ast week, I solicited questions through both email and my Twitter account. This week, I'll be answering them, as it's the end of Feedback Week! I've done this two times before, which you can find here and here.

Before I start answering questions, I'd like to give some feedback on the questions I got. Many of the emails I received were long, some asked several different questions that weren't very well connected, and a few asked questions that didn't look very interesting to answer. These all made me less interested in answering them.

You'll have the best luck getting a response from me if you ask one very short question that interests me. Twitter happens to be the perfect vehicle for getting such questions, as 140 characters is only enough space for one short question.

Congratulations, then to Will, who asked the only question asked via email that I'll be answering today.

Will asks:

How many turns is the perfect game of Magic?

From a game design perspective, this is really a question about narrative arc. Every game of Magic is like a little story. There are certain beats that we want to hit in each game, with a rising action, a satisfying midgame in which both players get to contribute, and a climax that decides the winner, hopefully happening just before the actual end of the game. The right speed of that arc depends on the format; players who want faster games with crazier things happening tend to play older formats, in which cards like Hive Mind or Tendrils of Agony can end a game as early as turn two. In Standard and Limited, though, we tend to keep the arc of a game a little longer, and most games last at least until turn four or five. In the end, there's no hard and fast answer to this question; as long as the ending of the game is satisfying to both players and the process that got to the end was enjoyable, then it lasted the right amount of time.

Now, off to Twitter.

Very early on. Design doesn't usually think about it a whole lot, but development usually has a good idea of what it wants to do almost immediately based either on the set's mechanics or on how fast the sets surrounding it are. As soon as I knew bloodthirst was in Magic 2012, for example, I knew that the format would need to be fast, and I was already leaning that way in order to contrast both Magic 2011 and Scars of Mirrodin block. I also knew how fast I wanted Innistrad / Dark Ascension Limited to be about a month into drafting triple Innistrad, although I can't tell you any details about that yet.

There absolutely is such a thing as "too much development," and it can take many forms. One form of this is when the cards are all so flat that none of them stand out as exciting, which is what happened with sets like Fallen Empires, Homelands, and Mercadian Masques. Sure, nearly everything in those sets is fair, but if that's what being fair requires then I'll take unfair any day of the week.

The other form that overdevelopment takes is adding random text to a card that is intended to be strong against a particular card or strategy, but that doesn't make sense on the card. My favorite example of this is Goblin Piledriver, which would be a gorgeous and perfect Magic card if it weren't for Protection from blue. It has that ability so that it can sneak past Psychatogs, but Psychatog is long gone, and if we were to reprint Piledriver now, that ability would be inexplicable to a newer player.

We try to keep the power level of sets fairly consistent across time. Remember the "perfect game of Magic" question from earlier? We want Standard to stay fairly consistent in speed, so that players know what to expect when they show up. Raising or lowering the power level would shorten or lengthen the average game, and we don't want to violate expectations too much.

When deciding what cards to push in power, we do our best to push things that are fun to be good. We often ask ourselves whether a world in which a card was very powerful would be a fun world. With something like Lightning Bolt, we decided it would be a fun world, but we have avoided making extremely powerful Fog effects, as playing a bunch of mirror matches between Fog decks at Friday Night Magic sounds really awful. Yes, this means that we have to make qualitative judgments about what's fun and what isn't, but that's secretly our job anyway.

A straw poll I just ran among the people sitting around me indicates that we don't feel very good about it, but we print things like Gideon's Lawkeeper and Blinding Souleater often enough that I'm not sure we all agree. Personally, I think common tappers are something that we should have around some of the time, but that we shouldn't have around all the time. Gideon's Lawkeeper was just so perfect for Gideon that I couldn't resist, for example.

I can think of only one example of doing this, and it was on a card that we'd like to make one day so I can't tell you much else. Much more often, we tweak cards to make them slightly friendlier to online play. For example, cards like Demonspine Whip and Helix Pinnacle let you activate for instead of just paying for +1/+0 or one counter, which saves a lot of clicks. We also avoid making things optional in very marginal cases in order to further reduce clicking.

It has given me more trepidation about giving people really good mana fixing. The best thing I ever did to my cube since I got here was cut all the Signets and Ravnica bounce lands. The Signets gave every color strong mana acceleration, the bounce lands gave slower decks free card advantage in their lands, and both cycles gave far-too-easy access to any color of mana you wanted. Removing both cycles meant that green had a near-monopoly on both mana acceleration and multicolor fixing, which changed green from a second-class citizen that didn't have much reason for being played into one of the better colors.

I've also gained more respect for building weird Constructed subthemes like reanimator into colors and making sure that each color and color pair has a distinct identity, but I think cutting the signets and bounce lands was by far the more important change. If you have a cube, you should try cutting them too!

Yes. The most recent example of this was white-blue Mass Polymorph, which we had as a very good deck upon the release of Magic 2011. Zac Hill brought it to US Nationals last year and only lost a single game out of around twenty he played with it during spellslinging the whole weekend. Zac has since remarked that he thinks that of the decks in the top eight of that tournament, only Josh Utter-Leyton's winning deck was stronger.

That depends on what we want to do. If we want to take a card completely out of consideration for competitive constructed on the grounds of being miserable, we add mana. If a card is way too good and we'd play it even if it cost more mana, we add mana until that isn't the case anymore. If we want to do more granular changes, we usually change other numbers. For example, we didn't want Garruk, Primal Hunter to go into decks based around Valakut, the Molten Pinnacle and Primeval Titan, so near the end of development we changed his mana cost from to , keeping his total cost the same but restricting the kinds of decks that could play him.

That depends on the timing. Early drafts are very casual, as we're always throwing opinions about cards around. We build the last few drafts of each set to include very strong players in order to simulate a Pro Tour or Magic Online draft, and at those drafts our competitive juices get flowing quickly enough that we tend to be pretty serious.

We mostly just compare notes. We don't draft so many times that analyzing by the numbers makes much sense. We're all pretty experienced drafters, though, so we usually don't have trouble figuring out when things are off kilter and what we need to do to fix it.

That depends on the kind of set you're working on. For example, early in Magic 2012 development Bill Rose made an offhand comment that it would be nice to have an uncommon dragon that is not creatively a baby dragon that we could put into ancillary products. I cleverly designed Volcanic Dragon, then discovered that it had already been made. That was fine with me, as I could just put Volcanic Dragon in the set. We've been more willing to put straight reprints in expansion sets lately, though, and you should expect to see a little more of that as time goes on.

People often assume that we control the creative elements of the card. While we can influence them, they are not our domain. I don't decide whether Lurking Crocodile is a Crocodile or a Hippopotamus, for example, but I usually don't care much.

In terms of powers I wish I had, I wish I had the superpower to always know what everyone around me actually wanted from me at any given time. That would be really useful! With regard to Magic R&D related powers, I can't think of any.

Just after a set enters Future Future League playtesting, quite often. Near the end of a set's development, quite rarely. As you might expect, we are more adventurous when we have more time to playtest before things become final. Changes we make later on tend to be conservative.

No. The whole card frame definition is kind of misleading, so let's reframe it entirely. Modern consists of all booster releases beginning with Eighth Edition and Mirrodin. This definition will remain the same forever, and cards will not rotate out of the format. New cards released outside of booster releases, such as those in Magic: The Gathering Commander, are not part of Modern, and will not become part of it in the future.

Okay, we're at the end. I guess I haven't shown you any dragon emperors yet, though. Luckily, I found one anyway! Here he is:

Thanks for all the questions. As always, you can find me on Twitter at @tomlapillemagic and via email at the bottom of this article.

Last Week's Poll

This Week's Poll

What color do you most like playing against?   White Blue Black Red Green