by Richard Garfield

There is no doubt what made me a gamer — it was Dungeons & Dragons. When I discovered that game, it changed my world. It showed me something of what is possible with games, and it made me ask myself what other games are out there that I don't know about, what games are waiting to be discovered?

D&D was unlike any game I had ever played as there was no winner or loser. It wasn't just long, it was indefinitely long; it might never end! Rather than being bound by its components, players could bring in anything they dreamed of to the game. The rules were often indecipherable and were almost always open to interpretation. It was completely out of the box.

I had found a wonderful and exciting world, and I still find games and game culture from history fascinating, with modern games exploring so many things. However, I felt an element of critique was missing; there were few places I could go to see what worked and what didn't work in the many games I was experiencing. This is one reason why I co-authored the book Characteristics of Games, which debuted in 2020; it felt like a common language for talking about games didn't exist.

We are in a golden age of games, and this has brought some excellent critique and standards to the medium. Hewing to standards has raised the average quality of games...but following standards too closely also limits us and will prevent us from seeing what is off the path.

What I aim to do in this article is talk about characteristics of games that are often seen as bad, as at best a necessary evil. Plenty of games use these characteristics, but these days they are often penalized for having them, which encourages designers and publishers to try to remove them or work on different games. What exciting and beautiful games are waiting for us out of the box?

Player Elimination

It is understandable that many players don't like elimination; you want to play a game, not watch everyone else play a game. However, elimination is not bad for all players all the time. For example, if you are playing in a place with a lot of available games (like a big game night, a convention, or a game store), or you have other things to do (maybe you are a student and can always fit in some study time), or you are simply happy watching and kibitzing, then elimination may not be a big issue. In fact, you may find it preferable to be engaged in a game for a long time in which you are not really relevant. Being logically eliminated from winning (while still being in the game) is a less recognized problem and often a consequence of a designer trying too hard to keep all players in the game.

A variant that is sometimes suggested is ending the game after the first player is eliminated — problem solved! This approach can lead to a good game, but it is usually a pretty different game. Using King of Tokyo as an example, if you ended the game after the first player was eliminated and made the player with the most VP the winner, then the only player motivated to knock someone out is the VP leader. There are other ways to determine the winner, and some are going to be good games in their own right, but none are going to give the same experience as King of Tokyo.

Typically, elimination is forgiven when the end of the game is near, but instead of leaning on this apologetic excuse — elimination isn't bad sometimes — I would argue that elimination can generate wonderful experiences, even when a game may go on a long time after you're out. Two games I got a lot of value out of are Titan and Diplomacy. Both feature player elimination, and the eliminated player might be out for hours or even days.

Perhaps that seems okay for certain hard-core hobbyists, but I think the appeal can be broader. My family has often played games like versions of Liar's Dice in which each player has a certain number of lives and is eliminated when those lives are gone. Liar's Dice is something that can be played around a table after dinner, with participants being eliminated, then talking, having coffee, kibitzing, and eventually watching the final showdown. The experience of getting the spotlight like that, and the small but real stakes of being eliminated has a real charm. The "penalty" of being out is not always bad; a game doesn't have to consume the attention of all the players the entire time, except perhaps in circles where that is the only thing holding the players together. In Bridge, one player sits out every time: the dummy. This was viewed in my circles as a mixed blessing — sitting out might be a drag, especially if it happens several times in a row, but the position often came with a sense of relief as you let your partner take over for a while.

Similarly, Werewolf/Mafia carry their player elimination well. The games are excellent for spectators and have a great feel to them, not despite the player elimination, but sometimes in part thanks to it.

In short, I see the removal of elimination as an adjustment for a different style of play and a different playgroup as opposed to correcting a problem.

Playing for a Loser

Games sometimes yield a loser rather than a winner. This can lead to a significant difference in how the game is played — and how it feels. This is something that often strikes modern players as "just wrong", but it's historically not uncommon. Some games were played to a loser, then the loser had to pay a forfeit, some sort of penalty.

My first real experience with a game that was played to a loser was Skitgubbe. This is a fast game played with a standard deck of cards that plays well with 3-5 players — and the loser would have to make a goat noise.

Occasionally some players grumbled that they wanted to know who won, but the game didn't work that way, and tinkering with it to make it generate a winner rather than a loser was not productive. The goal of the game is to get rid of all your cards, so one might say the person who got rid of their cards first won — but this is unsatisfying because the most interesting part of the game isn't the play around who goes out first, but the struggle at the end. Certainly the largest opportunity to show skill was not going out first, which was often just a matter of luck, but successfully navigating a more complex situation in the middle or end of the game. I did make successful scoring systems that rewarded going out later rather than earlier, and then you could play to find a winner over many hands — but more frequently we found that if we wanted to have a larger Skitgubbe experience, we would play for the megagoat: the person who lost three times first.

Later I found this thinking useful in my design of What Were You Thinking?, which is now Hive Mind. This is a polling game in which players are tasked with writing down things like three planets or five types of cat, then rewarded for matching other players. The original way I scored it was giving points to players so that high score wins...natural, right?

Well, problems became apparent pretty quickly. One problem is that it's hard to catch up once behind — and frustrating. After all, you scoring a lot of points means you matched a lot of players, so you aren't necessarily making ground on many of them. There was a feeling of hopelessness as you fell behind, so instead I tried an approach in which players got a strike for being the lowest-scoring player, with three strikes removing them from the game. This changed the way the game was played and very much for the better. Players always felt like they could "not lose". A more subtle effect was that doing well didn't demand squeezing every point out of the system, which meant that a person playing well could still afford to give fun and speculative answers, which made the overall game experience more enjoyable. For a while we played with elimination and going down to a few winners, but then I fully embraced playing for loser and simply ended the game on the first elimination.

You can see playing for a loser rather than a winner embedded in games sometimes, and learn how it affects the overall play environment. An example of this is Formel Fun, a game which has a unfairly low rating on BGG. This is a racing game, and each round the car in last place is eliminated. Players learn to adapt their play to "just don't be in last", and you find the spotlight focused on those struggles rather than those breezing ahead early in the game. The tension is inconsistent as you aren't always on the spot, but sometimes have really important decisions. The overall game is played to a winner — but the journey uses this characteristic of games.

Since I have played far fewer games to a loser than to a winner, I am not absolutely sure what it brings to the table, but I have found that it puts attention and stakes on those that are behind — which might be appropriate depending on the mechanisms. Again, in Skitgubbe this is very much where the most interesting play is. Also, while these games can be skill testing, it makes them more casual in that you need to push yourself only when you are under threat of losing, while at other times you are free to engage in more speculative play.

Long Games

The length of a game is one of its most important characteristics. Many playgroups have a restricted time to play and aren't interested in playing a game in several sittings. What games are we losing games by restricting ourselves too much to session length?

My youth was filled with games in progress, games that didn't fit into my schedule, so my friends and I had to adapt to those games. What's more, there was an entire genre of games, the play-by-mail games, that could go on for years. Many games that meant a lot to me would not reliably fit into a sitting; two that I have already mentioned are Titan and Diplomacy, but there were many more, such as Freedom in the Galaxy, the vastness of which is hard to explain. Wargames in particular could take a session just to set up!

That there is a hunger for a bigger game experience is clear; the success of the legacy genre and of games like Gloomhaven show that. These games impose limitations through controlling the typical session length, while allowing the sessions to be strung together in a bigger, more epic game. This is a solution to the problem, a solution that may not allow us to explore the entire space of possibilities offered by longer games, while still giving us a lot.

One place you can see the effect of this standard is in a game that feels like it doesn't last long enough to deliver the variety and scope that seems to be promised. I have run into this issue many times in design; one example is Founders of Reyvick, which is still being worked on. I want to make the game short enough for a reasonable session length, but not so short that the game doesn't give the right payoff. I think I have navigated that in Founders of Reyvick, but I have worked on (and played) games where it simply can't be done.

When I am working independently of a publisher, I usually don't aim for a game length. What I try to do is see how long the game wants to be. A 45-minute game can feel like it overstays its welcome and is better suited to be a half-hour game, whereas a two-hour game can seem like it really needs to go longer, maybe much longer.

Unreliable Game Length

Having the game not only be a manageable length, but also consistent in length is valued. This is useful for game groups scheduling games as they know what they are sitting down to...but are there interesting games we don't see because we value consistency in length?

My game King of Tokyo: Duel can end in a few minutes or take half an hour — and some players regard that as a problem. For me, though, it was part of the appeal. You don't play for a fixed time, then tot up the points; instead you might find yourself losing almost immediately. Once this happens, the dangerous balance of the game is understood — and a player needs to take that into consideration.

Again, Titan becomes a good example. I wouldn't count on the game being finished in less than six hours if people know what they are doing — and it could go much longer — but I once played three games in an hour with my friend Mons. That was not a letdown; it was exciting and memorable and colored all games of Titan going forward. This was a dangerous design, with players walking in a game jungle, not in a cultivated game garden. A game that can go long but which can also end at almost any point has a character that can't be duplicated with consistency.

Downtime

It is standard to keep turn length down, to make the wait between your turns as small as possible. This is great for many games as it keeps all players engaged for the duration. After all, it can be agonizing waiting for your opportunity to act in a game — and I must confess that when I got to this section, it took me a while to get started: Do I really want to go to bat for increased turn length? But for the right players and the right circumstances, it is not a problem and can even be viewed as a positive.

There are many reasons it may not be a problem to wait. Perhaps you want to think about your plans in the game. Perhaps you enjoy spectating. Maybe you just want a break. I have played many games in which participants would sometimes even leave and come back when they were curious how it was going — or when called for their turn. This experience is not for everyone, but it does have a large design space, and for some it is a rich and exciting area.

And you simply can't get certain game experiences with short snappy turns.

Leaning on Titan yet again as one of my favorite games that breaks almost all standards, players move on the master map and when they battle, they play a little skirmish outside of it. The skirmishes are fun — they are varied and skill testing — but each one might take twenty minutes, and there might be several in a row that don't involve you. I have made several versions of Titan for my own use, and getting the turn length down while keeping the skirmishes interesting is a challenge. There are bound to be many good games in that space, but they won't be strictly better than Titan simply because player downtime is reduced because the reduction comes at a cost.

Titan is not alone in being a game a lot of people love that might make you wait a while. Wargames and 4X games are often like this, as is the entire category of play-by-mail games. The 18XX games and Mage Knight are other examples. Many games will have significant downtime if you don't put a timer on them, such as Chess or Go. Of course in most of these games, the downtime is yours to use as you like, and there is plenty to be gained within the game by studying the current situation during other player's turns.

Short turns are at their best when you are reacting to other players' moves. This can give constant interaction on a rapidly changing landscape, which many enjoy, but many games that have short turns are less interactive, and then one can contrast the idea of playing a game with a longer wait and more substantial turn against a shorter turn. If the players are fascinated with the game, it might not matter much, except with the longer turns one can chat without slowing the game down, or get up and, say, use the bathroom without stopping the game. If you are worried your players might get bored and wander off if you don't force them to be at the table by pushing a turn on them every couple of minutes...it might be the wrong game for those players in either case.

Roll-and-Move Games

It is amazing to me that there is any need to defend roll-and-move, but many times I have heard a game disparaged as being roll-and-move, or a player thanking modern games for saving us from roll-and-moves. I can only imagine that they had such bad experiences growing up with roll-and-move games that they have condemned the mechanism itself — similar to how someone might hate vegetables because they were traumatized by elementary school spinach. Roll-and-move might be the oldest game mechanism, so it has a huge cultural importance to us as humans — but aside from the historical importance, any general criticism against the mechanism is rendered irrelevant by Backgammon.

Backgammon is one of the truly great games; whether one likes it or not, it is objectively a masterpiece. If someone dismisses it as being too luck based, they are speaking from ignorance. Any players not sure of that should try their hand against the top computer programs and see what sort of win rate they can get. The game has dropped in popularity as a gambling game over the last decade or two because players have begun to understand just how low their chances were against someone who knew what they were doing.

Roll-and-moves through history have had a hugely varied character, from completely random, like Chutes and Ladders, to games with a lot of room for skill, games that are filled with different boards with different types of spaces and varied pieces that have different capabilities and objectives. A couple of modern examples I love are Takashi Ishida's Magical Athlete and That's Life by Michael Kiesling and Wolfgang Kramer.

Many of the roll-and-moves do have a lot of luck, which at least partially accounts for their reputation, but I don't believe we should regard a 100% luck-driven game as bad, even though it may be bad for us. To evaluate such games, one must ask how well it delivers for the people who like them. Players of Bingo and LCR are obviously not playing for deep decision making, but something more like what people watching sporting events get when they cheer on a particular team. Is the randomness exciting, and does it allow for getting ahead while still permitting comebacks? Does it tell the players a story?

Even the most basic of roll-and-move structures — ones in which you have a single piece and you move it — has a wide scope of possibilities. You can introduce any level of decision making to the game by giving players a choice based on where they land. I have played games in which a lot of the options were delivered by a card or by rolling on a chart. Why would delivering those options by roll-and-move be worse?

In fact, there is a mechanical elegance to that because the possibilities change dynamically from turn to turn and can vary between players. Also, these possibilities are always visible and represented in ways that can be intuitive. There is the possibility for flavor and color that isn't possible with a deck or charts. You can introduce a cyclical rhythm to the game by looping the path or introduce options by forking it; you can make the path linear and have the stakes rise throughout with a natural endpoint. It is hard to write about all of this without wanting to go off and design one.

Unfair Games

In January 2025, I wrote an article on balancing games, "The Balancing Act", that describes some of my thoughts on game balance. In it, I lay out the case that there is not generally one correct balance because balance will change with one's experience, skill, and even play style. Aiming for the perfect balance will usually narrow the range of mechanisms you can use, or narrow the audience since different playgroups will have different ideal balances.

But the real issue for our purpose here is whether there are games that we will not get if we hold fairness as a standard that all games should have? The answer is absolutely. One of my favorite games, the one that influenced my design of Magic the most, is Cosmic Encounter. Each player gets an alien power that gives them special rules in an otherwise straightforward game, and the interaction between the aliens makes every game fresh, and the emergent behavior between the cards and aliens in the game is always surprising.

Cosmic Encounter was also a deeply unfair game as the aliens weren't remotely balanced. At the start of each game, we could often tell who was favored and who would be fighting an uphill battle. This increased the variety of experiences as sometimes players were the underdog, sometimes they were large and in charge. This did not diminish the skillful play available to players; it just meant the skillful players sometimes had to work much harder than others. If this game had been balanced and "fair", there would have been mechanisms the designers couldn't really use, and even if — somehow — they managed to get it fair-ish, players wouldn't believe it because they weren't skillful enough to recognize their failings, or they would get skillful enough that whatever minor unfairness was there would be amplified. Because the designers didn't really care about that, we have a sprawling and exciting game filled with variety and situations you simply can't get if you hold fairness above all else.

The Great Dalmuti was a design of mine based on shedding card games, and it was all about unfairness. The players at the top had advantages going into the next round, and the players at the bottom...well, generally they could climb a step or two and that might be a lot of work. There are many more fair shedding games, but they don't offer what The Great Dalmuti offers, and The Great Dalmuti exists because I didn't really care about fairness.

I'm not arguing that unfair games for casual players are fine since those players don't care about balance. There is no reason the most serious players cannot enjoy an unfair game. The key is that they recognize there are different values to a victory. It means more if I beat a better player at chess than if I beat a beginner. It means more if I use the faction that is weaker. It means more in Dalmuti if I claw my way from the bottom to the top than if I begin at the top and stay there. It does not mean I am a weaker or casual player if I choose to play a stronger player at chess, or choose to play that weaker faction, or choose to play a game in which I might randomly be thrown to the bottom and have to claw my way back up in a skill-testing (but unfair) challenge.

You can always make games fair if you like them enough and really want to measure who is best; the history of games is filled with this. That is why Chess is so often played with an even number of games alternating black and white, why komi is added to Go, why Duplicate Bridge exists, and why there are so many formats for Magic. Even The Great Dalmuti, which is relentlessly and deliberately unfair, can be made fair. I played a league in which we tracked all player records — and even though you may not be able to do well in a particular session, when you begin stringing all those games together, you can easily tell who the best players are. If a game is loved, players can balance it to their taste. If finding the best player is important, it can be done no matter how unfair the game is.

High Variance Games

One of the most common criticisms I see about games is when they are thought to be "too luck based". Games with higher luck are often seen as throw-away experiences; if a game wants to be taken seriously by players, it has to be careful how randomness in the game is expressed.

I suspect that some of this dislike of luck stems from a modern tendency to play more games less often — which gives less time for the true character of a more random game to be revealed. Sometimes this leads to an unintended side effect: when you reduce the number of random events, sometimes you actually make a game have more luck. It took me a while to understand that combat in Titan, with its hundreds of dice rolls, was less a source of luck in the long run than movement, which was one die roll a turn.

Luck is a wonderful and powerful tool in game design. It can be deployed to allow a broader range of players to have a meaningful game with one another. It can increase the variety that a game offers. It can test a host of skills like risk management and probability estimation that are hard to test in a deterministic game.

Games with high variance, lots of dice, or random events can have arbitrarily high skill. Wargames prove that. Poker and Backgammon prove that. By not appreciating what variance brings to the game, we are surely missing many great games. While I often view games from the perspective of a serious analytic player who is looking for game depth, there is also value in games that are less about player agency and more about being a spectator, such as the aforementioned Bingo and LCR.

One of my favorite types of games is a high-luck design with room for enough skill that there's actually somewhere an expert player could go, even if that expertise doesn't always mean victory. I would put Poker in that category, and it was an epiphany for me when working on the new version of Magical Athlete to realize that it didn't actually need that expression of skill to be enjoyable. I thought the drafting method included the original version — which was more skill testing — was very important to the game since it allowed for this expertise...then I discovered that switching to a less skill testing draft was faster and overall more fun.

There is absolutely skill to this very swingy game, but that is not why it is played, and leaning into the chaos worked better for it than working hard to introduce skill. I still want the selection and deployment of racers, as well as occasional decisions in play, to allow for clever and rewarding moments, but to think that the drafting strategy was ultimately the backbone of the game was simply wrong. There was nothing wrong with changing the game in that regard as the drafting was not what the players who loved the game were there for.

Conclusion

None of this is meant to tell anyone they are wrong about their taste in games or to declare there aren't amazing games that follow the standards of design.

It is about considering what else is possible that we don't know about. Those games might not be designed yet — or maybe they have been, and we can't find them since our current rating systems favor other games. This might mean that treasures are out there waiting to be discovered or languishing in obscurity, games that you might love or learn to love.

Even if you don't, however, if someone else would love them, our hobby profits from expanding the tent and welcoming more people. We are Homo Ludens — there are games for everyone.