General The size of one's cube

[OCPunisher]

Hello friends,

I'm wondering about how one goes about deciding the size of a cube.
- What makes you choose to include 450 cards, for example? Why not 360? Why not 720?
- Did you decide based on personal preference, or the size of the box you bought, or the number of good cards you owned, or something else entirely?

Personally, I settled on 450 because it meant we could draft with up to 10 people, which was likely going to be the most we could come up with at any one time. However, I've recently considered increasing the size by a fair margin to include more possibilities at various parts of the curve.
- For example, I currently run some pretty standard red four-drops: Avalanche Riders, Flametongue Kavu, Hellrider, Chandra, Pyromaster, Koth of the Hammer, and Sneak Attack.
- If I increased my cube to 720 cards, I could run things like Goblin Ruinblaster, Hero of Oxid Ridge, Keldon Champion, Chandra, the Firebrand, etc.
- That would allow people to still get their requisite number of four-drops, but not necessarily run the same stuff every time.

Is this a reasonable idea, or does it have some very ill-effect on a normal 8-10 person draft that I'm completely overlooking?

 

[Jonas]

I was at 450, but went back down to 360 again since I almost never had 8 people drafting, and just seeing half the cube got a bit boring.

 

[FlowerSunRain]

Functionally and ideally for regular drafting there are two cube sizes: perfect size, where the you draft every single card (usually 360, but that can vary) and oversized. The advantage of a perfect sized cube is pretty clear: the draft pool is always the same, so you can balance around it in a predictable manner. Oversized cubes give you better variety, but in exchange you have to account for dilution. Whether you are oversized by 50%, 100%, or 300% is just really dialing how much variety you want in exchange for how dilution you are willing to account for.

In practice, I think most people don't create separate cubes for separate player counts and perfect size cubes sometimes turn into oversized cubes when not enough people show up and force their cube to work in lots of different formats. It makes sense considering how expensive and time consuming cube construction can be, but its not really good game design. Its not limited to cube designers, though, board game designers force their games to be "playable" with a broad range of player counts because it sells more copies even though most board games play their best at one or two specific player counts (and are sometimes barely recognizable or intolerable at some of the "supported" player counts).

I chose my cube size because it was the best compromise between getting in the cards I wanted to play and having enough variance control to keep them playable. My number is not set in stone and has varied from 360-600 as my needs have changed.

 

[Jason Waddell]

I am out now, but from my phone... big cubes may have diversity in cards, superficially, but not a diversity in functional archetypes. Bigger cubes don't give you more design space.

 

[FlowerSunRain]

Adding more cards does provide more design space, but as you can increase the number of cards in your cube, the amount of design space that each card can add gets smaller while the negative effects of dilution get larger. The larger the effect of dilution, the more redundancy you need to account for it and redundancy eats away design space. I'm not sure at what point adding more cards actually starts taking away design space, but its pretty fast.

The only way you can really add more design space is by adding more PLAYERS, because that actually puts more cards and more decks into the draft. A 540 cube probably only has marginally more design space then a 360 cube (and is much less reliable at producing functional decks in its supported areas), but a 540 cube drafted by 12 players has much more potential design space then either of them.

Anyone who has tried to make a 90 or 180 card cube will probably be able to vouch for this, though there certainly is a ceiling on the number of players you can reasonably add.

 

[Eric Chan]

I think what you'll find is that at Riptide Lab, we have a very strong bias towards small cubes - mostly 500 and under - for many of the reasons outlined by Jason and FlowerSunRain above. It's mostly that you get to craft a tight, well-structured drafting environment for your players, and that you get to control all of the variables for any given draft. Increasing the size of your cube mostly serves to increase the variance of any individual draft, but doing that without some careful forethought might lead to unintended consequences. Just taking your red four drops as an example, if you were to run an eight-man draft, you might see six of them one night in the draft pool, all ten of them the next, and then only one the third time. Is this the kind of variance you're looking to promote? Would you be ok with all of those results?

If all you want to do is play some sweet cards like Keldon Champion, though, then I'd say to just play those sweet cards! Swap out that Flametongue Kavu for a draft or two and see if it tickles your group's fancy. In this way, you retain tight control over the draft metagame - that Champion is guaranteed to show up during your next eight-man - and you can gauge its impact more definitively. Worst case, you simply decide you don't like what you're seeing, and you revert back to the old card. To my mind, it's a better, cleaner way of increasing the variety of cards your drafters see while keeping other factors in your environment stable.

 

[FlowerSunRain]

I guess I should add (as someone who runs a 500 card cube) that while having more cards doesn't add any design space it is a nice "shorthand" way of swapping cards each draft. Manually picking and swapping is better design, but it is time consuming. The downside is that you get poor combinations. Like, lets way you run 4 cards in particular niche at 360 and 6 at 540. There is a non-negligible chance of getting 2 or 6 in your draft pool in any given 8 man, which is probably not desirable.

Personally I and may players like a looser draft experience created by variably supported but highly flexible archetypes and enjoys a huge variety of cards, so it works fine for us, but if you are looking for strong archetypes and tight decks a cube as small is possible is just, designwise, your best option.

 

[CML]

360 is probably ideal, though with money-drafts being almost always six people, I wonder if 270 is a but I know (aside from the rare practical difficulties that arise from only supporting 8 people) I'd be losing some of my design space by paring down from 450 to 360, there simply wouldn't be room for some beloved themes or some beloved spells.

The best analogy here is to the writing process, where endless revisions must both add some stuff you forgot and debride extraneous text that wasn't as good, and (if I may put on my Kristallnacht hat) "perfection is achieved not when there's nothing to add but nothing to take away." Obviously perfection is more of an asymptote / ideal than it is a reality, Cubes, like other creative projects of nontrivial complexity, are never completed, only abandoned. As much as I'm annoyed by mantras like "to add a card you must cut a card" or "restrictions breed creativity" but it's true that these thought processes lead to far richer design experiences, for drafters and pilots too, than "Eh f it I like it. Throw it in. Why are all these themes not working?"

 

[anotak]

to me the ideal number is perfect fit, it sucks when people want to hope for a specific card later in the draft and its just literally not in the draft

 

[FlowerSunRain]

to me the ideal number is perfect fit, it sucks when people want to hope for a specific card later in the draft and its just literally not in the draft

A sentiment very similar to this is why I moved all the land out of the main draft. The small number of specific 2 color fixing meant that the variance on these vital cards was much too high and it was upsetting to be in a color pair and see 0-1 lands in those colors.

 

[Jason Waddell]

Creating archetype diversity is mainly a function of:
1) compact anchors that can pull a deck in a unique directions
2) wide overlapping support that can be combined in lots of unique ways

This is more an issue of card selection, and not size, but inherently, there are a few main changes that, say, going to 540 would give me:
1) the ability to support more than 8 players (this never happens)
2) increased granularity in card choices. Currently I increment by 1 card at a time, but in 540 each card is basically 2/3rds of a card
3) added variability due to the shuffler affecting degrees of support
4) the ability to diversify my card choices

Note that, if done poorly, diversity of card choices is inherently antithetical to diversity of archetypes, because as you move down the chain you have fewer cards that overlap multiple archetypes. I am very conscious of filling these multipurpose roles, and I think something is sacrificed if you just want to slot in the "next most powerful ____ ___ drop".

 

[Laz]

I run 370. This is because of two major factors. Firstly, it it because I think 360 is the perfect size for most of the reasons listed above (tight design, reliably see certain archetype-specific cards). The second major factor is because I am really bad at cutting ten cards.

 

[Changling Bob]

to me the ideal number is perfect fit, it sucks when people want to hope for a specific card later in the draft and its just literally not in the draft

I actually completely disagree; I'd rather overfill by two to three boosters, such that while almost everything is always in the pool, there's always the chance that the corner piece you're drafting around isn't there. You do need enough overlap that if you don't pick up Perfect Card, you can pick up a couple of Suboptimal But Comparable Cards so your deck doesn't just fall over. It also means you can't just force your favourite thing every time, as it might not come together, and you have to actually adapt to what's in the pool today.

Also that way I can comfortably run booster tutor with a cube pack.

 

[CML]

Bob -- I used to think the same, but after drafting my 450 with 3 people it's apparent smaller groups have a real and (in my opinion) negative effect upon the drafting and deckbuilding process. Many of the themes just get ignored, missing their other halves (say, if you're drafting 6 people out of 450 for 3/5 of the Cube), and most of the decks end up being a bland-ish "3c midrange" kind of thing. I know this is the opposite of what one might expect, with signals being more important for smaller groups, and I'm not sure I have an explanation for this effect being weak, but there's a greater diversity of decks when we get 8 people (or are fortunate enough to get all 10). Maybe the Modo Cube needs a larger card pool so that there are 'endless' permutations of What Gets Drafted and What Doesn't, but I'm skeptical of even that claim, and given that the average 720 is drafted once a month ...

 

[Changling Bob]

I mean, I'm talking from an always getting 8 players perspective, which clearly a lot of people don't, and from 390 rather than 360, so it is just 4 extra boosters. I've not had the chance to actually test this properly, but yeah, with large chunks of the cube missing, I would expect what you're saying I think, I just want a small amount of variety for flavour. A lot of board games I play do this (take out some number of things before you start playing, so each game is a little bit different).

 

[Jason Waddell]

It also means you can't just force your favourite thing every time, as it might not come together, and you have to actually adapt to what's in the pool today.

The only people who force the same thing all the time do terribly here. Things aren't always open by virtue of competition, order in which bombs are opened, somebody taking your overlapping support.

I also consider 8-player drafts to be of substantially higher quality than 6-mans.

I would sooner maintain several slightly modified 360 card lists than one 400-card list.

 

[CML]

tell me about this legendary 7.5-card booster