General Sunk Cost Fallacy and the impact on the evolution of cube design

Random shower thoughts: as I read discussion about cubes elsewhere, I have noticed there is a strong tendency towards aggressive hive-minded defense over the established norms of power cube construction. I think that Cubes (for Magic) intrinsically have a couple of features that will lead to sunk cost fallacy for many.

An obvious one is that the cubes many build are horrifically expensive. A related point is that in many cases they can’t or won’t play test the cube before assembling it. An even more subtle point is that many won’t playtest a card before acquiring it, leading to a natural compulsion to put it in the cube and get their money’s worth.

This is not that novel an idea, but just something that came to me while thinking about how incredibly under-developed the cube concept is given how long it has been around and how many people make them.
 
Another thought: Maro recently asked on twitter whether people would want to see a cube retail product from WotC, and many people responded saying that a cube is a personal thing and people would be unhappy with the WotC cube. I find this ironic and really unfortunate for several reasons:

- The vast majority of cubes I have seen are absolutely not a personal thing and are just a copy of another cube with a couple tweaks, and in 95% of those cases they are a copy of 1-2 cubes in particular

- It would be extremely satisfying to see what R&D does for a format like this and what we could learn from it

- If WotC printed a non-powermax cube, it would grant some legitimacy to the idea that other cube concepts can exist

- I would just like to hear Maro talk about cubes on his commute to work for a few weeks
 
Random shower thoughts: as I read discussion about cubes elsewhere, I have noticed there is a strong tendency towards aggressive hive-minded defense over the established norms of power cube construction.

[...]

This is not that novel an idea, but just something that came to me while thinking about how incredibly under-developed the cube concept is given how long it has been around and how many people make them.


For most cubes out there the design element is small to non-existent. Card evaluation is the basis for making changes. "Oh look, they printed Dauntless Bodyguard! It's a strictly better Savannah Lions, so there's one easy swap" is the logic for including or excluding a card.

And I think I understand why? It's not because they are lazy lizard plebs and we are galaxy-brained riptide geniuses. They're applying their card evaluation and deckbuilding muscles to set curation. If you have a pile of cards in front of you that you want to form into a cube, or a cube that you want to refine, you can just put that muscle to work finding stronger cards to replace weak ones and ways to further support the dominant strategies. Just as you want the strongest deck you can make out of your draft pool, and the best draft pool you can get from your seat, you provide the best cards you can to your drafters (up to some arbitrary power limit).

If it were a conscious design choice, it would be a preference for power and traditional color pie. I don't think it is conscious though. That's just what cube is to most people: a collection of one each of the strongest cards available. And for most people this works! The cube owner is still expressing something personal in the packs, even if it's just their preferences and peculiarities of card evaluation. Drafters draft in the comfort of knowing approximately what's in the cube, because it's the same cast of beloved characters from the world of eternal formats, commander, and other cubes.

More people should play cube, even if it's the kind where the drafts are a lot more fun than the actual games. It's a gateway drug to the more refined, labratory-quality products available only here
 
I totally agree that it's exercising well-trained muscles, and thus a much more straightforward and easy approach to an otherwise complicated problem. In a similar vein, I know many people who have played constructed Magic for years but never construct their own decks. It's not for everyone. It's hard to design something without some constrains, so the "must be one of the 50 most powerful cards in this color" constraint makes things easier from a card selection point of view.

However, I think I would disagree that this basic power heuristic is a meaningful form of self-expression. This information is all over the place; detritus from constructed staples and limited pick-lists are ubiquitous if you follow Magic. Replacing Savannah Lion with Dauntless Bodyguard or Grave Titan with Grave-Titan-for-5CMC is not a particularly novel idea, heh.

Now, if someone designed a cube that is more like a non-existent limited set without the LSV cheat sheet, with a power level that is somewhere in the middle, then I think there is definitely some more interesting personal choices to be made. And some actual (sometimes uncomfortable) challenge in card evaluation.

Sorry if this all comes off as pretentious and whiny, but I've been away from the hobby since RTR and it's kind of sinking in just how little people have developed the corpus of knowledge behind building a limited format outside of here (and R&D, of course). I expected to come back and find rigorous peer reviewed white papers deconstructing the design of ISD and ROE limited, heh (and I know the type of people capable of that play Magic since I work with a bunch of them!).

One last thing, at the risk of sounding contrarian: I'm also not sure that the basic power-max cube is a good gateway. I have at least one(!) coworker who actually had a terrible time playing a typical cube, and a few more who have some healthy skepticism over whether it's worthwhile. They have enjoyed the set cubes I made, though, and seem on board to giving Grillo's cube a shot (especially since everyone is sick of M19 limited now, heh).
 
Also, while it's true some people love the comfort of their favorite cast of banned-from-constructed goons (heh), just as many people seem to go into a frenzy when release events come out. I imagine there are some that prefer comfort and some that prefer novelty (and some that like a balance of both).

EDIT: I totally admit that I have done nothing to push the format forward myself, either, and don't have first-hand knowledge about how difficult it is.
 
Well you can say you have arrived at the right web page. Riptidelab has been doing that since Day 1 which is quite many years ago.

Go through all the articles on the front page and all the forum-created threads here. You will find almost nothing but genuine creative ideas as for how to 'riptide' your cube from a powered cube into a 'riptide' cube.
 
Oh and on a complete tangent, if anyone has any links to people deconstructing successful retail draft formats, I'm all ears (eyes?).


It's been linked around here before but Jesse Mason has written a series on his blog where he reviews every Magic block from Alpha to Theros (http://blog.killgold.fish/ under Kill Reviews on the right). He talks about limited for the sets that he drafted or was aware of, but also goes into detail on other aspects of the design and presentation of each block. His writing isn't for everyone but I enjoy it.
 
Haven't posted in awhile, but this is a topic near and dear to my heart.

I'll play devil's advocate though just to try and balance the discussion somewhat. Power max design has led to a wealth of information that one person - or even a small group of people - would be simply unable to compile without sinking hundreds of hours of your personal time. Most people can't do that. And in a way, what MTGS is doing isn't that much different from what is happening on this and other sites. This site has a group of like minded cube managers exchanging information on lower power/more balanced cube designs. And there is a lot of sharing here too - you could even call it hive mind in nature. I mean, how many bounceland cubes do we have now? ;)

I've never played with walkers but I did run moxen and a lot of high powered cards. So I don't think I was far off from a typical power max cube experience. I'll be honest, it was fun. Was it balanced? Not really. I think environments like that have a lot more variance. That wasn't always bad though. Sometimes that worked to the benefit of our group. Not everyone is good at drafting. And if you have a mox and some stupid card, your deck can be bad and still compete. Problem of course comes when the good drafter makes a really good deck and then proceeds to break it in half by also having moxen and other GRBS. Even then though, you have games that get determined simply by the random distribution of those power cards. Even the best deck can lose. Did you draw your busted card or not? Sometimes that simply matters more than anything else.

All that said, I think power max design is inherently flawed. And my biggest beef with it is the idea that card X is strictly better than card Y, thus the only logical choice is to replace it - that concept completely ignores how a limited meta works. I tried arguing that countless times over the years, but it went on deaf ears with the power max crowd. I stopped posting on MTGS when they switched to Twitch. Just felt like the right time to leave. I occasionally drop by just to see if there's been any change in thinking, but to this day the "this or that" thread still has X > Y > Z with zero context. What can you do? Cube really should have become the dominate format for this game. It simply better than any other version you can play. But the community got super narrow minded with it and never found a way out of that, so it has lost a lot of it's appeal IMO. I think EDH is a lot more popular now and has become the casual format of choice (though I could be wrong about that).

This is a great site though. For new cube designers, I actually think they should start with the power max design. And that goes back to the original argument I made above - because there is less individuality, you have more data to make card choices. There's less initial work required to make a playable cube. You can only build one cube basically, but it's arguably a well tested design despite it's flaws. Riptide being more individual, it's harder to get really solid data. A card might be amazing in someone's cube and get a rave review, but that will often times have to do with other card choices in that individual cube. Power max design relies more on raw card value, so you are less likely to wind up with duds in your cube by following the community feedback. I think this is valuable to people new to cube.

For me personally though, I think the ceiling on what you can build with the data on this site is higher. It's just more work upfront to get there.

Cost is another problem with this game in general, but I feel like that plagues Riptide cubes too. I expect the Chinese proxy makers to eventually become commonplace with casual formats though - particularly Cube and EDH. And when that happens, the cost factor will be significantly diminished (to the benefit of these formats). I'm personally waiting for the day this 9 year running bull market finally ends and people wake up and realize how stupid it is to pay $2500 for Juzam Djinn. I'm selling or have sold most of my expensive cards in preparation for the inevitable.
 
Haven't posted in awhile, but this is a topic near and dear to my heart.

I'll play devil's advocate though just to try and balance the discussion somewhat. Power max design has led to a wealth of information that one person - or even a small group of people - would be simply unable to compile without sinking hundreds of hours of your personal time. Most people can't do that. And in a way, what MTGS is doing isn't that much different from what is happening on this and other sites. This site has a group of like minded cube managers exchanging information on lower power/more balanced cube designs. And there is a lot of sharing here too - you could even call it hive mind in nature. I mean, how many bounceland cubes do we have now? ;)

I agree that this is a complicated enough endeavor that necessitates the collective wisdom of a bunch of obsessive people on the internet! Perhaps cubes wouldn't exist at all if the power max crowd didn't initially pioneer things? Although, I wonder if slightly remastered retail set draft simulation cubes would have been more popular if there was a void to fill. Tweak something that already had 100's of hours of r&d put into it. Remove some of the pack rats and battle of wits, put taplands in the place of some of the unplayable commons, slightly beef up BR in ISD, etc. There might be a bit of a chilling effect on new designs because of the established presence of the typical cube and their advocates. Science advances one funeral at a time.. (sorry I've been looking at Innistrad cards)

I've never played with walkers but I did run moxen and a lot of high powered cards. So I don't think I was far off from a typical power max cube experience. I'll be honest, it was fun. Was it balanced? Not really. I think environments like that have a lot more variance. That wasn't always bad though. Sometimes that worked to the benefit of our group. Not everyone is good at drafting. And if you have a mox and some stupid card, your deck can be bad and still compete. Problem of course comes when the good drafter makes a really good deck and then proceeds to break it in half by also having moxen and other GRBS. Even then though, you have games that get determined simply by the random distribution of those power cards. Even the best deck can lose. Did you draw your busted card or not? Sometimes that simply matters more than anything else.

This is funny, because I have been close to starting another thread about variance and whether it's missing from both styles of cubes. The uncertainty of which cards you will see in a draft pool creates some excitement in retail limited that I think would be interesting to explore more in cubes. And not just big piles of cards like in the larger cubes, but careful distribution; build arounds seeded at uncommon, excitement at rare, stability and archetype reinforcement at common, like the most beloved retail formats (but minus the pack rats and battle of wits). Go ham on the AsFan. The bombs that aren't outright broken can usually be played around (and create tension), and so it might be worth the trade-off over the more narrow powerband.

Basically, it seems like riptide has evolved from taking a step backwards from power-max cubes, but it would be interesting to see the parallel universe where they evolved stepping forward from limited retail formats.

edit: I have played the modo cubes. They kind of feel like .. fan-fiction? They are neat at first but they also feel perverse somehow. And some games just play like the card game War.

All that said, I think power max design is inherently flawed. And my biggest beef with it is the idea that card X is strictly better than card Y, thus the only logical choice is to replace it - that concept completely ignores how a limited meta works. I tried arguing that countless times over the years, but it went on deaf ears with the power max crowd. I stopped posting on MTGS when they switched to Twitch. Just felt like the right time to leave. I occasionally drop by just to see if there's been any change in thinking, but to this day the "this or that" thread still has X > Y > Z with zero context. What can you do? Cube really should have become the dominate format for this game. It simply better than any other version you can play. But the community got super narrow minded with it and never found a way out of that, so it has lost a lot of it's appeal IMO. I think EDH is a lot more popular now and has become the casual format of choice (though I could be wrong about that).

Yeah, it's just surprising to me that after 5 years, nothing's changed. This thread is attempting to rationalize it as the consequences of sunk costs in those $2500 djinns you mention.

Cost is another problem with this game in general, but I feel like that plagues Riptide cubes too. I expect the Chinese proxy makers to eventually become commonplace with casual formats though - particularly Cube and EDH. And when that happens, the cost factor will be significantly diminished (to the benefit of these formats). I'm personally waiting for the day this 9 year running bull market finally ends and people wake up and realize how stupid it is to pay $2500 for Juzam Djinn. I'm selling or have sold most of my expensive cards in preparation for the inevitable.


I'm assembling a couple cubes, but I'm definitely proxying the expensive cards. That's one thing that has gotten much better over the last half decade. Try before you buy

I would still probably give 1/5th of a Juzam Djinn for a R&D researched drafting cube.
 

Chris Taylor

Contributor
Three things standing in the way of an R&D cube:
1) modo cube
2) are people actually going to buy a 300$ msrp product from wotc? (this assumes its penny pincher levels of cheap btw)
3) how different is this from a drafting perspective from say, the modern masters sets? Except this doesn't make modern cheaper and you can't ever experience it if all you have is 5$ (10$? I haven't bought a pack in ages)
 
It's been linked around here before but Jesse Mason has written a series on his blog where he reviews every Magic block from Alpha to Theros (http://blog.killgold.fish/ under Kill Reviews on the right). He talks about limited for the sets that he drafted or was aware of, but also goes into detail on other aspects of the design and presentation of each block. His writing isn't for everyone but I enjoy it.

This is really good. A breath of fresh air from the usual channel fireball non-critical style piece (which I read too, of course)
 
Three things standing in the way of an R&D cube:
1) modo cube
2) are people actually going to buy a 300$ msrp product from wotc? (this assumes its penny pincher levels of cheap btw)
3) how different is this from a drafting perspective from say, the modern masters sets? Except this doesn't make modern cheaper and you can't ever experience it if all you have is 5$ (10$? I haven't bought a pack in ages)

Why would a cube need to be 300 bucks, especially if it's penny-pincher levels of cheap.

Commander decks are 100 cards for 30 bucks. 30x5 (if we include basic lands) is only $150. Still expensive, but pretty reasonable.

Also, I believe $20 Deckbuilder's Toolkits have a couple hundred cards in them. I know you get 4 boosters, a bunch of land, and some common and uncommon filler in each color.

All things considered, I think we could easily see a cube starter kit that only costs ~$70. If card kingdom can create a cube product that comes with a full set of basic lands and card sleeves for $100, then I'm sure the mothership could release one for $70 (even with sleeves!)

Remember, we're just talking about fancy cardboard here. It costs the same amount to print a Scalding Tarn as it does Scornful Egotist.
 
This is funny, because I have been close to starting another thread about variance and whether it's missing from both styles of cubes. The uncertainty of which cards you will see in a draft pool creates some excitement in retail limited that I think would be interesting to explore more in cubes. And not just big piles of cards like in the larger cubes, but careful distribution; build arounds seeded at uncommon, excitement at rare, stability and archetype reinforcement at common, like the most beloved retail formats (but minus the pack rats and battle of wits). Go ham on the AsFan. The bombs that aren't outright broken can usually be played around (and create tension), and so it might be worth the trade-off over the more narrow powerband.

Basically, it seems like riptide has evolved from taking a step backwards from power-max cubes, but it would be interesting to see the parallel universe where they evolved stepping forward from limited retail formats.


I agree with this. When I first started working on my midrange/balanced cube, I was obsessed with removing bombs. To the point where drafting got harder (in a sort of annoying way) simply because there weren't any obvious first picks anymore. So I actually think a really flat power level is the wrong way to go. It makes signaling harder, finding a direction harder and it indirectly neuters synergy.

It's also worth noting that having a handful of really powerful (even broken) cards won't necessarily break your meta. I think we often look at constructed and say "well, card X totally warped the meta! So one card is enough to ruin things!" But in constructed you can have 4 copies - ensuring it factors into most games you play. And more than one person can run that card - meaning it will show up all over your meta. This is nothing like how a cube draft works. Not to go back to defending the power max crowd again, but this is something I feel gets downplayed a bit. And that group has been right to point it out. It looks like it's more of a liability than it actually ends up being in practice (having a few busted cards in your cube).

After I built and tested my kamigawa/innistrad themed cube - which was way lower power than anything I've done before - I came to a similar conclusion. Meloku, the Clouded Mirror is a straight up bomb in that cube. But it's a single copy of a card. And even if you feel it's so so good that if you open it you simply must put in in every deck you make, how many games will it even see play in? What is the true impact of that card in your cube meta? I used to snap pick cards like this in draft thinking I was going to always benefit from having higher powered cards in my deck. And this is mathematically accurate I think, but the impact to the games played in reality was always less than the raw value of the card. Sure, sometimes your bomb would win. But most of the time you'd either not draw it or draw it at a point where it wasn't going to save you. Further more, if I always pick Meloku without question, I am forcing myself to always play blue. Even if the color is being cut or the rest of my deck would rather splash black. It's almost a skill testing thing. Here's a bomb card that you should always pick... or maybe there are times you shouldn't?
 

Grillo_Parlante

Contributor
While yes the cost is a problem, the biggest obsticle that I had was just human psychology. Not only do you have to come up with a creative idea that your players can actually stand (which is brutally hard), but than if you actually stumble upon a good idea, you now have the nightmarish task of marketing and selling your ideas to other people, who for the most part, have no reason to care.

Remember, anything you come up with is going to be at best an impossible to evaluate curiosity, and at worse an existential threat to the integrity of that community.


I would consider the high cost of cards existing within the context of a broader frame, where it serves as a rational for someone to dismiss a thing whose newness is disrupting the comfort of their reality.

If the community was truly being rational (which it isn't), cube design would revolve around cheap cards. I kind of also suspect thats its always been much more niche than it looks, with the majority of players being established core players that have been playing for years. Its natually a very complicated format, they make it more complicated, and complexity tends to be a huge barrier to new player entry.
 
I do think cube is a complex format and that it appeals in particular to established (or at least experienced) players. It is pretty niche. But that might be underselling the format a bit. I left the game and cube brought me back. That might sound cliche but it's not an overblown or dishonest statement. For the last 10 years, I have played no other version of this game and have no desire to either. IMO, cube is that much better. It's literally ruined all other formats for me.
 

Grillo_Parlante

Contributor
I do think cube is a complex format and that it appeals in particular to established (or at least experienced) players. It is pretty niche. But that might be underselling the format a bit. I left the game and cube brought me back. That might sound cliche but it's not an overblown or dishonest statement. For the last 10 years, I have played no other version of this game and have no desire to either. IMO, cube is that much better. It's literally ruined all other formats for me.


Oh yeah, no doubt. <3 <3 <3 cube. I don't really want to hate too much on power max because one of the best parts of cube is that it allows you to somewhat revist the past, and get an idea of what was there.

I also appreciated your comment about bounceland formats. I didn't really know how to say it myself, but it would be nice to see some more creative design, other than my own set pieces, but its hard to say that without sounding hypercritical, and I understand how much work goes into it. Sometimes one person's success can become a reason for other people to not push their limits, and we have some talented people here. Would be a shame.

Nice to see you back, I was wondering where you had gone too. Thought you had quit the game.
 
I agree with this. When I first started working on my midrange/balanced cube, I was obsessed with removing bombs. To the point where drafting got harder (in a sort of annoying way) simply because there weren't any obvious first picks anymore. So I actually think a really flat power level is the wrong way to go. It makes signaling harder, finding a direction harder and it indirectly neuters synergy.

Hey, me too! I definitely had an issue in like 2015 where I was trying to make my power band as flat as possible, really amped up on synergistic picks. But the fact of the matter was that my drafters didn't share my enthusiasm for P1P1-ing between two equivalently-strong answers, or threats they couldn't tell were supported. I got some feedback to this effect, and it's what spurred my current design phase, which I've been thinking of as 'a Riptide dragon cube'. These are what I think the main mistakes I've made along the way have been (and who's to say that my current goal isn't also a mistake? Cube design is littered with beautiful imaginative idea threads that, six months later, get a final reply from OP: 'yeah, this didn't work very well').

-not playing enough Limited
-not playing enough Constructed
-not playing enough Standard
-excessively narrow power band (fewer exciting picks, drafting becomes 'hard work')
-excessively wide power band (the deck that curves Thoughtseize-Hymn-Lili-Jace beats the white weenie player every time)
-making too many little tweaks (by shoring up my new archetype celèbre, I weakened elements of the Cube I hadn't thought about recently. I had an issue where I kept cutting sac outlets for similar cards - except they weren't similar, because they weren't sac outlets! Oops)

There's probably more, but those are the big ones! What I find frustrating about this experience is that failing doesn't really teach you where you should go next; it just removes one idea from the infinity ideas available to you. Without a wealth of experience to draw on, or the - quite frankly - obsession common to the posters here, there are just too many ideas to possibly try!

So we limit ourselves artificially. Restrictions breed creativity, as Maro is fond of saying, but they also kill indecision, which might be even more important. A bounceland cube. A graveyard format. A combo environment. A mature attempt to recapture the emotion of young love (me, grillo, others).

But does it really matter that we have yet to find El Dorado? Maybe static perfection is the wrong goal. My Cube goes through cycles of sleepiness and the quick, chaotic growth of the first blooms after winter. To be honest, who really cares if it's still the same Ship? Theseus remains in charge, and it's still sailing, and probably going roughly the same place. Plus, I only know a couple people with their own ship. If experimentation is fun, and doesn't provoke existential dread, why not? Human brains crave novelty.

Anyway, that's how I used to feel. Maybe I still do? But in the last couple years, two people that I know of have proxied up versions of my Cube, and play them with their local playgroups. I feel a sort of parental responsibility to them, to make sure that my format doesn't have any obvious issues for people to stub their toes or hit their shins on. And so I went back to the drawing board, and said 'ah, how do I keep most of the same cards, but have a more coherent vision for them?'

And therein lies the rub!!!!!

I did exactly this sunk cost fallacy bullshit. Even though I could proxy any card I like, designed by Wizards or not, the sunk cost fallacy drove at me. The cost wasn't cardboard - it was my time. The vision was too large. And yet, I've rebuilt my Cube from the ground up three times, and every time I did, the Cube got a lot better.

Part of the fun of diverse Cubes, to me, has been seeing the echoes of the designer's personality in the kitschy, 'niche' card choices. Ah, I can say to myself. I bet you loved that Standard format. I bet you never played against this in draft! And the rarest, best version of these thoughts: 'huh. after reading that card, it seems awesome here'.

This has been an incoherent diatribe, I think? But I've expressed the deepest thoughts of my heart and soul.

PS: always slam Meloku. c'mon, life is too short not to live the dream.
 
Many things coming to mind after reading this thread.

An obvious one is that the cubes many build are horrifically expensive. A related point is that in many cases they can’t or won’t play test the cube before assembling it. An even more subtle point is that many won’t playtest a card before acquiring it, leading to a natural compulsion to put it in the cube and get their money’s worth.
This describes my methodology very well!

Even though I've started playing more with proxies for expensive cards, I still go ahead and buy a lot of foils just to try them out, and because they don't look that expensive, and maybe one day I might play commander again and who knows, and other excuses. Now I probably have a second cube worth of rejected cards populating my closet, and I've been thinking of just piling all these cards together and see if they look like a cube.




I need to formulate some trains of thoughts and come back with a comment tomorrow, but I'm not sure I completely agree that sunken cost fallacy is a driving force in power max cube. To me it feels like the traditional power maxing is more akin to a conservative personality, where there is a certain essential nonmaterial ideal that is being represented in that grouping of cards that resonates with the person, and the fact that there is a whole community that shares this simple and simplistic foundation feedbacks into peer acceptance for this defense of this view of the spirit of the cube. The antithesis to this would be something more of a materialistic world-view, which requires a great deal of deconstruction and criticality, but might lack the same type of cohesion the other mode of being has.

With that said, I don't think that it is fair to say that other places are a more like a hive-mind than we are, because I think we might be, but instead of defending a way of defining cubes, we defend a way of thinking cubes. And somehow most of us figured out that running Chromatic Star is cool.
 
Top