3.1 Total gas - a theory of deck sizes

Note: this article was originally written to be published on Stimhack, but since I haven’t heard back from the editor in a long time, I am posting it here. The prose is a bit different from the rest of the website, because the context and target were intended to be a little different. Also, this was written before EA Sports significantly innovated deckbuilding space with their 40-card Reeducation Azmari list.

Very early in the process of learning Netrunner, a player will ask others how many cards they should have in their deck, and they will get a very strict answer. More experienced players will collectively tell them that their deck should have “exactly N” cards.

Actually, that’s not how the story goes; let me tell it again. Very early in the process of learning Netrunner, a player will post a deck online, and others will jump at them saying that they’re playing the wrong number of cards, with a very clear idea of what that number should actually be, and a very clear idea about why. The beginner will sometimes tries to argue for their choices, but then will likely be persuaded by the very reasonable arguments the other players offer.

Yet, these same players that are very eager to give these suggestions will often end up puzzled at some tournament successful decks: I’m master thief!!, Big MaxX, World Tree Wu, Guantanamo Bay Bay, ProsperAri, mulch, Venti Matcha Latte, The Worlds Grind, and so on. These decks all violate common understandings of how big a deck should be, and all of them have good reason to do so. Who’s right here? How many cards should you put in your deck?

The classical stance#

There is a commonly held stance, especially among people who have been playing Netrunner for some time. If you ask experienced players, the advice you will usually get is that in Runner decks, your deck size should be exactly the same as your minimum deck size, and in Corp decks, it should be exactly that number plus four.

The arguments in favor of the classical stance are generally pretty simple, and quite persuasive. However, there is one major problem with them that will come back to bite us as we review the opposite side: these arguments are borrowed from other card games, most importantly Magic. In fact, if you want a particularly well-developed and mathematically detailed exposition of these arguments, you should probably go and read Patrick Chapin’s Next Level Deckbuilding. Carrying knowledge from one game to another is very common, and often correct, but there are cases where a heuristic is based on some specific dynamics within a game, and it’s easy to not notice that some important things are different in a different game. In a lot of card games, arguments developed within Magic work well; in others, they work terribly.

How well do they work in Netrunner? Before we look at the cases where they break, let’s review the positive arguments for the classical stance:

Card quality#

Netrunner has good cards and bad cards. If you build an HB deck, you will likely start by putting Luminal Transubstantiation (a very good card) in it, and end up with zero copies of Nerine 1.0 (a very bad card). But what about everything in between?

We could conceptualize card quality as existing on a spectrum, where every card sits in some position between the staple and the unplayable. So when we look at, say, a 45-card Criminal deck like cableCarnage’s sableCarnage, we may point at different cards and say that they are the best card in the deck or the worst card in the deck. Of course, the worst card in the deck is still pretty good, or cable wouldn’t have put it in his deck. So the average card in sableCarnage is somewhere between really, really good and pretty good. Say that we give the best card in the deck a rating of 10/10 and the worst card a 7/10; we put it all together, we do the math and we can say the average card in sableCarnage is a 8.2 out of 10.

Now suppose a new set comes out, and we want to add Alarm Clock to the deck because we think it would be good in there: we think it would be at least an 8/10, way better than the worst card. We add it, making the deck a 46-card deck, and then compute the current average, and it turns out to be… 8.15/10! Well of course, we forgot to take the worst card out. If we remove the worst card (which as we said, is a 7/10), then the average increases to 8.25/10. We did it, we improved the deck!

The argument about card quality goes as follows: cards in a deck can be ranked from best to worse (and that ranking is a hard and grueling process); in a deck that is bigger than the minimum deck size, if you find the worst card in the deck and cut it (which is a hard and grueling process), you will end up with a better deck, because the average card you draw will be better.

All good for Runners; but this doesn’t work exactly the same way for Corp decks. One problem when building Corp is that you don’t only get to choose which cards you draw from your deck but also which cards your opponents draws from it, and among them there have to be agendas. While they can be strong, Agendas tend to be cards you’re just not super stoked to put into your list in big amounts, because while they can be good for you to draw, they are very good for the Runner to draw, and the overall outcome tends to be negative. However, you have to include agendas, because of how deck construction rules work.

So we have cards that tend to decrease average card quality in the deck and we’re forced to add them. How many of them? Well, we have to add more of them for every 5 cards we put into our deck. So we cheat a little bit. Since any good non-agenda card we add improves the average card draw in our deck by being better than agendas, we add as many of those as we can without having to add extra agendas. That translates to playing your minimum deck size, plus four. [footnote 1]

Synergies#

“But cards aren’t just good or bad!”, I hear you say. “Cards shouldn’t be given a rating in a vacuum, because how good they are depends on synergies in your deck: they work well or badly, together”. Well, there can definitely be such a thing as better or worse cards, but you are also right: a lot of decks are trying to get something specific going: they want to take cards that work together and play it in the same game.

But this is also an argument in favor of minimum deck size. Let’s look at sableCarnage again. This deck plays two copies of The Twinning, a very powerful card that does nothing on its own. In order to do anything, it asks you to have Paladin Poemu, Mystic Maemi, Cezve or Miss Bones in play. There are unfortunate cases where you draw The Twinning, but none of its enablers show up until late, so it remains a dead card; you want to avoid these cases as much as possible. Obviously, the more unrelated cards you keep adding, the more this synergy package suffers: you become more and more likely to draw only one half of the engine without the other half. That’s a good reason to stay at 45; and of course, it’s just as valid if we’re talking about Reeducation + Neurospike, Boomerang + Diversion of Funds, Coalescence + Aesop’s Pawnshop, Loki + Trieste Model Bioroids, Pharos + Tree Line, and so on.

Cards have different times at which they want to be played. Some are strong at all times, some are best saved for last. Some cards are especially powerful if you play them very early, ideally on the first turn. Lago Paranoa Shelter, Tranquility Home Grid, La Costa Grid, Patchwork are examples: these cards provide you with extra value pretty much forever, so every turn you wait before playing them is a turn you’re not getting value. If your deck relies on cards like these for its economy to click, they are, so to speak, one-card engines, and you should try and maximize your chance to play them as early as possible. Obviously, the bigger your deck is, the harder it will be to make this happen.

Self-balancing draw#

This is a classic argument from Magic, where it works very well. In Magic, you don’t just have good cards and bad cards: you have cards that you want to be drawing in certain ratios. In particular, Magic’s mana system makes it so that you want to draw a certain ratio of spells to lands. Depending on your deck, you may want 40% of the cards you draw to be lands, and 60% of them to be spells, and so you build your deck to make this likely to happen. [footnote 2] Of course, disaster can strike: you can get mana flooded (drawing too many lands) or mana screwed (drawing too few) - and both situations can be very bad.

But luckily, things are not as bad as they seem. If you have drawn too many lands, there are fewer lands in your deck to draw from, so you will be more likely to draw more spells going forward, and vice versa. Generally speaking, the more you have drawn a certain type of cards, the more likely you are to draw a different type of card. This is true of every game where you draw cards from a deck, and we can call it the natural ability of card draw to be self-balancing, or self-correcting. In Magic this is very good, because it ensures that your draw sort of balances towards that ratio we talked about earlier: you want it to be average.

An important property of draw being self-balancing is that the higher percentage of your deck you have drawn, the stronger this effect is. As you draw through 25%, 50%, 75% of your deck, the odds of drawing different things from what you have drawn so far keep increasing. Obviously, the smaller your deck is, the easier it is to see a higher percentage of your deck quickly, and the stronger self-balancing gets. Since self-balancing is good in MtG, this is a pretty good argument towards smaller decks.

Is self-balancing positive in Netrunner? Yes, especially for Corps! In any scoring deck, agendas are the typical kind of card you want to see an average amount of. If you’re agenda flooded, your HQ clogs up with points you can’t score, and little money or ice to accomplish anything. If you’re agenda screwed, you become unable to convert a superior board into actual points, and you give the Runner time to establish inevitability. A similar thing can be said about ice: with none of it you are helpless and vulnerable, but excess ice can also end up being useless, since it costs too much to install and you don’t really get much out of gear-checking the Runner multiple times.

Self-balancing helps you not to stay in either flood or screw for too long: whenever you draw cards, you are a little more likely to draw things you haven’t seen yet. That is ice and money if you’ve seen too many points, and points if you’ve had more ice than you can actually use. As we’ve said, this effect is stronger at lower deck sizes. If you’ve played Rush PD lists and also Asa lists with a similar gameplan, you may have found out that Asa has a bit more of a tendency towards awkward draws, and gets flooded a fair bit more often. PD’s lower deck size is part of the reason why it sort of feels like you just draw the cards you need.

Everyone else is lazy and sinful#

As we’ve said above, cutting the 46th card from your Runner deck is a hard and grueling process. For a lot of people, there is obviously a lot of aversion at the idea of going through this: deckbuilding is supposed to be about expression and creativity, and not about the painstaking ordeal of optimizing.

The conclusion is obvious: everyone who claims to have good reasons to play more cards in their deck is actually just being lazy and the justifications they are providing for their choices are post-hoc rationalizations.

Expressed this way, this is obviously silly. Enough “big” decks have succeeded in important tournaments - with multiple of them winning Worlds - that there has to be something to it. However, there is something to this argument when it is taken with a grain of salt. Players at all skill levels have a hard time when faced with the challenging prospect of having to find the cut, and playing 46 cards “just because” is very tempting. It takes a lot of games to figure out that 45 is actually smoother, anyway. So even though there are a lot of reasons why a bigger deck may be better, sometimes it can be reasonable to assume that someone who is not able to provide those reasons may just be feeding their aversion - especially yourself.

Something that you will hear pretty often is that someone is playing a 46 card deck, because “staying close to minimum is important, but there isn’t much of a difference between 45 and 46”. This article - and generally most of the things I say when thinking about Netrunner - is concerned with the idea of trying to do what’s best. The question isn’t whether the difference is large, but what direction it’s in - if a 46 deck is just a little bit worse, then why are we playing the slightly worse deck when we could easily be playing the slightly better one?

Alternative arguments, good and bad#

Oh, but there are many valid reasons to play a bigger deck. Netrunner boomers don’t always believe them, even when faced with a mounting pile of evidence. As great philosophers have pointed out when talking about topics only tangentially related to Netrunner, a widespread and deeply rooted way of thinking will almost always provide ways to explain away evidence that can falsify it - for example, “this deck is misbuilt and unoptimized, it only did well at Worlds because the players who piloted it are good players” is as common a reaction to strange new brews as it is insulting to those who come up with them. Many still consider Big MaxX’s success at Worlds 2021 to be some sort of aberration that teaches us little about “normal” Netrunner. It’s clear that some will never be convinced by more evidence: we need arguments that sound as persuasive as the ones above; we need a model, a theory.

I believe that some of the arguments that I’ve seen people propose for bigger decks look reasonable on the surface, but don’t actually hold up in practice; that some don’t really make sense at all; that a lot of them are actually sound, though they may require specific conditions to appear in the meta. I’ll just discuss one that I disagree with, and then try and dedicate a lot of space to what I think are actually good arguments.

Concerto replaces itself so it doesn’t count#

(this section is affectionately dedicated to my friends from testing group Muntal Bost. I’m not sure how serious of a case for 48 card Sable they were trying to make, but some people do believe in this idea or parts of it, so I feel obligated to try and dispel it)

You could have a 45 card Sable deck. But what if we added three copies of Concerto to it? On face value it would be a 48 card deck - likely a worse one. But when we play the Concerto, we are making a run, which is something we would want to do anyway, and drawing a card from the Concerto itself. The Concerto has “replaced itself” so seamlessly that it barely took a card draw. It didn’t cram our deck, it just smoothly became something else, so easily that it should barely count as a card. Our 48 card deck actually has the consistency of a 45 card deck.

I believe that many things are wrong with this line of thinking, and not all of them obvious. But the first thing to point out is that the phrasing “replaces itself” is another relic of Magic terminology that carries connotations that similar effects don’t deserve to have in Netrunner. Magic, and games like it, have a completely different resource system, where among your most important fundamental resources you have cards in your hand, and something you spend to play them (mana). But you only get to draw very few cards! So a very desirable property that a card can have is that in addition to doing its thing in exchange for mana, it then draws you a card, replacing itself. In a game where you get to draw few cards (less than half of your deck in most games), this feels like a “free draw”, and players learn to respect this effect as deceptively powerful, because it can often help you achieve “card advantage”.

Needless to say, Netrunner is completely different. You have cards, a resource to play them (credits), and a resource that interacts with both by being spent to play cards, get more credits, and get more cards. The result is a game where card draw is useful but also plentiful, where Diesel isn’t completely game-breaking and it is sometimes correct to not include it, where “card advantage” as a heuristic fails completely except in a few uncommon situations. If the goals we want to achieve are so different, why should we copy Magic players' terminology for something like Ponder or Electrolyze and apply it to Concerto and Jailbreak?

Let’s assume you take a functioning 45 card Sable deck and add three copies of Concerto to it, and think about what can happen as a result. You spend a click to draw a card, and it’s a Concerto. You could have drawn a different card - in fact, if the Concerto was not in the deck, you would have drawn Card X just below it, which now sits on the top of your stack.

Now you could spend a click to draw Card X. If you do that, Concerto has been dead weight: its effect has been to make you click twice to get to Card X. But what you are more likely to do is to play the Concerto, making a run with maybe a few extra credits to use during it and drawing you Card X. You have still clicked twice to draw Card X, but one of those clicks has been spent to make a run. If you were already planning to run before drawing, then this is basically no cost, and that’s the situation where it feels like there’s something “free” about the Concerto: you are packaging two things you wanted to do already into a single click. However, there are so many situations where this is not the case. It happens so often that you want to draw, then run: for example, you know you could draw into a Dirty Laundry or a Miss Bones, so you try and draw into them before you run. With Concerto, you are forced to make the run without being able to alter your sequencing depending on what the card you’re drawing is.

Just consider this situation: you are against PD, they are pushing a key agenda in the remote, you need to run in your first two clicks because that’s the only way you can get past Skunkworks, and you need Boomerang to actually get past that Gatekeeper. With your first click, you draw, hoping for the Boomerang. You find a Concerto, which you play, and it draws you a Boomerang. You have the card, but now you have spent two of your four clicks, and it’s too late to get past the Skunkworks. Yes, as Sable you can get a click back if you can run the mark successfully, if there is no Border Control there, if you have your breakers out already, if, if… In the average case, did Concerto really “turn” into another card at no cost?

There are more cases where this breaks down. An obvious one is that there are situations where making any run would be damaging, like when there is a rezzed Tributary that you can’t break. In this case, you don’t want to play the Concerto, so it sits in your hand and actually ends up being dead weight.

One more: the optimist view of Concerto sees it as “not really counting as a card, because it turns into another one: it just is another card with a positive effect stapled onto it”. But the question is what card is it? Sable invariably plays The Class Act, which often asks you to pick between a card to draw and one to bottom. You’re supposed to take the one that helps you the most right now, but if Concerto is a mystery card with a little extra on top, suddenly that choice becomes a lot more difficult to make, and TCA’s power is diminished. You should easily be able to see that a similar thing applies to keep/mulligan decisions, making every starting hand with Concerto in it much more difficult to evaluate.

The idea that Concerto enables 48 card decks that play like 45 is definitely wrong for the simple fact that if it was “free”, it would enable 45-card decks that play like 42, and their superior consistency would invalidate 48-card Sable. But even more important to understand is that drawing a card doesn’t make anything a “free” card, because card draw already came with your basic action card. The idea that these cards “improve consistency” is a leftover from habits developed elsewhere, and it simply doesn’t carry to this game.

Toolboxes#

koga’s I’m Master thief!! is and will forever remain a puzzling deck. It features a lot of very peculiar slots, making for a strange list that somehow dominated the tournament it was played in, but was so idiosyncratic and difficult to play that no one else picked it up and did well with it. It is also a 48 card deck, at a time when the received wisdom of “stay at minimum” was generally followed much more adamantly. koga’s only comment on this choice is “yes, 48 cards baby”.

Let’s try an provide an explanation of our own. Kabonesa Wu is a very strong identity with innate tutoring. Wu’s problem is that the programs you get tend to be short-lived, which makes most of them bad candidates for her ability. The classic solution is to play some SMCs, because if you Wu for the SMC, then crack it for the program you actually want, then the program stays, though you had to pay a small additional fee for that effect.

koga’s list leans on Wu’s ability much harder than that: it plays 3x Scavenge and 2x Rejig, both cards that “flicker” your tutored program so that Wu doesn’t remove it from the game at end of turn. This enables a toolbox deck, where you have ample ability to tutor many things depending on what the situation is. The deck plays Deus X, Imp, Misdirection - all programs that you may only need in specific matchups, but will save your life there - and then has the ability to always have them when needed.

The idea here is that if we had to force the deck down to 45 cards, we would have to cut some cards - and that would end in either cutting economy, making the deck too poor, or some of the less important tech programs. But you can have both, because the tutoring makes up for the reduced consistency you lose by making the deck bigger. Yes, you are slightly worse at finding Diversion of Funds in your opener, but you are sure you can have a reasonable amount of early game economy, very easy access to one or two Rezekis on turn one, and tech pieces for everything you may need later on. It’s not an obvious choice, of course, it’s a tradeoff, and a bold one - but I find it difficult to state that it is a clearly wrong choice.

A toolbox doesn’t necessarily have to come from tutoring. Just look at the Worlds 2023 mulch list, which plans to mill itself relatively quickly and then keep playing Labor Rights to get Simulchip back. If you can do that, Simulchip can fetch you anything you may need from the bin: money, multiaccess, Clot, a breaker to deal with anything the Corp might throw at you. Here too, making the deck a tiny bit bigger allows us to have a larger toolbox than we could otherwise! This means we can include cards that we install pretty rarely (like Revolver or Clot) but that can be completely game winning when we do, and will warp Corp play for the single fact that they are in your heap.

Did this make mulch less “consistent”? It depends on what you mean by that. mulch wasn’t very good at finding a specific piece of tech by turn two, or at having strong openings with a lot of money in them. However, it was good enough at contesting whatever rush decks like PD or Ob were doing, and then it could be 100% sure it would have whatever tech was needed by turn eight or nine. This is what mattered, because the threats that needed answering - Stavrun, Audacity, Skunkvoid - required that you had your answers by the mid-late game, not earlier. mulch only ended up struggling against fast asset decks that asked you to find Imp very quickly, such as kaNEHl. The takeaway here is that there is a difference between being consistent at having very strong openings and being consistent at drawing into your tech by turn eight; and some big decks that struggle with the former may well be able to do the latter, by finding pieces that you couldn’t possibly fit in 45.

Garnets#

Tutors allow you to fetch the card you need at any time, and that’s already an impressive thing. Sometimes, however, tutors go beyond that, giving you additional benefits on top of that. Let’s look at DeeR’s World Tree Wu from late 2022.

This deck is build around getting World Tree down early, selling Mayfly and other cheap programs to assemble your rig, and reach a complete inevitability state that Corps could do nothing about. Its main problem is that it is rather poor, because getting the Tree down costs a lot of money, and Shaper didn’t exactly shine for its great money options at the time.

Tree is very helpful in that regard, though: it saves you three credits on the card you’re tutoring. That means that assembling your rig costs a lot less: it’s a twelve credits discount on three breakers plus Turbine, which is insanely good. The deck absolutely needs this to function properly. But what happens when you draw one of your icebreakers? You have to install it from hand at full cost. A lot of the deck’s worst games involved you drawing two of your breakers, falling behind on tempo, and being unable to contest a fast corp.

It’s easy to look at the deck and think that the very large size - 65 cards! - is due to this being a toolbox deck that wants to have many options available, but really the most important reason is that in a smaller deck, you will draw your breakers, and that can lose you the game. “Drawing your cards more consistently” is actively bad when you don’t want to draw them, and if you want to make that happen less the best thing you can do is make your deck bigger.

I’ve been told that Yu-Gi-Oh! players use the term garnet for a specific kind of card: one that you want to find with a tutor effect but you don’t want to draw naturally. This requires that you have especially beneficial tutors, which don’t come often, but World Tree definitely provides a case for this.

Extremely low density#

We talked about how agendas are problematic cards to have in your deck, as much of a strength as they are a liability. In fact, sometimes they are entirely the latter, because decks may not want to score them, winning by other means instead while the Runner tries to find the few agendas that are in the deck.

A classic way to deal with this is to only play 3-point agendas. This ensures that you have to dedicate as few slots as possible to them, and also that stealing is kind of awkward for your opponent: a completely “flat” distribution of agenda points means they actually have to go for nine. The lowest agenda density you can achieve with “normal-sized” decks is 6/44 = 0.1363, which is great. You have very few agendas and your opponent has to steal half of them, while you try and do something else.

But could we go lower? Yes, and the only way to do it is by diluting your agendas in a bigger deck. You could play with a 8/59 = 0.1356 ratio, or even 10/74 = 0.1351. The King’s Guantanamo Bay Bay and sebastiank’s Venti Matcha Latte respectively are decks that do just that. The benefits are obvious: you are making it more difficult for the Runner to find your points while you push your non-agenda win conditions (Dr. Vientiane Keeling, or Mutually Assured Destruction + double End of The Line). What about the negatives?

The King’s deck can afford a drop in “consistency” because it plays such an annoyingly impenetrable ice suite that it will always buy enough time to find its win condition; it also plays some very powerful draw cards that can dig deep into R&D, such as 3x Simulation Reset. sebastiank’s deck has many redundant pieces that can substitute for each other (Mutually Assured Destruction only asks you to have many cards, after all: which ones they are is secondary) and has impressive tutoring capabilities, coming both from Ob’s ID ability searching for your favorite assets or upgrades and Pivot+Gaslight finding your kill pieces.

There’s another problem that we talked about, and that’s self-balancing draw getting weaker in bigger decks. The main consequence is that you will have fewer draws where you find an “average” amount of agendas, and more cases where you draw none and keep drawing none, or you draw many and keep drawing into more. What does this mean for these lists?

These decks aren’t particularly interested in drawing agendas in a “normal” quantity allowing you to push some in a scoring remote without having a flooded HQ: they never score anyway. So drawing no agendas is not really an issue. What about extreme flood, which is more likely for these decks than for 6/44 density decks? The answer is to have plenty of tools that can dispose of multiple 3-pointers at once. Attitude Adjustment and Drudge Work aren’t simply cards that manipulate agenda flow, they are answers to torrential floods: they are actually better in decks that flood hard when they do flood, because you get more money back for your trouble. They are perfect fits for huge lists.

Sometimes deckbuilding rules force you to make strange choices that don’t have that much to do with density. For example, Sokka’s The Worlds Grind really had to play Above The Law, something it couldn’t really do with the “recommended” 44 card size. The solution was to simply go to 49.

Total gas#

A few years ago, the Standard format was completely defined by Rezeki. The card was a huge deal: it meant that there was no way to starve and grind Runners and run them out of money. The game feels very different now: Runners can have a lot of money in their deck, but the well can run dry after a long game. Longevity is a potential issues, and there are often trade-offs to be made between it and speed.

Just look at two Reg Hoshiko lists: Sokka’s You do always come back from Worlds 2023 and Nemamiah’s 2018 called, it wants its Val deck back from UK Nationals in the same year. The decks are really similar on a surface level: they play the same console, breakers, wincons, and even tech. But they make significantly different choices when it comes to economy: Sokka’s deck burns fast by drawing quickly with Diesel and getting quick profits with Moshing and Strike Fund, while Nemamiah’s goes a little slower but makes immense amounts of money with Liberated Account. The results are respectively a deck that can quickly find its answers to rush strategies but can end up with a short stack and no money against yellow taggy decks, and one that can be just a little too slow for PD but will always be able to contest R+.

Both choices were correct for their specific environment: Sokka had to deal with a field full of PD and Rush Ob, while Nemamiah competed in an event that is traditionally stuffed with yellow-blooded spammers. There’s an obvious interpretation here: some Corps require you to have fast money, whereas others mostly care about how much money is in your deck overall, giving less care to how fast you can accrue it. That is to say, in some metas, a really important property of your deck is total gas: the amount of money you can make given a very long time. Decks that check you on that metric tend to be grindy taggy asset spam, hyper-glaciers, or asset-based prisony archetypes. This is a fundamental difference between Magic and Netrunner, and a reason why classical arguments only go a certain way: Magic decks aren’t designed for games where you’ll draw all of your cards. You will draw 15-25 cards from your deck in most games, and you need to make sure that that’s a very good sample of it. In Netrunner, you will see most of your deck in the vast majority of games. This turns a lot of preconceptions on their head.

At first glance though, it’s not obvious what this has to do with deck size. After all, in the example we just described, removing cards to make space for others was enough to add a lot of total money to the deck. But there are cases where that isn’t enough.

An obvious case is when you purposefully put cards in your deck that cut into your long-term economy. KyraWNY’s Less is Mawrie is a relatively fast setup deck that has usually built an inevitable rig by turn five, something you probably wouldn’t guess about a 55-card behemoth. The main reason is that it plays three copies of both The Price and Gachapon, which are able to install your key pieces at close to no cost. But both cards give you speed while taking away your longevity. Experiments in 45-card variants of the deck ended up fruitless because you would have no money after setting up a rig. The best solution was to add money to the deck just by adding more money cards to it.

Sometimes it is Corps that ask for Runner cards. In the late Parhelion meta, PE asset decks were a force to be reckoned with in a meta where Corp options were pretty varied. It is intuitively very obvious, although contrary to received wisdom, that a very large Runner deck would be a good way to not run out of cards: that’s what Council’s Uno Mas Hoshiko did to great effect. The deck could play in such a way that PE could never mill it alive, and against other opponents it would exploit powerful cards that run the risk of making you run out of gas quickly such as Zer0, Bankhar and Moshing - again, by simply never running out of cards.

An important thing to notice about this deck is that its “consistency” was still pretty good because a lot of cards could sub in for each other to do each other’s job: Bravado is Raindrops Cut Stone, Earthrise Hotel is Dreamnet is Zer0, Botulus and Bankhar are any icebreaker, etc. There is such a high volume of strong cards with comparable effects that the drop in “card quality” as the deck gets bigger is barely palpable, and they all work on their own so well that there’s no need to assemble a specific synergy package. This was part of the strength of Big MaxX, where there was so much generic economy that you could be guaranteed to mill any money card just to draw another one.

Running bigger is not the only way to increase your longevity: Nemamiah’s route of just playing “bigger cards” is one we’ve seen, but some recursion cards also do something similar. Since Rebellion without Rehearsal dropped, players traditionally very dedicated to Mawrie have switched to 45-card decks playing a combination of Ashen Epilogue and Labor Rights, such as Jai’s Technikko. The idea there is that Epilogue takes some tempo to play but ultimately adds an incredible amount of money back to the deck: it’s in some ways like playing a bigger deck. Similarly, mulchy Loup decks such as DoomRat’s Buffet of Chips often avoided mulch’s big decksize, which was partly necessary to be able to feed Audrey v2 for a very long time, choosing to obtain that longevity by playing Buffer Drive instead. This was unavoidable in a deck that really needs Imp in its opener in order to work.

Putting it all together#

So, how many cards should I put in my deck? The answer doesn’t feel obvious to me. I believe that the classical stance’s arguments, while based on theory about games that work a little different, are still mostly sound. Most decks definitely benefit from staying at +0/+4 cards. But I don’t think we can state that that’s always or even almost always correct anymore.

The most important thing to realize is that the benefits of sticking to the +0/+4 rule aren’t absolute: high card quality, consistency at deploying engines, self-balancing draw aren’t necessarily the things your deck wants the most, nor things your opponent is challenging you to have. There are alternative qualities that you can achieve by making different choices, and what you will prefer will depend on a variety of factors.

We can try and summarize these factors in a nice little table:

There is no royal road to good deckbuilding: it’s all judgment calls, guesswork and experimentation. Netrunner wouldn’t be such a good game if there were easy answers. Learn when to be big and inevitable, when to be small and efficient, and all the ways in between them. ABR!

(1) There are legendary tales of Corp decks that really want to draw into their agendas because they want to rush out super quickly, or they need to execute some sort of combo that requires maximum consistency, and so what’s best for them is to play 46 or 40 or some other strange number of cards that’s smaller than minimum plus four. The idea is exciting and definitely underexplored, but despite recent developments such as OF15-15’s nyoooooom, there is still little concrete proof that there is something to this idea.

(2) There’s one more trick to it: the ratio with which you want to draw card can actually change during the course of a game. Specifically, you want a bunch of lands in your opener and then fewer lands later on, so you play fetchlands to thin your deck as early as possible and have different ratios at different points of the game. A similar dynamic is a subtle part of what makes Rush Ob decks consistently powerful: they can draw openers like a 17 ice deck (which is great), tutor some ice, and then draw like a 13 ice deck in the late game (also great).