the game ends, right? When the clock runs out, you wait until play stops naturally, right? Like a whistle and you say, okay, the time ran out and there was a whistle was blown. Let's check the scoreboard. Okay, team A has 90 points and team B has 85 points. Well, we're going to take the higher scoring teams number and we're going to add eight. So team A is ahead with 90. We're going to add eight. That's 98. We're now going to play basketball with no clock. We are going to have the shot clock so you don't just pass forever. But we're not going to have a game clock. We're going to play and whoever reaches 98 first wins. And this is... Now think about what this does. You get a lot of the benefits of a timed game. You have a pretty good sense of how long the game will go because only the last bit is on time. So you can get a lot of those benefits while also making the end more fair and not running into the one ball thrown at the very end that determines that after the whistle was blown and the clock has run out. Well, you still can get that, but the benefit is there is no way to have overtime with an Elim ending, right? Someone will hit that score of 98, even if the... It's like, oh, the other team has 97. One team has 96. The 96 team scores a two-point shot. Boom, immediate game over, right? The game always ends on a score, right? You never have a situation where like basketball, a lot of times the game will just end while someone's dribbling and the time runs out and that's boring. In Elim ending, the game always ends with the ball going through the hoop and that's great. Okay, so obviously... We talked a lot about endings so far. Yeah, we have talked about endings. And we've also talked a little bit about winning, right? Because there's this sort of unavoidable attachment between the victory conditions of a game and the ending conditions, right? They're not the same thing. The rule that says player A wins if and the rule that says the game ends if, right? Are not necessarily the same rule, but they are often tied together very strongly. So we have to address this, right? Okay, so sometimes the game ends when someone wins, right? Like we talked about, you know, tennis, right? Someone gets to match point and they might lose even if they got to match point. The game keeps going. Then they hit match point. Okay, the game immediately ends over GG, right? Yep, you don't go into the third phase of tennis where they do an extra scoring round and then reveal the hidden points or any nonsense like that. The game's just over. Right, if we're going to play five sets and I win the first three sets, we don't play two more sets for no reason. The game ends when someone wins. The game ends then someone wins, right? Okay, the game ends when this happens. Then we're going to tally up the score and figure out who won afterwards, right? You see this happen a lot in games where the score is like a secret, you know, players are hiding their score behind some sort of shield and then everyone reveals at the end. Okay, game over, how many points does everyone have? Or you might see it in games that are complex where it's not clear who has won until a bunch of calculations are done and the final board state is determined. Poker, right? It's like, okay, no one's raising anymore. This is who's in. Everyone's done folding. There's no more betting happened. Reveal cards, it's over. Reveal cards, find out who won, right? Yep, there's no postcard reveal action you can take in a poker hand. Not that any poker variant I know of. Yeah, right? Okay, so if you have that situation, right, with certain types of endings, especially the timed ending, right? The winning might not happen at the ending. You might need to do even more playing and you know, it's arguable that, yeah, technically the ending rule hasn't been met because no one was winning. You're going to keep playing, but you get the idea, right? Depending on your ending rule and your victory rule not being aligned, you may have to introduce some sort of tiebreaker or overtime or additional stuff to figure out who won after it ended, possibly with more play. These are crucial. If you've ever played a board game where everyone's tied at the end and then someone flips through the book to figure out what the tiebreaker is and there is no tiebreaker, you know how unsatisfying that is. A hockey game that just ends in a tie, at least in the days when that could happen, really unsatisfying, even the shootout's pretty unsatisfying to be honest. Football ties are kind of weird in football, but like playoff hockey, play until someone wins because time is not a problem. Fairness is what they're trying to solve for in the Stanley Cup Finals. So they'll just keep playing hockey until someone wins, even if it takes almost nine innings of hockey. Yeah, but you know, like I said, this can be a good thing or a bad thing, right? You know, you might like one period of overtime hockey with an exciting golden goal finish. You're not going to like nine extra innings, right? In a baseball game. I will. Yeah, in a hockey game, I want nine more periods. I'm good with that. But you know what I recognize? Because I have, you know, the empathy of a human. I suspect the players are not super jazzed about that. I expect the people who work in the arena are super not jazzed about that. All right. Okay, so we talked a bit about player elimination, but let's talk about it a bit more. It tends to be a taboo a lot because of just the social ramifications, especially just, you know, for people playing games with their friends, right? You know, it's like Counter Strike. I remember when that came out, people were super upset. You know, luckily the rounds are short, right? So if you die and you're eliminated, you get back in pretty quickly in the next round. But if you got four people over your house playing board games and someone's eliminated and there's no one else to play a game with, that person who got eliminated too early, what do they have to do? Chill out and do nothing? They're not at a convention like PAX, right? Like we are now. What are they going to do? Player elimination really sucks for them, but it's good in that you don't have players who can't win a game still playing a game. And that is key because if a player is still in the game, but the game is not over and they can't win, they are trapped in garbage time. You've all been there. This is probably the worst part of tabletop gaming, but it affects all games. We use a lot of tabletop examples. I've seen garbage time in hockey before. Yeah, garbage time is when a game is still going. The ending condition has not been met, but the victory condition probably, effectively already has been met. So players are playing for no reason at all. In sports, this could risk injury, right? Why am I still playing this dangerous sport for no reason? But even in non-sports, it's like, yeah, why are we still playing this for no reason? It's already been decided. The ending rule is misaligned with the victory rule and people are just playing a game for whatever. And especially the players who cannot win anymore. They're still, you know, in garbage time, you might have a situation where, okay, two of the players can still win, three of them can't, and that leads to king making, right? Yup, if I am behind, I cannot win. Scott and Joey Jojo are both like duking it out for first place. I'm in last place. I am mathematically eliminated. There's nothing I can do to win the game, but there's no escape hatch for me to go after and there's no way for the game to kick me out and there's no way for me to quit. I get to decide who wins now. You've played in games where third place decides who's in first place. Maybe on that, it's Scott because he beat me at root last week. So I mess Scott up so that he loses and Joey Jojo wins. And now we've made the game less fair, but I had more fun. Right, sometimes king making is intentional. If you've ever played a lot of Bomberman editions, you will set it up so that when a player dies, they will become a ghost on the side of the map who can then haunt and throw a few bombs into the playing field to try to attack the player, probably the one who killed them and maybe even get a chance to come back into the playing field. It's a whole nother discussion. Probably less fair, but a lot more fun. Yeah, king making can sometimes be intentional, but if you have a game that's seriously competitive where people care about winning and losing, then it's usually a negative. You should try to avoid it, which means you should also avoid garbage time. And that's all based on having a good ending rule that is in alignment with the victory rule. All right, we talked about this a little earlier. His race to the galaxy again, right? Player control over the ending in regards to victory, right? A player who is winning a game or believes they are winning a game is going to want to end the game quickly, but give less time to other players to catch up to them. A player who is losing a game wants to extend the game. It's like, well, I'm losing right now. I could win if I had like five more turns to catch up, right? But the game might end before then, right? So if you have an ending condition, such as in race to the galaxy that allows some amount of player control over the ending, that is something to be very, very aware of and you absolutely need to calibrate it very perfectly. Yep, because you could get into a situation. I've been in this situation in let's just say poorly designed games that I might have seen on Kickstarter where a player can cause the situation that ends the game, but that player is not winning. Meanwhile, the player who is like 99% likely to win can't actually do the thing that ends the game right away or you get an even worse situation where a player is losing. There is something they could do that might give them enough points to like have a chance to get ahead. But because the way the game set up, if they do that, that will end the game immediately and then they lose. So the player's choice is do the thing that guarantees I lose or do things that also guarantee I lose. All right, so in a lot of games out there, the eternal struggle. Most famously in CCGs, but also in plenty of other kinds of games, there are usually this dichotomy of agro strategies and control strategies, right? Aggro strategies, aggressive strategies are where players try to score a lot of points very, very quickly in the game, expend a lot of resources very early on and hopefully end the game quickly, right? Control strategies are all about surviving. Assuming that other players are going to be playing agro, you just need to not lose and not let the game end. And as long as you can succeed at that and get to the later stages of the game without the game ending, the agro players will have expended all their resources. They're all out of gas. They just, you know, they shot all their ammo out of the machine gun and now they're empty, right? And now you're in control. You can play big cards that cost a lot of mana. You can cast a big magic spell, right? And the other player can't do anything about it. And all of this is going to be very, very important for your ending rule. If you have an ending rule, especially one that allows player control, but even without, if your ending rule ends the game early, agro strategies will dominate. If your ending rule ends the game late, control strategies will dominate. So if this dichotomy exists in your game, your ending rule has to take that into account and make sure to balance for whichever one as a designer you want to favor. Do you want both to be viable or do you want one to be more viable than the other? What it really comes down to, right, is people love close finishes, right? With any kind of ending you got and any kind of game you got, a close finish is better for everybody. Yup, imagine if the curve of the agro player running out of ammo and the control player build themselves up. If the game ends right at the moment they could cross each other. Mwah, chef kiss. That is just a beautiful game right there. Right, so the close finish is totally dependent on that alignment, right? The victory, right? When are the players who are winning and losing as close as possible to each other, right? And then end the game right then and that will be the most exciting and the most fun. If you end too late, nuh, if you end too early, nuh, right? Do not end a game too soon. This is hard to do. It's really hard to think of games that end too soon, but it is possible. Games that end too soon are games where players didn't have a chance to play with all the toys in the game, right? So if a game, if a game has like, you know, 20 different, you know, they have a starcraft, right? It's like, okay, the game ended from infantry. I didn't get to build any big giant spaceships, right? All that fun stuff is in the game, but I never get to use any of it because the game ends too soon. Yup, big, big point salad, German style, Euro game and you're building stuff and there's all those big stone buildings that do something crazy and cool and no one gets to build them anytime you play. No one ever builds them. The game always ends before then, right? Voltron went in to go for a fight. One lion got showed up and beat all of the bad guys. They never fought Voltron. They never used blazing sword. They just won right away. It's like, that's too soon. Not only does it really happen, but it's not that bad because if the game ends too soon, you'll be unsatisfied, but also you're free. The game's over. Sure. This rarely happens though because I think most game designers can recognize that all of the fun toys that they've built into their game when they're not being used, that's like, you don't need someone to tell you that you need your game to go longer, right? What you really need to worry about is games going long, right? Ending early is better than ending late. If you end a little bit early, right? Okay, people only got to swing the blazing sword of Voltron twice. They still got to swing it once. Good enough. They want to swing it again. They can play again, right? And it's too late. I've been swinging Voltron sword for three hours now. Yeah, it gets boring. You fall asleep, right? People are playing in garbage time. It's a monopoly situation. It's always better to end early than to end late. Okay, so with all these things in mind that we've learned about game endings, different types of endings and how they align with victory conditions, we are going to go through an unproven but still I think decent method for if you are designing a game and I need to come up with a rule, right? I'm writing the rule book for the game. The game ends when this happens. How do you write this rule? A specific procedure for doing so? Let's go through that procedure now. Okay, so to go through this procedure, we're going to need a game to do this with, right? So I came up with this game. I probably did. I'm probably not the first person to come up with this incredibly simple game, but I chose this incredibly simple game because it is very convenient that with this game, the math we're going to do later becomes very, very simple and straightforward. So here's the game. Or even possible in some cases. A lot of games you can't actually math them. True. So here's the game. It's real simple. It's the D6 game. Everyone gets one D6. You can have any number of players you want. I don't care. Everyone rolls D6, right? Whatever number shows up, that's your score. Then you roll it again and you add that number to your previous number and you keep every time everyone rolls the same number of times and everyone adds up all their rolls together. Whoever has the highest total number across all rounds is going to be the winner. If there's a tie for first, no, then no one wins. That's just, that's a loss. You have to be ahead by one to be winning. And look at this lazy game designer up here, not actually designing a tiebreaker for this game. Right. But also, most importantly, we don't have an ending rule for this game. When does it end? We don't know. We're going to design. We're going to figure out when should this game end, right? How many times should we roll the D6, right? How many rounds, right? Let's come up with a rule. So to come up with that rule, we do some math. Here's some math. This is, by rolling D6, the percent chance to win based on the current relative score. So just to understand how to read this, right? Let's say we're in, doesn't matter what round we're in, I'm ahead of rim by three points currently in the game. That means if there was only two rounds left in this game, it was going to end after two more rounds. I'm ahead by three now. My odds of winning after two more rounds are 76.08%. Not bad odds. Yeah. So obviously, because this is such a simple game based on only probability and nothing else, I was able to calculate this entire chart. On the left of this thing, you see it's all ones and zeros because that is zero rounds left. If we got to the end of the game, there's no more probability of someone won or someone lost. There's nothing else to look at whatsoever. And obviously, in most games, you're not going to be able to calculate these numbers so perfectly. That's why we're doing this with the D6 game. In fact, it's like as Rem said, it might be impossible to calculate these numbers at all if not incredibly difficult. In fact, you ever watched like, look at a football game and you see that graph of like percentage chance of winning over time of the two teams? One, those graphs are insanely complicated to make. They use a lot of math and a lot of statistics, but also they involve a lot of subjective analysis and they involve looking at a history of like football games in the past to determine what percentage chance someone had to win at every given point. But it's not objective math. That is not a law of the universe. That is experts eventually just making an educated guess. That's why you can see a game go 90%, 0%, suddenly when something crazy happens. They're using lots and lots of statistical models, big computers and vast troves of data. Here, I'm using a D6 to figure this out. It's basic probability. So looking at these numbers, when should we end this game? What should our ending rule be? We could play a set number of rounds. We could say, yeah, we'll play four rounds, but what if after four rounds, the players are still really close, right? Then sure, we'll get the close ending that we want, right? But is the player who wins really deserving? Are they truly luckier? Are they luckier enough, right? We should sort of maybe wait to end the game until someone breaks away. If someone breaks away, right? We want to end the game immediately, right? If somebody rolls on the first roll, if I roll a six and Rim rolls a one, right? It's like even if you were to play five more rounds after that first round, my odds of winning are still 79.5%, even if we play five more rounds, right? So a decent rule might be, once a player is up by five, then either the game ends or the game has a set number of remaining rounds before it is actually over. Yeah, not even five. If you look at these numbers, right? It's like, there's an argument to be made to end the game immediately if someone's ahead by two or three, right? Yeah. It's like, yeah, if I'm ahead by three, even if we play three more rounds, I still have a 72% chance to be ahead after three more rolls, right? Why waste everyone's time with the three more rolls? As soon as we've determined you're the winner, the likely winner, right? Just call this thing off and we can play again and put the person who's down by three out of their misery. They only had a 20% chance to win if we play three more rounds. See how easy this is? Why don't we actually look at some numbers? Yeah, so you don't have numbers. So what are you going to do? Because your game is not this simple and straightforward, right? You're going to play test with no end, right? So when I say play test with no end, I don't just mean play test a whole lot, which everyone should. You got to do that too. You got to play test a thousand percent more than you think you need to when you're still not going to play test enough. Yeah, that's obvious, right? We're talking about play testing with no ending rule. You didn't design the ending rule yet, right? So don't put one in as a placeholder. Just play test and don't end the game. Let everyone just keep playing and don't stop until, you know, it's obvious that you should stop, right? At some point, your players will start quitting on you because it's been four hours, right? You'll start, you'll feel it in your gut like, okay, there is nothing more to be done here. And that that's your ending rule. I guess that is an ending rule, but the ending rule is everyone collectively decides. Yeah. I mean, not many games can get away with the designer watches you play and says, okay, stop now that person is one. Like there aren't many games that can accommodate that. Yeah. If you do send the game to play testers far away and the play test happened without you, set the ending rule to just be end when you want, right? And then have the players report to you data so that you can then figure out the ending using the method. So here's the method, right? You're going to take one step back, right? You did these play tests. You saw when people thought the game, like they had enough, right? We had enough on turn five every time. We had enough after 10 minutes every time. We had enough as soon as somebody got built that thing, they became so strong that we couldn't beat them. So as soon as someone builds that, we've had enough, right? Whatever that thing is and make your first draft of your ending rule to be one step earlier than whenever that thing is. If people got tired of the game of five minutes, say, all right, my first ending rule for my next set of play tests is going to be end the game at four minutes. Oh, you're almost like frame by framing the game to find the exact moment that someone really won or the exact moment that you as the game designer wanted to be the pivotal moment, the deciding moments. Yep. You moved it earlier. You're going to obviously keep play testing with your now first draft rule, right? But you're also going to start collecting a crap ton of data. You can't calculate the probabilities for your game like we can for the D6 game. So you're going to collect all your data from all your players and you're going to try to make graphs where the x-axis is time, right, or the course of play, whatever that means for your game. And on the y-axis is going to be some metric that you are trying to end the game according to, right? So here we go. We've got chance of winning over time. I generated this graph with random data. I just put a random function into a spreadsheet and that's why it looks the way it does. A pretty swinging game looking there. Though, look, this game actually, it's funny. This is random data, but the graph shows something interesting. It appears that this game is very swingy and then toward the end, if it goes on long enough, the players sort of converge. Yeah. So you can see here on the x-axis, we have turns or time, right? That could be minutes, turns, whatever. And on the y-axis is a percent chance to win. That's why you can see, you know, there's only two players. So that's why you can see them sort of, you know, being very symmetrical, right? If one player's percent chance of winning goes up by 10, obviously the other player's chance goes down by 10. And you can see here the game sort of converges, you know, twice, right? But really the final convergence seems to be around turns 17, 18, where kind of finally stabilizes the blue player is on top. The red player is on bottom. This is only, of course, one playthrough of the game. But if you were to build this graph averaging together several playthroughs with several different players, you might notice those convergences happening around the same time period and you'll get a really good idea of where your ending spot should be, right? You can also get a lot of interesting subjective data from these kinds of graphs. Like that end bit almost implies that the game gets to a point where players have sort of built or done all the stuff they can do and now they've plateaued, neither one of them has a win-more mechanism to like force a victory. So even though blue is ahead, blue is not ahead by as much as blue was at other points earlier in the game. So there's a lot of subjective data you can get out of this beyond just figuring out exactly when the game should end. It can also tell you, have you let people play with your toys or do people run out of toys? Yep. Also, you know, percent chance of winning is even if you can calculate it somehow or estimate it somehow. But you probably can't, right? If you change your ending rule, that will change the chance of winning. If we change the rule for this game, right, this game looks like it ends after 25 turns or so, right? If we change the rule and say this game ends after turn five, that actually changes the graph, right? Because an agro, if there was an agro and a control player, if we end, if we decide the new ending rule is turn five, suddenly the agro player's percent chance to win is going to be way high and the control player is going to be way low. So every time you figure out an ending rule or tweak your ending rule and then you want to get more data, you got to throw out all your old graphs, make brand new graphs with that ending condition because it's a completely different experiment now, right? But you don't have to make new graphs if you make a different one from the Y axis. For example, total score over time. So here's a four player game where players can only score points and never lose points and it shows score over time. And again, this is random data. So if we play tested a game and this is the data we collected and the four players, right? We're using four different strategies, right? What would we conclude? We conclude that the players using the yellow strategy maybe want to fix something in the game relating to that strategy. It's clearly not a winner. Or it might be more that in a four player game there are only three like strategies that are viable. You see this in a lot of games like Settlers has this. Green might be playing Settlers and doing like the brick and wood strategy. And then red's like, oh, I'll do the city strategy instead. The third, the third player is like, oh, I'll do the Stiffy Cards and the sheep strategy. The fourth player, what the hell are they going to do? Those are the only three strategies in Settlers. They're just like make and do. They've already lost from the beginning of the game. Yep. But what can we learn about the game ending from this? Well, we obviously might want to consider if we don't fix the strategy divergence, we might want to consider some player elimination. Why is yellow still in this game? They pretty much never had a chance to win or come back, right? Set them free. We also see that even though red was able to hang, red sort of fell out early. It was between blue and green. Green couldn't, you know, could sort of keep up with blue, but was only able to overtake barely and wasn't able to maintain that overtake. We probably should have ended this game around, you know, turn five maybe, right? In between the four and the sixth there, right? Blue overtakes green. You got your correct same final ranking there. And all those turns from five to 10, nothing really changes with regards to rank, except for that brief moment green is ahead, right? Which might be the blue run out of things to do. Right? Why did we play those final five turns? That was a waste of time, right? The closest ending most satisfying for all the players would be turn five and maybe fix whatever's up with the yellow player's situation as well, which might not be related to endings, right? All right, let's look at another graph. Also random data, right? Here's total score in a game where score can go up and down, where players can lose points. And on the x-axis rather than time or turns, we have the wheat remaining in the game. Assuming it's a board game where wheat is a resource that is limited and over time players are taking and eating the wheat somehow, the wheat supply is decreasing down to zero and the game ends when there's no wheat left. You can see right away, right around turn, right around when there's like seven wheat left. It's almost like duchy panic time. You've got wheat panic time because suddenly you see the yellow player skyrockets up in victory points as the wheat is running down to zero. Something happened there with the wheat. Yep. So, yeah, you can see that that's going on. So the game does seem pretty exciting. Players going up, players going down. Yellow, who ends up obviously winning, wasn't really in the lead at any point during the game, right? It's up to you as a designer to decide which of these strategies the players were using. You believe is the one that deserves to win, right? Maybe you believe that the red player played the best and they should have won. You might want to decide your ending role should be at turn 10, right? Or maybe even, you know, well, not turn 10. 10 wheat remaining or 13 wheat remaining, right? Just take 13 wheat out of the game. The other thing, yeah, you might recognize like you said, remove some wheat because you see the game's pretty stable until when the wheat starts to run out, make the wheat run out sooner to force the conclusion more quickly. Yep. You would get, if you just took five wheat out of the game, so there's 25 wheat in the game and I guess they runs out to zero, just remove five wheat, change your wheat supply to 20 and the game at zero and you'll see, oh, we had the same ranking when there were five wheat left as when there were zero wheat left. So we basically didn't change anything, right? Yep. So... Now there might be confounding variables. You'll have to do this with all your resources because wheat might be a red herring. It might be that the wheat runs out because the players who are losing are desperately trying to buy wheat to catch up to the winning player or it might be because the winning player is consuming the wheat through some thing they got in the game that ends up being a win more that lets them just dunk on all the other players for a while, even though the game's effectively already over. There's no way to know. That's something you can't know from the graph. You're the subject matter expert here. You design the game. You should know what the significance is of the wheat in those last five weeks. And this is why you need to overlay, right? When you've made all these graphs, you need to take transparency sheets or virtual digital transparency sheets. Draw a red line on each graph individually like a vertical line and say... Who at the audience of Pax Online has ever seen a transparency sheet? We're older than a lot of them. Draw a vertical red line on each of your graphs indicating like when that graph indicates the game should end and then put all those graphs on top of each other. Make sure the x-axis are in alignment with each other, which could be hard to do. Like, okay, there were five wheat left on turn seven, so you need to line that up, right? And then see the red lines bundling up together. When you start to have a good ending rule, if the red lines are all over the place, you need to work on the game a bit or at least pick one maybe. But when you get a good ending rule, you'll start to see those red lines sort of come together. It's like, ah, I like the game to be 10 minutes long. I like the game to go until this 20 wheat as a good starting supply. It should be about a 20 turn game. And when you see the lines lining up, ah-ha, you're on to something, right? If they're far apart, need some work, right? Overlay. It also gives you a nice subjective piece of information around where might be the most interesting parts of your game. When things cluster in an area, that's almost like a critical moment in the game. It's worth it to step back and figure out what that critical moment is, because basically you'll eventually determine like there might be four critical moments in your game. And then your final decision is which critical moment should end the game. All right, so we can't stress this enough, but aiming for the photo finish is the key, right? In the real world, in most games, photo finishes are kind of rare, you know, more rare than they should be. And when they do happen in any kind of game at any level of play from the ultra professional down to, you know, two people in the backyard, right? A photo finish is exciting, right? People who lose close are not as upset as people who get crushed. Spectators love to go see their team. They'd probably rather see their team win close than win by a lot. It's more, you know, it's more emotional. I mean, I'll take it if the Rangers win like 10-0. I mean, yeah, but still, you get the idea, right? But if they win at the last second in a really close game, I'm actually happier, but I'm only happier after they won. Photo finishes usually mean that not only is the game more fun and more exciting for everybody, but also that the ending was correct, right? In that, ah, we didn't play any extra, right? We figured out who was winning and exactly, okay, they're ahead by one point and we can tell that they're gonna win and they deserve to win because we're the designer. We decide what deserving to win means. It's up to us. It could be whatever we prefer, right? The person I prefer to win was winning by one point and the game ended exactly then. Perfection, right? Aim for that photo finish, right? Like I said, let the right one win. You say, right? If you are ending at a good time and some other player is winning and you say, I don't like the way that player is playing, right? But they keep winning every time, but I hate how they play, right? You're probably ending at the wrong time, right? The player who's playing the way you like, right? The employing the strategy that you believe deserves to win. Your game is ending too soon or too late for them to be effective. Yep, because if you're a game designer, really the core thing you're deciding here is what does this game reward? What behaviors are rewarded in this game? And if you want the game, like you make some sort of political entry game and the player who just assassinates everyone wins every time. One, you've probably got a degenerate broken strategy. Two, the game's not doing what you wanted it to do. And three, the game is probably as a result unsatisfying because no one plays with all the actual diplomacy tools. But at the same time, you have to decide, maybe you want the game to be that way. It's entirely subjective. You get to pick what your game is or isn't. All right. Eliminate sad players. We talked about player elimination a lot, but people don't like it, but it's actually quite beneficial when appropriate. But you've got to decide when and how. How do you eliminate players? When do you eliminate them? The point is we've emphasized this several times, but it bears repeating. Players trapped in a game that they cannot win is bad for those players, usual almost all the time, and almost all the time bad for the players who are in the game who still can win. The longer someone is trapped in a game, you are just being rude to them as a designer and making their life not fun, right? Do not do that. When they start, if they start trolling or griefing or king making, all those types of like, they're not seeking the utility of the game anymore. They're not trying to win anymore. Can you blame them? You trapped them there. You put them in a play box and said, don't play. So they're going to start team flashing until you kick them out. Right. So if you don't want to eliminate individual players and the whole game, why not? Right? Just as soon as somebody can't win. Yeah. Yeah, some of those games, if one player goes bankrupt, the game is immediately over score. That's it. Yep. So then you get to a point to where some players might actively assist a player who is bad at the game to prevent them from being eliminated to keep the game going so they can win. You'll have your players almost rubber banding the other players. That is difficult to pull off, but I've seen games that do it. All right. Create the escape hatch. So we talked about the escape hatch early on, right? But escape hatch might be hard to design, but it can save you from doing that individual player elimination if you really don't want to end your game in its entirety, right? Okay. I don't really want to end my game so early. They haven't played all the toys yet, right? But this player can't win now. Create for them some sort of escape hatch. If you're playing hearts and you take the queen of spade, that's 13 points. Why keep playing hearts? You're just going to lose this round. You've taken too many points, right? You don't care anymore, but in hearts, you can shoot the moon. There's an alternate victory condition. It's not a different really ending condition, but it is for a hand, not the whole game. But it's, I mean, yeah, but, you know, the victory and the ending are sort of in alignment there, right? Where they happen at the same time in hearts. But the point is you've taken the queen of spades, right? You can't win, but you keep playing because you could once in a blue moon, shoot the moon out there in space. Or potentially tattooing. Potentially this way. If a player can't win the game anymore, they're eventually going to grief and trolling cause problems. They're going to mess your game up. Give them something fun to do and they won't do that anymore. But it's actually more nuanced than that. If you give them something that technically, mathematically could win like the dominance cards in route. You could win with it. You're already in last place. You're like, you're not likely to win. You're basically giving this player something to do and also tricking them into making the game more exciting because if they succeed and start to get close to that weird escape hatch victory, suddenly all the other players are invested and have to deal with that. And suddenly the game gets real exciting in a hurry. Like the birds and the cats are tied and the games like approaching 30 points and then odd and nowhere. The otters are just like, I'm going to get that rabbit dominance and then the whole game blows up. That can be super fun. All right. Allow quitting. A lot of times people feel really bad for quitting a game. Create an official way to quit your game that is within the rules to free people from that psychological trap. If you couldn't design some other way for the game to end perfectly every single time because your game is too complicated, allow a legitimate way for people to forfeit or quit or escape that doesn't make anyone upset because it is official and within the rules. Yeah, especially longer games. Just this is so rare. But the longer a game is, the more likely it is that someone will want to quit. Maybe maybe they have to poop. Like there's a lot of reasons someone might want to quit. But maybe they have to poop a lot of games. Yeah, is that even if they don't actually flip the table as this vibrating gentleman is about to do in the slide. If you quit most games, you metaphorically flip the table. Even if you are polite because by leaving the rules have no way to resolve that situation and it's almost impossible to continue most games when even one player quits. By letting players quit, you avoid that problem. The player can quit without flipping the table metaphorically or literally and while not ruining the fun of the players who still want to keep playing. It's literally a win for everybody. All right, rinse and repeat, right? Take all those things that we just learned, right? And play test without end, right? Keep making those grass. For a earlier play testing over and over. Keep aiming. However much play testing you think you need. Yeah, keep aiming for that close finish, right? Consider player elimination, right? Do it again. Change your ending rule. Do it again. Change your ending rule. Do it again until it's feeling pretty great. And now if the rest of your game is good, the ending rule will not ruin the game, hopefully. If the rest of your game is bad, even your great ending rule isn't going to save it. I'm sorry, okay? So TLDR, right? We want to make a really good ending rule for our game, right? Every game has to have an ending rule, right? Ending earlier is usually better than ending late. Don't let your game go on too long, right? When in doubt, if you cut a little bit off, you're probably better off. Just like in movies. A movie that's edited a little more, even if it's a little shorter, more likely to be better. Like this panel is going to be edited a little bit shorter because you've gone over time a little bit here. Okay, close finish, right? Aim for the close finish to happen like every time. Could you imagine if a horse race actually had a photo finish every single time instead of just sometimes, right? That would people would love horse racing way more than they do, right? And a bad ending will ruin the rest of your game. You might have designed an amazing game. That's super fun. And if the ending rule is messed up in some way, if it goes too long, especially, that could turn a game that would have otherwise gotten a 10-star perfect review into like a six-star because it's like, yeah, the game was super fun in the beginning, but then everyone got bored, right? If you're unlucky enough, your game might end up having its own. Like if you Google for the name of your game, it is famous for its bad ending, right? If you had just ended that game that was super fun at the beginning right before the bad part, everyone got bored. This game was super fun. The whole time I was playing it, right? Don't let the bad ending rule spoil the whole game, right? Thank you for watching this panel. You got time to sit at home and watch packs because you didn't get to go to packs. Yeah, that's how this year is going to roll for us and you. If for some reason you enjoyed what you watched over the course of the last hour-ish, you can photograph this QR code here to access all of our... Or just type that word. Just type that word into Google. It'll find all our stuff. It's 2020. We're past the point of having to give you a URL. Yes, or type that word into an email to packs telling us that this panel was awful. It's up to you. You have the power. Yes, goodbye.