Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Accatitippi]

I don't know magic well, but self-milling is discarding your own cards from the deck to get a powerful benefit, right?
Unlike discards in Dominion, discards in Magic are quite permanent.

If so, I think a better comparison might be trash-from-top-of-deck effects, which are some of the most polarising attacks in Dominion.
An analogue of self-milling would then be Leprechaun or Cursed village hitting War or Locusts, which while occasionally hilarious can lead to some negative experiences (I'm still trying to recover from the time I killed my own goat with locusts).

[Awaclus]

I don't know magic well, but self-milling is discarding your own cards from the deck to get a powerful benefit, right?
Unlike discards in Dominion, discards in Magic are quite permanent.

If so, I think a better comparison might be trash-from-top-of-deck effects, which are some of the most polarising attacks in Dominion.
An analogue of self-milling would then be Leprechaun or Cursed village hitting War or Locusts, which while occasionally hilarious can lead to some negative experiences (I'm still trying to recover from the time I killed my own goat with locusts).

It's true the comparisons aren't entirely accurate, but they're inaccurate in the other direction.

Removing random cards from your deck doesn't make any difference unless your entire deck gets depleted — there is a number of random cards that you don't get to draw because they're at the bottom of your deck, and if you mill some cards from the top, the number of random cards that you don't get to draw stays the same, only some of them are a different bunch of random cards that are now in your graveyard instead of being at the bottom of your deck. That's how it works in Magic because you typically only draw 10-15 cards per game from your deck that has 60 cards in it, and when it doesn't work like that, it's because your opponent is playing a deck that can force you to mill your entire deck and doesn't particularly care how many cards you have in it. In addition, Magic has a ton of reasons why it's good to have more things in your own graveyard, such as cards that give you benefits based on what's in your graveyard and cards that can be played from your graveyard.

[Donald X.]

Scrabble is a different effect. Using Mark's language, it is a Timmy experience. Even if you lose a game of Magic, it might be fun if you got to build a 200/200 creature (a fun experience!). But Spike will be unhappy. Scapegoats are for Spike. Timmy doesn't need scapegoats, and a game creating its value through Timmy things will have no trouble with losing. Here's an example similar to Scrabble: people don't mind losing Oregon Trail or wargames, because they have no mechanics anyway and it's all about the experience. Another example is Monopoly:

I don't see how Scrabble is for Timmy, not at all, and it isn't mechanic-less. Scrabble is about the joy of anagramming. That joy isn't tied to your score, that's the key thing.

Here's another question:

Players hate self-milling, and they will hate this. As one of many examples, look how little Arc-Slogger is worth compared to how powerful he is. Just stop submitting it - everyone
...
As it turns out, players generally hate hate hate self-milling so much that we would never actually do that

Whatever that card was, they still do self-milling all the time; an example from the most recent expansion: https://gatherer.wizards.com/Pages/Card/Details.aspx?multiverseid=466856

You found a similar effect with Tribute.

Some people feel like it's attacking them, since it can flip over good cards; I think it tends to help as much as hurt, but so what, I don't need people to feel bad over a non-attack

Logically, it makes no difference when something is milled - they just draw the card below that. But people hate milling anyway because of loss aversion. They weigh the loss of a discarded good card far higher than the benefit of a discarded bad card. So even though their emotional reaction makes no logical sense, they can't help it anyway.

The natural thought experiment is: suppose the milled card was face down, so they don't see which card was milled. I think this wouldn't create the impression of an attack. Is that true? Do you have experience with cards in either Dominion or other games that have this face-down-mill effect? How do players react to them? (Face down mill isn't possible for Tribute because of cheating issues.)

Milling in Magic and milling in Dominion are really unrelated. In Magic you give yourself a resource - cards in the graveyard - while maybe feeling bad - I would have drawn that. If you don't run out of cards, it's a positive effect, that some people don't enjoy. In Dominion you don't draw that card that shuffle - it's good or bad depending on the card. [And in Dominion, "milling" refers to lowering piles without gaining cards, e.g. Remodel Province to Province.]

I don't see what I would get out of secretly moving a card from your deck to your discard. I don't flip over cards just to flip them over; I look at the card and care what it is. Just moving the card isn't harmful enough to be an attack or penalty, or beneficial enough to be a bonus. It's busywork. I think it would feel like busywork; not an attack or benefit, but stupid. In Magic e.g. "put the top card of your library on the bottom" (if it goes to your graveyard, we'll know what it is) would be completely meaningless. Okay sometimes you'd have some information about your deck order, but aside from that, it's just shuffling a shuffled deck.

Tribute isn't self-milling; you do it to your opponents! Lots of cards can flip over your opponents' cards; what was special about Tribute was that it didn't say "attack" on it, and people didn't like that they couldn't Moat it. This is why Chariot Race doesn't discard the opponent's card (which some people then dislike for different reasons).

[whatwhen]

I don't see what I would get out of secretly moving a card from your deck to your discard. I don't flip over cards just to flip them over; I look at the card and care what it is. Just moving the card isn't harmful enough to be an attack or penalty, or beneficial enough to be a bonus. It's busywork. I think it would feel like busywork; not an attack or benefit, but stupid. In Magic e.g. "put the top card of your library on the bottom" (if it goes to your graveyard, we'll know what it is) would be completely meaningless. Okay sometimes you'd have some information about your deck order, but aside from that, it's just shuffling a shuffled deck.

Tribute isn't self-milling; you do it to your opponents! Lots of cards can flip over your opponents' cards; what was special about Tribute was that it didn't say "attack" on it, and people didn't like that they couldn't Moat it. This is why Chariot Race doesn't discard the opponent's card (which some people then dislike for different reasons).

You're correct that it's a pointless mechanic. I'm interested in how players respond because it helps me understand the overall design space. Some of the ideas I'll mention will be quite bad, but illustrative of concepts. Your comment that "it would feel like busywork; not an attack or benefit, but stupid" is perfect.

I didn't realize that players disliked Tribute because of the lack of the "Attack" word on it; thanks for that insight.

Here's my next question.

Start the game with a few easy options and slowly raise the number of choices as the game progresses.

In particular, Mark says that that a game's first few choices should be easy. We can look at how other games handle it.

Heroes of the Storm, a multiplayer action game: your actions at the beginning of the game make little difference. Respawns are instant, so killing your opponents barely matters, and dying also barely matters. You can't push towers or affect the board meaningfully. Early game is like warmup.
Dream Quest, a single-player deck-building game: equipment you pick up early can't be removed easily. That means early equipment decisions will clog your run later. So early decisions matter as much as late decisions.
Drafting games: all deck decisions are made up front before the game starts. The performance of your deck is mostly decided early.

Consider the balance between early/late decisions as a general concept, untied to the specifics of these examples. If early decisions are easy to make, it reduces the impact of those decisions. Should the consequences of early decisions weigh the same as late decisions, or should early decisions be weighted less? What about how obvious those choices should be? What are the underlying things that Rosewater sees, that make him feel his statement is correct?

(Let's pretend that early/late weighting, or early/late obviousness of choices, is a tunable knob that doesn't affect anything else in the game. For example, early decisions having outsize impact is an unavoidable downside of drafting games. It can't be changed without fundamentally altering the nature of the game, but we could close our eyes and pretend it can!)

Scumming is a consequence of early decisions mattering, but it can sometimes be solved, so let's ignore it. For example, if you are playing a multiplayer physical board game and you do poorly early, you won't flip the board and reset the game. Your friends would be mad at you.

As for what counts as early vs late, clock time seems like the most logical scale, since it's the scale on which you think and interact, as long as there is zero downtime.

One aspect I considered for this problem was Spike. If Spike fucked up early, then he will be suffering for the rest of the game and will want to flip the table. He doesn't care how fun the flavor is and how great the building-up feels. He wants to win, and feeling like a loser for a long time will make him miserable, and he wants it to end quickly. (You prefer to make games that avoid this problem. But let's assume the person is hyper-competitive; he will quit playing and find another game if he thinks he's losing, no matter what.)

However, Spike may not be the relevant factor that matters for early/late. In designing which choices are worthwhile to have in a game, my document states:

The purpose of choices is to promote learning. You make a choice and then receive feedback on that choice, learning about the quality of your decision.
If choices are too obvious, then you already know the right decision and so you don't learn.
If choices are confusing or nonintuitive, with unclear connection between the choice and the consequence, then it's hard to evaluate your decision.
The player should make a decision whose correct answer is not obvious, and should be able to relate future consequences to that decision. A decision being high-impact will make the connection between consequences and decision easier to see, but the impact is otherwise not important. What's important is the clarity of the connection.

The player judges quality of a presented choice by how much he expects it to teach him about something he cares about. Obvious choices will teach him nothing, and meaningless choices will also teach him nothing (or nothing that he cares about).
He may perceive an obvious correct answer, and later discover his "obvious" answer was wrong. This is a strong and excellent learning event, but it requires that later discovery to happen. Otherwise, his initial judgment of obviousness, and hence uninterestingness, will stand. (See: healing)
Note that if connections are clear, the player will learn very fast. That's good if you have lots of things to teach the player. If you run out of teachable things because the player has learned everything, the game becomes dead.
Teaching everything in a short time is not a good thing for a designer, even if it delivers all the same learning value in a shorter time. Because player perception of what they learned is not decided by how much they learned, but how much time they spent learning. So stretching out the learning is useful in tricking players to see a game as "deep".
Having fewer choices makes them more obvious and less confusing (by eliminating ones that the player shouldn't consider).

In particular, as the game progresses, the player should have some feeling of how the ongoing events are related to what he did before, helping him evaluate whether he made the right decision. If a choice is a black box, then the game designer should remove that choice. Is this the correct framework in which to evaluate why Rosewater wants early decisions to be more obvious? How come early choices require different treatment from late ones? Perhaps because they affect more duration of the game, and hence require a clearer connection between the choice and its consequences?

Perhaps the correct framework is instead clock time between information and the choice given. For example, in Heroes of the Storm, playing a new character is like a large burst of information. Then, you need some time to digest this information and form good decisions. Early/late weighting is a true tunable knob for Heroes that doesn't affect any other aspect of the game. So its early game being a warmup is a conscious decision by the game's designers, and they feel it's best this way. Its early decisions are not "easy" so much as "unimpactful". The connections are still clear, but they don't matter for victory.

[spineflu]

Do you use any sort of tracking/sorting/project/"productivity" type software for working on your games (like trello or hack2plan or whatever)?

[Donald X.]

Start the game with a few easy options and slowly raise the number of choices as the game progresses.

In particular, Mark says that that a game's first few choices should be easy. We can look at how other games handle it.

I don't know what Mark meant, but what I would mean by that is, a game should start low-information and progress to more information. The early decisions are easy in the sense that you have fewer decisions with fewer options, not in the sense that you obviously know what choice to make.

I have this problem in many of my games. In Dominion you have to read 10 cards on turn one. That's a lot to ask of players. Many of course don't do it, they just read the $4's if they have $4, even though to play well you want to read them all and think about them. It's better to present less information turn one; I just didn't because well I had ideas that involved lots of reading turn one.

Dominion does at least give players fewer decisions per turn early on.

Drafting games: all deck decisions are made up front before the game starts. The performance of your deck is mostly decided early.

You mean drafting e.g. Magic. "Drafting games" are games where you draft all game, like Greed; when the drafting is done the game is over.

Consider the balance between early/late decisions as a general concept, untied to the specifics of these examples. If early decisions are easy to make, it reduces the impact of those decisions. Should the consequences of early decisions weigh the same as late decisions, or should early decisions be weighted less? What about how obvious those choices should be? What are the underlying things that Rosewater sees, that make him feel his statement is correct?

I think Mark is just talking about something else. If someone actually thinks early decisions should be trivial, well, I don't see any reason for that.

However, Spike may not be the relevant factor that matters for early/late. In designing which choices are worthwhile to have in a game, my document states:

For me, Spike is not a relevant factor period. I just never think about the Magic psychographs when working on a game.

The game has to start out interesting me enough to work on it. Then I have to enjoy it enough to want to keep working on it, and the players who try it have to like enough too. If I pass those hurdles I will eventually think, okay, who is the audience for this game, and how can I make it better for them specifically. I'm never thinking, "but what about Spike."

The purpose of choices is to promote learning. You make a choice and then receive feedback on that choice, learning about the quality of your decision.
If choices are too obvious, then you already know the right decision and so you don't learn.

For me the purpose of choices is to have a fun game. You can only maximize one variable; I am maximizing fun. I don't need to teach anyone anything, or simulate the experience of learning, or any such thing. Fun.

I don't like having trivial choices; they waste time (which is no fun). In certain cases a trivial choice makes sense because it accomplishes something else, e.g. having "you may" on Moneylender so that we don't need to keep you honest (what if you want to play it but don't want to trash a copper, a rare situation but not an impossible one).

You can get some fun from exploring the strategic space of a game, which can include learning about choices. That's just not the root perspective I have though; I am maximizing fun, not learning. A choice can be fun with no exploring-strategic-space element.

Teaching everything in a short time is not a good thing for a designer, even if it delivers all the same learning value in a shorter time. Because player perception of what they learned is not decided by how much they learned, but how much time they spent learning. So stretching out the learning is useful in tricking players to see a game as "deep".

Making a game "deep" is trivial. It's never a focus for me because I can always get as much of that as I want.

I do think you need to sometimes push stuff in the players' faces that they want to be there but might miss. There were all those people who tried Temporum, didn't change history, and then said, "well changing history was sure nothing." The box doesn't include a good player to show you how they make out from changing history. So it wasn't enough that it had strategic value; it was too subtle for some players (and could have been less subtle).

In particular, as the game progresses, the player should have some feeling of how the ongoing events are related to what he did before, helping him evaluate whether he made the right decision.

In this day and age, it's "What they did before, helping them evaluate whether they made the right decision." Yes I am actually correcting you there.

If a choice is a black box, then the game designer should remove that choice.

It's not great to have decisions where players don't have a basis for deciding what to do, yes. It's also bad if they can "do the math," and you end up just sitting there doing the math all game. You want choices that are clear in how they matter but still don't let you easily tell what's better.

It's especially a problem with bidding games, though I've enjoyed some anyway.

Is this the correct framework in which to evaluate why Rosewater wants early decisions to be more obvious? How come early choices require different treatment from late ones? Perhaps because they affect more duration of the game, and hence require a clearer connection between the choice and its consequences?

I suspect he was just talking about how you want players to get to ease into the game. If you read a card a turn for 8 turns, at that point you've read 8 cards, and it's way less demanding of the player than reading 8 cards turn one.

[Donald X.]

Do you use any sort of tracking/sorting/project/"productivity" type software for working on your games (like trello or hack2plan or whatever)?

No.

In the 90s I used the Borland Pascal editor to make text documents of ideas, then wrote little programs in Pascal to make card images. These days I use Wordpad to make the text documents, and Paint to make the card images. I shrink google'd art in Gimp because Paint just takes every nth pixel, if you can believe that. Paint sucks, and in the future I have big plans to use e.g. Paint.net instead, but you know, haven't sat down and checked it out yet.

For Dominion, looking at Renaissance, there's a main file where I keep a list of the cards and work on ideas, a junk file that I move ideas into when I don't want to see them anymore, a file of the full card texts as they stand and all updates to it that I made along the way, a file with the recommended sets of 10 and some notes on those, and a file of articles - teaser, preview, secret history, artist notes, work on the funny paragraph. There is also a rulebook, but I have that in another directory where I have stuff that involves the publisher. And then I have the prototype card images files, grouped by normal/landscape and then just in the order I made the images.

[whatwhen]

You mean drafting e.g. Magic. "Drafting games" are games where you draft all game, like Greed; when the drafting is done the game is over.

You are right. I was not aware of Greed's mechanic.

For me the purpose of choices is to have a fun game. You can only maximize one variable; I am maximizing fun. I don't need to teach anyone anything, or simulate the experience of learning, or any such thing. Fun.

My "learning" comes from my analysis of the word "fun"; I'll provide a more careful description. If you look at things that are fun vs not fun, I think it comes close to "players have fun when something improves the way their minds organize things. Like realizing something new that matters, or a worldview-changing experience". I haven't seen examples that contradict this definition. Non-interactive things are fun too: books, movies, music, art, humor. A game has choices, which allows the realizations to be tailored to that person's individual profile.

The only reason to use this definition is if it allows us to analyze what is fun vs not fun. I hope it captures the behavior in a clearly understandable way. I'm also looking for better ways to think about fun, and ways to describe it more clearly.

Here's an example of fun:

In 2007, I attended the World Championship. One of my responsibilities was to play in a free-for-all multiplayer game and subtly show off a preview card from an upcoming set that no one knew I had in my deck. That card was Chameleon Colossus, and I wasn't particularly subtle...doubling its power eight times, I attacked with my Chameleon Colossus for 27,648 damage and gained 55,296 life.

A game designer shows up in your game, uses a card that doesn't even exist yet, creates a giant creature, and steamrolls everyone with it. All these things are new and unusual experiences, changing how people's minds organize things and the world - showing them something they didn't expect. So we expect it to be extremely fun. If Rosewater repeated it 10 minutes later, the impact of the fun would be way smaller, because the changes in everyone's minds have already happened. It won't change anyone's mental organization much further than it already did.

For me, Spike is not a relevant factor period. I just never think about the Magic psychographs when working on a game.

That's a very interesting perspective, and I think an ideal one. (Yes, I want to be that kind of designer too.) I think the reason you don't take Rosewater's psychographics seriously is that his psychographics are not totally correct, so they have unusual bits that don't make sense. Instead, I'll redefine them. Vorthos is part of Timmy. Melvin fits along with Timmy and Spike. Johnny doesn't exist; he is a different dimension and partially applies to all three. Here's the correct psychographics:
Timmy wants to have fun. Anything which shows him something fun is great.
Melvin wants to have fun, just like Timmy. But his learning is from inside, not outside. He figures out the solution, instead of the solution being presented to him. The realization is introspective. (Comparison: Timmy can have fun netdecking, but Melvin will have fun with deck creation.)
Spike wants to prove to others how great he is. Any effects on himself are a byproduct and not the direct cause - he might value winning, but only in the sense that it makes other people realize how great he is. Everything is in the eyes of others, and how he estimates others will see him. He may even take actions that reduce his chances of winning if he thinks it will improve how others view him, because winning is not the direct goal.* "Stuck in the Middle with Bruce" described seeing it happen:

He claims innocence by being unaware. While he may have had his head buried in the sand, Bruce had his eyes wide open...I'm referring to plays that you make that you know are incorrect. I'm talking about finding an excuse to lose. Ah there is no problem losing when you have a good excuse. Heck when you are all set with a battery of excuses it is so much easier to swallow the bitter pill of defeat.

Why care about psychographics? Because people are linear combinations of these three. In particular, the fact that Spike is such an extreme caricature doesn't affect the analysis; there's no threshold at which any of the Spike effects disappear. Everyone has a proportion of that extreme caricature. If someone is 10% Spike, he will have 10% of Spike's consequential properties - his low proportion of Spike-ness doesn't make him immune to Spike-related feelings. That makes it useful to extract the Spike-ness from him and analyze it separately. Spike is a consequence of how human friend selection works, although that's going out of scope for game-design discussion.

Why do I think your perspective on not designing for Spike is an ideal one? Because designing for Spike creates unpleasant design choices. I design for Spike for business purposes; I want a larger audience. To illustrate, one type of Spike exists on forums: in an internet argument, Spike will pick over small grammar and wording mistakes, argue over definitions, and argue over what everyone exactly said. He wants to win the argument and prove to others that he is better than his opponent. Discovering anything is not his main goal. He will spend many pages arguing about pointless minutia, and others will say, "why the fuck did you waste so much time on that", but it makes perfect sense to Spike. (If you see people like that on this forum, they are also purchasing and playing your game.)

Or think about Sirlin: his Playing to Win article wants to show others how he is better than the common house-rules crowd. In public self-descriptions, he hypes up his own games as receiving recognition beyond what they actually received, projecting himself as a genius designer. He didn't credit Matthan, because he wants to be respected as Puzzle Strike's sole creator. Sirlin has a strong Spike component. Whether his Matthan gambit was successful is separate from what his intention was. (I think it worked. People are rarely aware of the history behind Puzzle Strike; Steam reviews don't mention it.) So then, should Sirlin feel ashamed about stealing from Matthan? (Ok, he was caught.) But he probably sees credit-stealing as humanizable, because he views things through a Spike lens. Thinking that others are like him, he expects them to empathize with his action. That's why he feels less shame about being caught than someone less Spikey would.

Spike also gets in the way of Timmy. Spike is partially the cause of time overruns - long after Spike has a reasonable decision, he wants to double check and triple check everything to maximize his chances of winning. Timmy will say, "this is boring" and play quickly, so the stopwatch is not for him; it's for Spike. Online, at least. In person, friends can leave Spike out of the next game. But it's helpful to have some in-game mechanism that reduces it.

Johnny is for choices made from a huge selection, culling out many useless options to find the right one. Anti-Johnny is for choices made from a tiny selection, making accurate estimations better than others. In Magic, this usually manifests in a deck-origination vs piloting split. As an example of why my definition is more accurate, Gifts UngivenCabal Therapy feels closer to Johnny than anti-Johnny, despite being piloting. It gives that same freedom and will evoke that same feeling. Johnny and anti-Johnny are dimensions of all three of Timmy, Melvin, and Spike.

Making a game "deep" is trivial. It's never a focus for me because I can always get as much of that as I want.

I have trouble doing so. One obstacle to depth is randomness: in a random event, connections are no longer clear in the before and after, creating a barrier to learning. I've never played Ascension, so let's pretend my following description of it is true, even if it isn't. In Ascension, random events happen regularly. Nobody is able to predict the future very well, so when players make decisions, their future performance is not connected to their past in any clear way. After a game ends, how should they change their behavior? They don't know. Even though there is certainly room for them to improve their decisions, they are not able to, and hence the game is not deep. The depth is there, but they can't achieve it, so it effectively doesn't exist. Poker is like this: depth is there, but learning is slow, and many people stop learning before they hit the maximum.
How to solve the randomness obstacle and improve depth? One way is to unrandomize. Flip future cards upside-down. Instead of seeing the next 6 cards in the Ascension buy-pile, now they see 30. (Ignore the massive reading problem this creates, for now.)
And it creates a triviality problem: depth only matters for ways that people care about. This same unrandomization technique applies to many games. For example, in a card game, you can make everyone turn three cards face-up from their library each turn. They only learn 3 new cards per draw, solving the reading problem. When stuff happens in late game, they can trace consequences back to previous bad choices, since the information is available. But players start thinking, "this is just a memorization game where people compete to see how many increasingly trivial and irrelevant things they can keep in mind, who cares about that?", and it's not deep in a way anyone likes.

A really good game like Go manages to create depth in a way that people can keep improving; in addition, players continue to care about what they are learning. But I've never designed anything as good as Go.

After unrandomization, the read-many-things problem can often be solved. But it's hard to create depth in a way that people think, "this type of depth matters". Often designers veer into the "this game is deep because people can't keep 100 things in their head, and they need to" territory. Or veer into "this game is deep because people can't add 100 numbers". Or "people can't read 100 things quickly". Different players care about different things too. Clicking quickly on pixels sounds like stupid-type depth to many players, but other players like FPS games and Osu. The type of depth players care about is significantly influenced by how it's framed within theming and drama.

As an actual illustration, imagine if Dominion used 10 cards the first game, then added another 10 cards the next game, 20 total. Then every game, rotate out the most stale 10 cards and add in 10 more cards, so it stays at 20 until the group breaks up. Whether Dominion works with 20 cards is not my point, I have no idea if it does (I haven't played Dominion either). But note what happened - the "I don't want to read 20 cards" problem has been solved - they're still only reading 10 cards at a time. But now we have the "players don't like needing to keep 20 cards in their heads" problem. Maybe some do.

Another illustration is graveyards. After playing a card, it goes in your graveyard instead of being shuffled back into your deck. Only when the draw pile empties is the graveyard shuffled in. This improves consistency and creates depth, but counting and remembering cards in graveyards is "stupid-depth" for many players. The main benefit is really the consistency. (And also, the fact that constant deck shuffling is not possible. Bags and videogames can continuously shuffle, but things in bags have limited text and art.)

I do think you need to sometimes push stuff in the players' faces that they want to be there but might miss. There were all those people who tried Temporum, didn't change history, and then said, "well changing history was sure nothing." The box doesn't include a good player to show you how they make out from changing history. So it wasn't enough that it had strategic value; it was too subtle for some players (and could have been less subtle).

Wow, that theme is great. I thought you sucked at theming but now I completely changed my mind. Temporum's theme is really well-chosen, and is still strong even as a theme chosen before mechanic.

I agree, it's not nice when the player just doesn't see something. Rosewater told a story about Magic:

During the first such playtest, we noticed that when players got an Eldrazi out, they wouldn't attack with it. R&D had quickly realized the power of the annihilator mechanic, but the less experienced players didn't seem to feel safe throwing their giant creatures into the fray.
We took notes and talked about it, but changed nothing. The next playtest the same thing happened, and the playtest after that... we added the text "CARDNAME attacks each turn if able." Once players saw how much damage the Eldrazi did when they attacked, hopefully that would encourage them to attack with the other Eldrazi.

Basically, when players don't realize something, the designer forces them to do it and see the fun. It's hard to predict when this is needed without newcomer playtesting. Magic's huge financial base really helps them here.

Healing is a personal example: after playing 5 games of Magic, I didn't yet realize that late game I would have sufficient mana to do whatever I wanted. So if I had spare mana to heal and nothing else to do: I would heal. It didn't cross my mind that there was even the decision to be made, of delaying my heal to obscure the opponent's reach. Someone had to tell me. So my judgment was, "healing is trivial". But I was wrong. It looks nice after I learned it, but if I hadn't learned it, I would still keep thinking, "it's trivial".

In this day and age, it's "What they did before, helping them evaluate whether they made the right decision." Yes I am actually correcting you there.

Sure. I don't follow that for my own writing, but I understand both the cultural context and the linguistic context of your choice. You're not wrong. Related is my preference for writing "publically" and singular "data" despite them not being common. I switch some of these choices depending on the audience I'm writing for. I keep Spike strictly masculine because he is strongly associated with negative connotations; this is from awareness of a cultural context. Otherwise, I often switch between random he and she for uncertain third parties.

It's not great to have decisions where players don't have a basis for deciding what to do, yes. It's also bad if they can "do the math," and you end up just sitting there doing the math all game. You want choices that are clear in how they matter but still don't let you easily tell what's better.

This relates to the "I don't want to add 100 numbers" type of depth. Unrandomizing things makes connections clearer, but the created depth should be interesting, not pointless math. (I heard that Mage Knight is like this.)

I suspect he was just talking about how you want players to get to ease into the game. If you read a card a turn for 8 turns, at that point you've read 8 cards, and it's way less demanding of the player than reading 8 cards turn one.

Thanks, that fits in with "clock time between information and the choice given...to digest this information and form good decisions." But your perspective is better, because the player sees interactions play out to help make a better choice, not just spending time internally introspecting on the choice.

Here's my next question.

if players are taken out of the bidding by say winning an auction, they will get screwed over by the other players turning out to undervalue things; ideally good players can stay in the auction as long as possible

I played a multiplayer game called Offworld Trading Company. In this game, there are exclusive locations, each accommodating one player only, and some locations are better. Players choose exclusive locations, one at a time, so later choosers get locked out of the best locations. To determine the choosing order, an auction system is used: a clock ticks down saying "if you choose now, you will get $X debt", and X ticks down over time. Players decided how much debt each choice was worth, then pressed the "choose now" button when it was worth it. The final chooser got 0 debt, to remove collusion between the final two players. However, the issue you described came up: players only have the ability to influence one of the auction slots, and then were taken out of the auction.

The model can be summarized as: many options, players bid on the privilege of choosing first, then on the privilege of choosing second, etc. What auction systems improve on the one described, of bidding one at a time for each successive slot? (Such as reversing choice order, or bidding on differences between choice prices instead of absolute choice prices.) How do I make the most confident (best) price-estimator arbitrage his estimate against others, to give more input on prices? This was a major problem in Offworld, that new players did not know how to value things, lacked confidence in their estimates, and hence felt shut out from the very beginning. Offworld's designer Soren Johnson originally had an exploration mechanic to replace this auction one. However, the competitive community strongly advocated for the auction mechanic, so Soren removed the exploration mechanic, leaving new players in the dirt. The experienced players had no issue with auctions, and Soren had no ability to receive feedback from newcomers (they weren't in his Discord), so this caused problems that he could not see. The exploration mechanic was terrible, but it was more newbie-friendly.

*Note that winning tournaments for prize money is also desired. But we'll accept Rosewater's definition of a game as including no real-world consequences. I think it's a very accurate definition.

[pacovf]

I feel like you may want to make your posts a bit shorter. But maybe that’s just me.

[chipperMDW]

In designing which choices are worthwhile to have in a game, my document states:

Is this referring to a document you're writing?

[Accatitippi]

I don't know magic well, but self-milling is discarding your own cards from the deck to get a powerful benefit, right?
Unlike discards in Dominion, discards in Magic are quite permanent.

If so, I think a better comparison might be trash-from-top-of-deck effects, which are some of the most polarising attacks in Dominion.
An analogue of self-milling would then be Leprechaun or Cursed village hitting War or Locusts, which while occasionally hilarious can lead to some negative experiences (I'm still trying to recover from the time I killed my own goat with locusts).

[. .]
In addition, Magic has a ton of reasons why it's good to have more things in your own graveyard, such as cards that give you benefits based on what's in your graveyard and cards that can be played from your graveyard.

I wasn't aware of this, thanks. That being the case, I really can't see the problem with self-milling...

[Accatitippi]

From the Time Machine series, we know about the mechanical stuff you'd change if you had full power to change the already-released, already sold sets.
Is there anything non-mechanical in Dominion that you'd change - or ask RGG to consider changing? I'm thinking about stuff like card names, card art, card sizes, layouts, box size or shape, even marketing strategies, sale model, theme, whatever. For instance, would you change the card backs if you could?
We already know about Harem and the Event back miscoloration blunder. Is there anything else?

[GendoIkari]

I don't know magic well, but self-milling is discarding your own cards from the deck to get a powerful benefit, right?
Unlike discards in Dominion, discards in Magic are quite permanent.

If so, I think a better comparison might be trash-from-top-of-deck effects, which are some of the most polarising attacks in Dominion.
An analogue of self-milling would then be Leprechaun or Cursed village hitting War or Locusts, which while occasionally hilarious can lead to some negative experiences (I'm still trying to recover from the time I killed my own goat with locusts).

[. .]
In addition, Magic has a ton of reasons why it's good to have more things in your own graveyard, such as cards that give you benefits based on what's in your graveyard and cards that can be played from your graveyard.

I wasn't aware of this, thanks. That being the case, I really can't see the problem with self-milling...

The problem is purely psychological. When you put a card you wanted or needed to draw into your graveyard, you feel bad. Even though your chances of drawing that card were exactly the same if you put cards from your library into your graveyard or not.

[Donald X.]

My "learning" comes from my analysis of the word "fun"; I'll provide a more careful description. If you look at things that are fun vs not fun, I think it comes close to "players have fun when something improves the way their minds organize things. Like realizing something new that matters, or a worldview-changing experience". I haven't seen examples that contradict this definition. Non-interactive things are fun too: books, movies, music, art, humor. A game has choices, which allows the realizations to be tailored to that person's individual profile.

I don't think fun is so tied to mind organizing. Eating a chocolate bar is fun. Some games are about the joy of making decisions; many are not. The first chapter of Knizia's Dice Games Properly Explained is all games with no decisions.

If Rosewater repeated it 10 minutes later, the impact of the fun would be way smaller, because the changes in everyone's minds have already happened.

I personally am a big fan of novelty. But it's not the only fun thing.

I think the reason you don't take Rosewater's psychographics seriously is that his psychographics are not totally correct, so they have unusual bits that don't make sense.

I think it's reasonable to say, some Magic fans like combos, let's make sure there are combos for them; some like big creatures, let's have some big creatures. You can give names to those people if you want, and maybe that helps you have conversations, so they did. Which psychograph likes 1/1's and lifegain, that's what I want to know.

I am not working on Magic though. In Dominion, like, there are players who especially like attacks, and players who hate attacks. It's not Timmy Johnny Spike, none of those hit on this. I should try to make attacks for the players who like them and non-attack interaction to have interaction in games where the players decide not to have attacks. I don't have enough of the relevant conversations to name the players who like and hate attacks.

I design for Spike for business purposes; I want a larger audience.

I don't get that; consider that for Magic, something like 1/3 of the players are Spikes (from some article I barely remember - I'm not picking 1/3 because there are three psychographs). They buy more total cards than non-Spikes, despite there being fewer of them. But you know, if your game doesn't bleed your customers, you'll make more money selling to non-Spikes. And like, the classic mass market hit genre is party games, not thinky euros.

Making a game "deep" is trivial. It's never a focus for me because I can always get as much of that as I want.

I have trouble doing so.

I am always pursuing variety and interacting cards (or other elements with rules interactions), and that just naturally makes it take longer to learn everything there is to know. For a simple abstract, it's harder, you have to find a way to generate the variety and interactions out of your simple system.

As an actual illustration, imagine if Dominion used 10 cards the first game, then added another 10 cards the next game, 20 total. Then every game, rotate out the most stale 10 cards and add in 10 more cards, so it stays at 20 until the group breaks up.

For a while at first we actually tended to play with 10 cards, then swap out just 2-4 for new ones for the next game.

To determine the choosing order, an auction system is used: a clock ticks down saying "if you choose now, you will get $X debt", and X ticks down over time.

That's called an auction clock or Dutch auction clock. The Dutch used them!

What auction systems improve on the one described, of bidding one at a time for each successive slot?

I think bidding games are fundamentally flawed. I've had fun playing them though so what does that matter. I don't have a perfect system, but I did make a bidding game, Gauntlet of Fools, and can tell you about that.

The basic bidding is based on Evo's, just slightly tweaked. We take turns bidding on heroes. If you have a hero already you don't bid. If you don't you can either outbid someone and take their hero, or bid on an unclaimed hero and take it. When someone takes the last hero (for free unless they're really boasting) we're done.

This still has the issue of, I can place the perfect bid on a hero, expecting perfect play from my opponents, only to find they underbid on everything remaining. Or I can place a minimal bid on a hero, and then watch everyone else place perfect bids. But in practice, it worked well; you can start out cautious and see how things go.

You could make something much more complex that went further. Everyone could have a bid on each hero, all starting at zero. We take turns, on your turn you either raise a bid on one of your heroes to be the highest bid on it now, or pass. When we all pass, we go around and claim a hero we have the highest remaining bid on (someone could win multiple heroes but I'm assuming we still want each player to just have one).

Gauntlet of Fools had another wrinkle to its bidding. Instead of bidding VP as is most common, you bid rules that made your hero worse. I can win with one arm tied behind my back! One Arm Tied Behind Your Back made your 1's and 2's into 0's; With a Hangover gave you worse attack and defense until you'd killed a monster, and so on. So a bid would vary in value from hero to hero - a hero with 5 dice is hurt more from tying the arm than one with 3 dice. And you could saddle a hero with a bid, not expecting to get that hero, just wanting that hero to be weaker whoever gets it. You could put a bid on a hero that you thought would make that hero look bad to someone else, to try to get it cheaper. And sometimes the bids interacted between heroes, e.g. one hero (weapon, they were class/weapon combos) can finagle which monster we're facing a few times, which really hurts anyone With a Hangover... unless that hero also has With a Hangover.

Gauntlet of Fools was a dud and is out of print; reviews would tend towards, we didn't know what we were doing so we didn't bid, in this bidding game, and then the person with the best hero won, that was stupid. As if you would play Medici without bidding and that would tell you how good Medici was. In retrospect I combined two game elements I enjoyed, bidding and rolling lots of dice, that tend to be liked by different people. I should have made it an economic game instead. It was clear that it was a direction, I just was enamored with the boasting flavor - "I can win while juggling and without breakfast."

[Donald X.]

From the Time Machine series, we know about the mechanical stuff you'd change if you had full power to change the already-released, already sold sets.
Is there anything non-mechanical in Dominion that you'd change - or ask RGG to consider changing? I'm thinking about stuff like card names, card art, card sizes, layouts, box size or shape, even marketing strategies, sale model, theme, whatever. For instance, would you change the card backs if you could?
We already know about Harem and the Event back miscoloration blunder. Is there anything else?

I didn't have any authority on most stuff in the early days. I would certainly change the name of the game (it sounds like a war game), the logo, the card back, all the art I don't like. The card size worked out okay, Magic size would have been nice but I don't know if it would make the game more expensive or what, and people can now get sleeves for this size. The box size, man, it's addressing a problem while looking good on a shelf. I'm not even interested in marketing stuff. The theme is fine. I'd probably change some names, sure, it's something I didn't spend much time on in the early days. In the prototype the VP symbol is a crown, which is just better than a shield, man, shields are for defense. The frames could be prettier for sure.

I'm not sure what you mean about Harem, unless it's just, why is there a card called Harem. Yes Events were supposed to match randomizers, though these days I would probably ditch randomizers and have you use a card from each pile like I do.

It would have been great to go straight to some real established company for online Dominion. It would have been better to not make small sets, although then there would be a larger Alchemy set that some people didn't like. The rulebook could have been better but hey now it is. I should have devoted some time in the early days to making spin-offs without requiring them to be much different from Dominion.

I would at least try out icons for the basic +'s. I didn't at the time because I didn't have good ones and it's not much different. The space savings now that I have a bigger font would be great though, cards can use up so much space on the +'s, meaning that some ideas just don't leave space for them, and are stuck with e.g. "can this be +$3 or +3 Cards." But as icons I could put them all on one line (and have it look nicer than it would if we did that with the text ones).

[Wizard_Amul]

Gauntlet of Fools had another wrinkle to its bidding. Instead of bidding VP as is most common, you bid rules that made your hero worse. I can win with one arm tied behind my back!

In retrospect I combined two game elements I enjoyed, bidding and rolling lots of dice, that tend to be liked by different people. I should have made it an economic game instead. It was clear that it was a direction, I just was enamored with the boasting flavor - "I can win while juggling and without breakfast."

I personally don't particularly enjoy bidding games or dice rolling, but I really like the sound of this idea conceptually.

[Jashezilla]

Hey Donald! Just wondering, will there ever be any good head-way made with a Dominion phone app? Been itching to play it on the go for ages now. Thanks heaps!

[whatwhen]

Rosewater talks a lot about market research and focus groups. In Drive to Work 333, he describes some of the starter questions he asks: "how do you feel about X", where X can be important mechanics of a block, overall flavor, or certain defining cards. (This episode is not too worthwhile.) At the end of a focus group, the question asked is, "Would you buy this product", as a more-objective overall indicator of performance, even as a Yes/No single bit of information. Their market research isn't very sophisticated, but its sheer volume creates significant and valuable information.

You don't have focus groups, but you do a lot of playtesting. Players usually don't bring up things until prompted. Even when asked Rosewater-style questions, they're not coherent in expressing their underlying emotions. For example, people might dislike attacks, or enjoy certain feelings like going infinite, or feel that complexity is too much, or feel discouraged by certain choices required of them. But they'll usually point to some concrete and specific symptom rather than the underlying true issue.
Usually, starter questions reveal an initial symptom. Then careful followup questions guide players to discover the core issue, helping them realize why they feel a certain way. After they realize the source of their feelings, they can then tell that insight to the interviewer.

What starter questions are useful in finding the initial indication of an issue, when players may not understand it or be able to articulate it? Followup questions are easy and situation-specific, but I find that the initial discovery of those issues is the hard part, so my question focuses on starter questions.

Example: Magic's market research found a pattern that players don't like downsides on large creatures. Individual players point out specific cards they don't like, usually as vague feelings, but can't express why they feel that way. Then Magic connected the dots and found the underlying pattern. (I'm simplifying; the pairs are more subtle than just "large creatures", and some types of downsides are ok.)

[crj]

I feel like you may want to make your posts a bit shorter. But maybe that’s just me.

It also feels like it might be better to fork off the detailed compare-and-contrast with Magic into its own thread.

[Donald X.]

Hey Donald! Just wondering, will there ever be any good head-way made with a Dominion phone app? Been itching to play it on the go for ages now. Thanks heaps!

Yes, there has been no formal announcement, but someone is working on it now, a real company with published games.

[Donald X.]

You don't have focus groups, but you do a lot of playtesting. Players usually don't bring up things until prompted. Even when asked Rosewater-style questions, they're not coherent in expressing their underlying emotions. For example, people might dislike attacks, or enjoy certain feelings like going infinite, or feel that complexity is too much, or feel discouraged by certain choices required of them. But they'll usually point to some concrete and specific symptom rather than the underlying true issue.
Usually, starter questions reveal an initial symptom. Then careful followup questions guide players to discover the core issue, helping them realize why they feel a certain way. After they realize the source of their feelings, they can then tell that insight to the interviewer.

What starter questions are useful in finding the initial indication of an issue, when players may not understand it or be able to articulate it? Followup questions are easy and situation-specific, but I find that the initial discovery of those issues is the hard part, so my question focuses on starter questions.

Some playtesters are more talkative than others. Some just don't want to say bad things, and you have to glean what you can from how the game went. Some will say, I hate that card, I love this one. You only have so much choice of who playtests, so you just have to get what you can from who you've got. Things like "they hate attacks" are never subtle though.

I find out how much people like things by playing games. Sometimes I'll ask them what they think, but generally I can tell when people are having fun. For external groups, I get what they volunteer, and then sometimes say, okay can everyone say a sentence about each card, can you rate them. For a new game, even if people aren't talkative, you will get a measure of how much people want to play it.

It's easy to find out what's hard to understand, at least for the caliber of player that's playtesting. Players will not understand it. In my games they will ask, wait what; for the external groups, they'll say, hey we have a rules question. It's harder to know what very casual players will find tricky; you have to find some and play with them. You can stare at the cards though, and try to see, how could you misread this. Head playtester LastFootnote plays some with more casual players.

My operation is just not at the level of a big company with focus groups. "Starter questions," I have no such thing. "Let's play some Dominion" is where I start.

[Hockey Mask]

Do you have  playtesters of different calibers? Is it important to have greener players to ensure that wording and gameplay are “dumbed down” enough?

[Donald X.]

Do you have  playtesters of different calibers? Is it important to have greener players to ensure that wording and gameplay are “dumbed down” enough?

The playtesters have a range of skill levels. This isn't because I carefully picked them out; I'm playing with the people who want to play. It's great to get in some games with some more casual players, when it's possible. The Hexes in Nocturne were too complex, and this wasn't obvious until the release party, where I saw those players having to deal with them.

[GendoIkari]

I just noticed that Empires has no events that are once per turn, or once per game; while Adventures had several. Was this a specific design decision; something you didn't like about how the once per turn played out in Adventures?

[Donald X.]

I just noticed that Empires has no events that are once per turn, or once per game; while Adventures had several. Was this a specific design decision; something you didn't like about how the once per turn played out in Adventures?

There's no specific draw to doing either, is the main thing; I'd rather let you keep paying for the Events to keep doing them. I'd only use either if it seemed the only way to do something, which for a few of those it did. "Once per game" is especially exotic and just didn't come up for Empires. "Once per turn" did have an issue in Adventures - some Events had both "once per turn" and "+1 Buy" and I notated this poorly, so that it confused people. You can just notate it correctly though as the later printings do.