Testing for Exploration vs Testing for Fine-Tuning

I've played a few Sealed matches of our virtual M13 set and I've been watching how people draft the set in a Community Draft that I'm running in the forums. (A community draft is like a simulated draft in the Limited forums. The forum viewers collectively vote on the picks made by one seat in the draft, while the other seats are drafted by the moderator or some collaborators.)

My impressions are:

• It's actually kind of fun!
• The fun might not last...
• Bond isn't actually doing anything in the set??
• This is going to be hard to playtest...!!

Here are the details of what I mean.

It's actually kind of fun!

This set is somewhat fun already, which was genuinely a surprise for me. I expected it to be terrible at first.

First of all, we are sticking to mostly core set stuff, and have lots of guidelines as to what an ability should cost. As MaRo once wrote, Magic with just normal cards like French vanilla creatures is still very fun, and partly it was fun like that. Another part must be due to design members individually having tweaked the set through playtests with their playgroups, removing unfun cards in the process.

Second of all, our power level is way too high compared to an ordinary set. Which is a problem, but this kind of creates a feeling of intoxication as the cards are better than what I'm used to in an ordinary core set. Look at these Uncommon cards. All of them are a savage beating. (Click on the images to enlarge).

How does this compare to a real core set? The tendency in Uncommon from M10-M12 is:

• Each color gets 1 clear "Ace" creature.

This is your Air Servant/Cudgel Troll/Serra Angel/Howling Banshee slot. Red has gotten cards like Dragon Whelp or Volcanic Dragon here. Sometimes the 2nd best creature is almost as good, such as Water Servant or Awakener Druid. But in our set, we have 4 of these guys.

• Utility, efficiency, build-around, defense...each color gets a variety of creatures.

The real life core sets didn't have all awesome beaters at Uncommon. But that doesn't mean it's about nerfing them to boring levels, but rather about making them good in a different way. A White Knight is different from a Serra Angel, which is different from a Roc Egg

In M11, Red had Fire Servant and Earth Servant as the big guys, but it also had a tiny but a powerful utility creature in Prodigal Sorcerer (like other colors might get something like Alluring Siren or War Priest of Thune here). Ember Hauler is efficient and nifty rather than overbearing, much like Black Knight. Cards like Chandra's Spitfire and Fire Servant provide a build-around-me factor. Also, while Red doesn't have one, most colors have one defensive creature such as Wall of Frost or Reassembling Skeleton.

So it's not "Uncommon = powerful beater," but it's also not "make some of them awesome and some of the boring in order to be balanced." It's about making them each good in a different way.

Anyways, when I open up my Sealed pools and see all the numerous potentially game-winning Uncommons and Rares in each of my colors, it's kind of exciting and I think "how do I put in as many of them as possible to abuse them." But there's not that much depth to that and as I get used to the power level, I don't expect this kind of excitement to last.

The Offensive Commons

The second reason it feels fun (for now) is that we have too many good common attackers. Both sides are attacking. I guess we did a good enough job of nerfing defensive creatures, and there are many evasive creatures, some with intimidate. I have no trouble getting a power 2 or 3 attacker on the board right away, and the opponent also does so to put me on my toes.

The variety in creatures is low. The creatures are mostly offensive beaters of one type or another. Here's a list of the 2-drops for each color: (Once again, you can click on card images to enlarge them.)

As you can see, in every color except Blue, EVERY 2-drop has power 2 or higher (or the equivalent of that, as 2 tokens). What we are doing is funky to say the least. In most sets, if there are two creatures in the same converted casting cost in the same color, usually one of them is an offensive creature with on-curve power and the other is some kind of utility creature with some nifty ability. Here are some examples from M11:

With the M11 commons, every color has at least one creature with off-curve power that serves some utility/defensive purpose. Green and White, being the creature colors, have multiple 2-drops with on-curve power, but Red and Black each have one on-curve creature and one off-curve creature. Blue being a weak creature color has no on-curve creatures at 2cc.

What this achieves is that the creatures feel more varied, and you can build your curve differently based on each deck's plan. If they were all offensive beaters, they will all be interchangeable substitutes for each other. Some would be better than others, but they would all serve the same purpose of attacking quickly. (Of course, another point to consider is that decks can differ not just in what kind of 2-drops they want, but also in how many 2-drops you aim to draft for that deck. But still, it helps if decks can differ based on the kind of 2-drops you draft, not just in the quantity of 2-drops that you draft.)

Also, when every 2-drop has on-curve power, it could lead to a hyper-aggro environment, so putting in utility creatures helps slow down the pace of the environment.

The fun might not last...

Currently, it's somewhat fun because these threats appear on both sides of the board and attack. However, I also get the feeling that maybe in 4-5 more sealed matches, it would start to become stale really fast. I feel as if almost everything is around the power level of a Stormfront Pegasus. Everything serves a similar role and there isn't the feeling of building a deck that has its own inherent strategy as you do in real core sets. Many of the individual cards are kind of fun but there isn't the depth that makes you want to play it over and over yet - at least that's the feeling I have now.

The paradigm of differentiating the roles of creatures within the same mana range might not be an iron rule, but it allows different decks to have a different early game. It's my opinion that it's just plain better to do so, allowing people to draft decks deliberately.

While I think it's ok to deliberately change things for a new plan, in our case we just weren't planning slots for "on-curve creature" vs "not on-curve creatures." Now that we're in the Development stage analyzing this, I would like to call on the design and development people in this project to come up with development-designs for common creatures of any converted mana cost that don't have on-curve power.

I would also like to call for reprint suggestions for creatures that aren't on-curve, because looking through examples of these creatures help us think about the effects that these creatures could possibly have whenever we produce our own designs. We also need reprints anyways.

At the same time, it might be fun to leave a portion of the cards with high power. It would be an extremely warped environment where it's right to keep all three power-2 2-drops in Red, but even in a core set it could make sense to leave three power-2 2-drops in White so that it behaves effectively as a weenie color. (Although personally, I don't particularly see a point in making one of the good attacking 2-drops a tapper since I believe cards should be as different from each other as possible rather than be cool for doing everything).

One hypothetical paradigm would be to have a higher number of efficient low-drops than usual, and to also counterbalance that with a higher number of defensive creatures like Vine Trellis or a 3cc 2/3 that are good at bouncing off these 2/2s and 2/1s. Core sets have been avoiding this so far in order not to stagnate the board, but it might be ok to have these defensive cards if there could also be many cards that help aggro creatures push past these defensive creatures. Some examples are Goblin Shortcutter, Vulshok Heartstoker, or some kind of Invoker. It's important that the effects that help aggro creatures to continue attacking do not require the investment of a card like Panic Attack, because one-shot cards like Panic Attack would dilute a deck's threat density when aggro decks need a lot of redundancy. If such an effect can be devised, these attack-enabling effects would be much more useful to aggro decks than they are to control decks. Aggro decks can grab those blocker-negating effects while control decks go for value, hopefully creating a divide between fast decks and slower decks.

Full Zendikar block draft (Zendikar-Zendikar-Rise of Eldrazi drafts) was available on MTGO for a limited time, and was actually quite intriguing despite not having been developed to play together, because there were tools for both aggro decks and slower fatty decks. It seems draft cubes often feature a contrast between aggro decks vs. slower card advantage decks that use signets etc. to ramp into late game bombs. It would be interesting to do this except with a more core set feel. Having a slightly higher number of aggressive 2-drops might actually help in doing so.

Bond isn't actually doing anything in the set??
 

As I mentioned earlier, I've been running a community draft of the set, and being the moderator, I've had the opportunity to watch each player's picks. (The community collectively plays one seat in the draft, but there are live drafters behind the scenes for the other seats.) Although we are only about to reach the end of the first pack, it's been an extremely interesting experience.

The drafters each have their characteristic drafting styles. I've seen many of the drafters pick a card that I disagree with at first, but then they follow it up with similar picks and as the picks progress I can see a legitimate deck strategy forming. There is a player who picked Blood in the Water highly despite it being a weak indirect form of removal, and then followed it up with more and more removal, many of them good ones like Chandra's Outrage.

I've definitely experienced situations in Limited where the more removal you get (even the bad ones, such as Weakness), the more they cover for each other's holes (since they're each good at removing different things), and the weak removal spells allow you to save up the better removal like Doom Blade for the biggest creatures, so that their overall value grows as a whole. He even managed to secure a late pick Mind Rot and a Wyeth's Bounty which are not amazing on their own but in such a removal heavy deck will help attain card advantage so that the player may actually be able to remove everything the opponent plays.

Another player picked some high-drops in the beginning, then followed it with a high number of mana Elves and land-fetching creatures for a solid strategy.

Another player picked flyer after flyer, which would produce a steady stream of damage. If he could either pick some tempo tricks (unfortunately Frost Breath is currently not in the set) or some blockers, that would form a very scary strategy.

It's heartening that this is happening in draft, even if Sealed feels like it might turn monotonous after repeated sealed matches. It means that it's possible to draft different styles of decks. (Although I don't feel that it's nearly as archetypeful as it should be.)

That's the good part about the draft. A more concerning part is that the Bond creatures, the highlight of our set, don't seem to be creating any interesting dynamics in the draft. In the first pack, a Windborne Pteron wheeled around the seats. The people either had something better to take, or they weren't in the exact color pair of WU and didn't want to splash.

It's still unclear what the Bond mechanic provides for the set. The original point of the mechanic was that it could tie in with Ravnica dual lands that we guessed would come after M13. Then, some of the appeal first mentioned for the mechanic was that Kird Ape was a cool card, and it feels good when your card is receiving a bonus, especially the turn you play the required land and it "turns on."

As we examined the strengths of this mechanic, we theorized a style where people splash colors lightly, in a way that is different from the all-out multicolor of Ravnica or Shards of Alara. It's easier on beginners who don't know how to build a complex mana base. It's like "multicolor lite" where the multicolor costs are optional and less punishing if you don't get it, rather than having the card stuck in your hand when you're color screwed and can't pay a cost like "2WU."  When people are using splashed lands to turn on Bond, the moment you draw the splash land is more dramatic compared to casting the Bond creature when you already have the requisite land. We wanted to encourage splashing colors so that those dramatic moments would happen more often. To encourage merely splashing colors rather than heavy 3-color play, we decided there would be plenty of cards to incentivise commitment to your main colors, and there would only be a normal level of mana fixing.

One problem was that while the Bond creatures were arguably very likable, they didn't affect the way you play the game at all, unlike something like Bloodthirst.

We figured that at the very least, it could create interesting dynamics in draft and also make the act of building a sealed deck less than trivial. We also considered the possibility of a deck type that "builds around" Bond by combining land-fixing spells (Green mana fixing and Blue mirage effects) with every Bond creature you manage to pick up. We also tried to create as many cards that affect lands as possible without making unfun land destruction effects.

To make Bond creatures create interesting dynamics in draft, we decided that their costs and sizes would be such that they are decent without turning on, but better than what is usually allowed for their cost once they turn on, so that you could pick them up first and then gradually figure out which ones are your splash colors as you see more picks.

Currently, this hasn't happened. I find playing Sealed with the current set to be an exercise in identifying the color with the highest concentration of power cards (such as the Green cards below), then finding the color that forms the best curve with it. Since I often get many power cards in every color, there's great freedom in picking a color, and the Bond creatures are often not a big factor in choosing my colors. Sometimes there are potentially splashable Bond creatures but I don't want to for the sake of consistency.

And as I mentioned, people in the draft that I've hosted so far haven't been incentivised to splash colors. (Although this is only one draft of course).

This set really wants one more colorless mana fixer such as Terramorphic Expanse to facilitate a little more splashing. We've been avoiding it since it can create surprise interaction with Bond at instant speed, which might be a cool interaction in an expert set but possibly too chessy for an introductory set. But perhaps we should say what the heck and do it anyways. (However, it may mess up the common card count if we do both Terramorphic Expanse and Expedition Map.)

But mostly, I blame the power level of the Uncommons and some of the Commons. I have a feeling that they might be making the Bond creatures not feel special enough to splash.

Another problem that the high power level is causing is that we can't tell if the Bond creatures would cause a shortage of creatures for players who aren't in the two colors of that creature. For example, let's say a Kird Ape in a pack is unattractive to players who aren't in the exact color combination of RG. For those players, it will be as if there's one less common creature in the pack to choose from for their decks. Each color has 2 common creatures with Bond, each with their color constraints.. Depending on how easy it is to splash colors, and depending on how willing players are to play these creatures without activating Bond, there may be a shortage of creatures during drafts. This needs to be accounted for and calibrated somehow, but currently the overall high number of playable cards might be masking this problem so that this shortage is not felt.

This is going to be hard to playtest!

I imagine there to be at least two very distinct types of testing: testing to explore many possibilities, vs. testing to fine-tune the set and make everything mesh. To compare to a painting, it's as different as trying out many different sketches vs. putting on the oil layers.

The test I mentioned is about testing how Bond really affects Limited if the power distribution of the surrounding cards were balanced, and fine-tuning the non-Bond cards we designed is not yet an issue.

I imagine that with most Magic sets, the early exploratory tests for a set would only need to establish whether a particular kind of play pattern or a particular mechanic is fun and actually works. For example, it might be about finding out whether pairing creatures with Soulbond is fun or not and what the best implementation is. When testing to figure out the right implementation for Soulbond, you don't need to get some Uncommon Air Elemental variant to the right power level in order to find out things such as "it's best when you are allowed to re-pair pairs."

In this set, though, we have a peculiar situation in that our mechanic that doesn't affect play much, but affects the way you draft or build your deck. The power level of the Bond cards in relation to the other cards in the set is very important for testing what kind of effect it has on Limited. We actually have to get the power levels of the surrounding cards mostly right.

To solve the power level problem, the next few tests could use M13 Bond-related cards in a mix of M11 cards. Perhaps we could generate 6 M13 Boosters and 6 M11 Boosters and only use the Commons and Rares from the M13 Boosters and only the Uncommons from the M11 Boosters. We could allow any land-related cards from the M13 Boosters to also be used.

Another reason why this mechanic is difficult to develop is because while strategically our Bond creatures only provide about as much game play decisions as French vanilla creatures do, in terms of complexity, the Bond creatures have long text and they can't be counted as French vanilla creatures. (Although their status seems to be easy enough to keep track of.) Each color will have 2 vanillas, around 2 French vanillas, and 2 Bond creatures, which makes it very vanilla-heavy in terms of strategy. Although French vanillas like Garruk's Companion are not a blank in terms of strategic significance and can definitely contribute towards the formation of an archetype, it would still be difficult to craft many different playstyles in the set if there is less space for creatures with original text. Perhaps this can be solved by devoting a higher number of common slots to creatures than usual. The Bond creatures call out for specific color combinations, and in sets where cards require to be in a certain archetype to be good, booster packs are allowed to be more dense with playables.

So those are my thoughts after my initial testing. I hope to be back with more as testing progresses.