It appears the voting periods are largely agreeable so I'm going to go ahead and state that we will be doing M-F voting periods for each of the sections.
Guild voting will go 10 deep and split ally/enemy on different weeks. Shards/Wedges will only go 2 deep and will be all 10 on the same week.
We can go with the intrinsically best standard for color classification. So basically, I think we are mirroring 2016's classifications with a few notable changes:
1) Noble Hierarch is Bant again. (I'm not keen on the idea of having a card voted upon in multiple sections and the arguments here overwhelmingly favor classifying Hierarch as Bant).
2) DOJ and *** are grouped together this time.
3) Bomat Courier and Scrapheap Scrounger will be red and black respectively.
I'm still lacking a concise definition to use for our voting criteria. I've read through the comments and these are my thoughts:
@Karshtakavaar: I think you're right in that this project is somewhat specific to the MTGsalvation audience which is generally more interested in what is strongest and not about diversity when it means sacrificing power level. I don't want to disinclude anyone, but I also would expect someone with a top 20 of mostly budget or fun options may not see much, if any, of their list reflected in the top 20--though it would be on the spreadsheet. If you're okay with that then by all means give us your top 20! I think that is preferable to a lot guesswork as to where cards you don't play or have experience with might fit.
As we get more options cube diversity is expanding, but you're presenting power vs diversity as if they are an either/or choice and they aren't necessarily. We're lucky enough to be able to achieve both goals with greater ease now. Many cubes are upping their size due to the amount of powerful choices that have become available to us. In short, I don't think the power axis is being edged out so much as the number of options on this power level has increased. Archetype design is very important, but there's also a lot of powerful cards that don't fit cleanly into an archetype but are definitely worth cubing. I think a healthy cube has some clearly defined archetypes, but isn't beholden to them. We should see aggro, midrange, and control supported in every cube, but when a cube is running, say, Hardened Scales to support a +1/+1 counter theme in lieu of stronger generic cards, I personally feel that is a mistake.
@wtwlf: I don't agree with your statement about the 2016 definition using an "arbitrary/inconsistent powerlevel metric that nobody can possibly agree on". In fact, if anything there was a little bit too much agreement which is partially why I want to incorporate the use of spoiler tags this time. I strongly prefer less sentimental criteria such as "favorite" or "most loved" in favor of "most played" "most impactful" or "contributes most to a win". I want to keep hearing feedback and hammer down our definition for voting criteria in the next couple days. Thanks again for your feedback.
I don't think spoiler tags solves the "problem" of people being influenced by other voters' lists. I mean, just open the tag and look at the list. I understand what it's trying to do, but if you really want everyone to be submitting their lists uninfluenced by other lists, then you should do this via PM.
I also don't think we'll be able to come to a true unanimous agreement on the voting criteria. Even if you tell folks to post their top twenty "most impactful" white cards and everyone agrees on that, you'll still gets posters who interpret that differently. What does impactful really mean in terms of cube? Is Armageddon the white card that's had the most impact on games for me? Does that make it rank higher than Moat or Balance? Maybe I'll just shortcut impactful to mean P1P1 or favorite since these are easier for me to interpret. I don't believe there's really a way to avoid this metric being mostly arbitrary based on the individual posters.
I do not play any shard cards (besides Sphinx of the Steel Wind which is rarely hardcast so might as well be colorless), yet I play Noble Hierarch as a green card.
I really have two favorite cards lists if we go that route - one as a player and one as a designer. The first includes cards that I like to play personally, and is biased and influenced heavily by the decks I love playing. The second factors more heavily issues like archetype support, playability and balance. For example as I player I value Searing Blaze over Abrade, but as a cube designer it is the opposite.
I suggest that if we do decide to use favorite cards as a criteria, we should be after the second type of list - favorite cards as a cube owner/manager/designer and not as a cube player/drafter. The second is a lot more valuable. Of course our biases as players affects our judgement as designers so it isn't a cut and dry separation. But at a cube manager's perspective we need to cater to many more needs - the likes of several players and the entire environment, which is less prone to outliers.
I still think P1P1 is the better criteria as it is well-defined. Of course it will be different from cube to cube, that is why we do this project. The individual vote doesn't matter that much, it's the aggregates that count.
Yeah I agree that Hierarch should be a green card because it's good without being in Bant, whereas Yasova and Alesha are garbage without access to their other color identities so should be considered as wedge/shard cards. Ultimately it's only one card and we really shouldn't spend so much time on it, but without pedantry we wouldn't be cube designers I suppose.
(BW, this is your project and you're going to be doing the leg work on it. You've already been generously solicitous of everyone's opinion, and we're all dead split on it so you're not going to find consensus. Just tell us what to do with Hierarch and let's all move on.)
I also think a P1P1 (or most powerful, or best) makes the most sense, especially since maximizing power level is the overriding philosophy on this forum. If people want to have an opportunity to highlight their favorite cards or whatever, just add a "Favorite Cards" section of like 5 or something at the end of the real rankings.
P1P1 or most powerful or best are all wildly subjective, and completely contingent on the list you're drawing the data from, and the playgroup you're drafting it with.
If we want to do P1P1, we should do far less than 20 cards per section. How many white cards are honestly P1P1-quality cards? Like 5 maybe? Also, P1P1 value is split into two very separate camps. There are those that think that safe, universally playable cards make the best P1P1 choices, and there are those that think that snagging the most powerful card for the right deck is worth gambling on. As much as I like Path to Exile, I don't honestly know how awful my pack would need to be in order to P1P1 that card. Best is subjective. Most powerful is subjective. ...I just don't think those metrics provide useful data anymore. Cubes are too diverse and varying powerlevels are too contextual.
Am I seriously the only person that thinks that a list of peoples favorite cube cards wouldn't be more valuable? It can be the list of "the 20 best white cards to include in cubes" ...which is data everyone can accurately vote on, everyone's opinion is equally valid, and you can use whatever metrics make you excited to cube with certain cards. BlackWaltz3, this is YOUR project that you're championing. If you want to pick a metric and have people try to conform to it when voting, just do it. But I can't imagine a powerlevel definition being outlined that is both accurate and inclusive.
I agree that P1P1 is the metric that is easiest to parse. Would I take Card A over Card B? If yes, then rank accordingly. However, the biggest problem with that is the number of true P1P1s in any given color fall off fairly quickly. Maybe there's five or ten white cards that I would realistically take P1P1 given other options, but what does a pack have to look like for me to take Path to Exile over options from other colors (or colorless)? And Path should almost definitely be on a top twenty list for white.
I guess an important question to ask is what is the purpose of the ranking? Is it to generate a pick order list for someone who is drafting? or is it more for cube designers a list that they should think about including in there cube (provided they are building a cube to the same standards of most powerful cube etc). When I first started building my cube I found these rankings mostly useful for that purpose. If that is the case, the metric could focus on that aspect. Maybe something along the lines of "how important this card is to the overall cube experience?" Any metric we pick is subjective, but this might capture the heart of what the list is trying to do.
Also, if Noble Hierarch is Bant, then the Mox should probably be in there respective colors and not in colorless? This would also make an argument that there could possibly be a [mana]WURBGa] section, and Black Lotus, City of Brass and the like would go there since they are intrinsically best in 5 color decks. (not arguing that we should have a 5 color section, but simply pointing out the flaw in using the "intrinsically best" isn't quite the line that makes sense to me.
I much prefer "most important to my cube" to "p1p1" since it reflect more of what makes cube cube to me, and also doesn't give weird priority to bombs that may not be super relevant to multiple archetypes.
P1P1 or most powerful or best are all wildly subjective, and completely contingent on the list you're drawing the data from, and the playgroup you're drafting it with.
I mean, "favorite cards" or whatever is way more subjective, since by definition it's nothing but personal opinion. At least P1P1 has some basis in objectivity w/r/t perceived power level.
More to the point, I don't know what the point would be of the project you're proposing. Like, if it's just "I want everyone to know about this cool card", that's fine, but why not just have a big thread about people's pet cards? Or hell, just make an SCD about it. God knows this place could use some more activity outside of spoiler season. But what difference is it to anyone that Parallax Wave and Porphyry Nodes are two of my favorite cards? Who does that help? And why do we need to have a vote on it?
Anyway, if P1P1 is such a difficult thing for people to grok (I really don't think it is, but whatever), it's not hard to come up with a different way of thinking about it. Something like this:
Edit: Calibretto said it better. Would I take Card A over Card B, all else equal.
I mean, "favorite cards" or whatever is way more subjective, since by definition it's nothing but personal opinion. At least P1P1 has some basis in objectivity w/r/t perceived power level.
Correct. But if the goal is obtaining good data, we can get that with asking people their opinions on their favorite cube cards. It's better to ask a subjective question with the goal of obtaining data that can genuinely be answered by it than to ask subjective questions if the goal of obtaining objectively useful data. The voting sample size won't be even remotely large enough for the variance to be ironed out by averages. Not by thousands and thousands of voters. But with peoples favorite cards, the information is honest and equal (and honestly? about a million times more valuable).
Quote from calibretto »
And Path should almost definitely be on a top twenty list for white.
Then we're asking the wrong questions. Because if that's making a list of "top P1P1 cards", the data we're getting is pretty meaningless.
Quote from braid of fire »
I much prefer "most important to my cube" to "p1p1"
Me too. Everybody can vote, it doesn't matter what metric(s) you use to determine which cards make the list for you, and it doesn't matter what kind of cube you prefer to draft. I'd rather do this and err on the side of inclusivity.
A suggestion - if we decide to go by most important cube cards, can we not order them? Otherwise we need to decide if the best six drop is more important to the cube than the best mass removal which are impractical. I'd not play a cube without either.
I agree with the notion that an agglomeration of favorite cards is not very useful. I also don't think it would lead to more sensical results necessarily - if Goblin Welder is the only artifact matters card in the list, it would be ill-advised to jam it without support. You still need to know the context of the cards on the list, and it will still be subjective, there is no getting around this issue.
But again, those inconsistencies are more acceptable when trying to determine a list of peoples favorite cards. There's underlying context within that prompt that we can deal with. But if you're using the data to try and determine something meaningful; data that can be followed ...those inconsistencies are less acceptable.
And Path should almost definitely be on a top twenty list for white.
Then we're asking the wrong questions. Because if that's making a list of "top P1P1 cards", the data we're getting is pretty meaningless.
Agreed. That was my point. It's rare that I'd ever take Path at P1P1, so if it's on a list of the top twenty P1P1 white cards it raises a lot of questions about how useful the data actually is.
Why not just say "These are the first twenty cards I'd include in X section in order of importance"? These lists should be here to support new cubers looking for a resource when building their cubes.
Wow, alright, lots to address. This is unfortunately the point where I need to be less generously solicitous (thank you Hoodwink) in the interest of moving forward with the project--imperfect though it certainly will be.
As far as voting criteria, my inclination is to have not just a one factor definition such as "best" "p1p1" or something like "favorite" but rather a short series of factors (maybe two-three sentences) to consider as guideposts for people's lists. I'm thinking the most important measures are "how critical this card is to winning (not upon resolution)", "how often would you main-deck this card", and "most important to my cube". I don't claim these aren't subjective, but they aren't as widespanning as favorite/most loved or as narrow or uninclusive as "p1p1" or "best" I think they are good middle-ground guideposts. It improves upon the definitions used last time, but doesn't change them so radically that this can no longer be validly compared to our past results.
Noble Hierarch is going to be whatever there are the most comments in support of. I started with Green and then all the Pro-Bant people commented, so I said "Fine, Bant" and then the pro-green people chimed in....lol.
I count myself, Hoodwink, and allred for team Green and a few others have commented for team Bant. I'll just let the numbers tell me what to do and spare me the headache.
Okay now some user specific feedback:
@calibretto: PM's probably would accomplish my goal of limiting influence from others lists, but I am NOT going through a bunch of PM's to compile the data! Hehe. I don't want people to feel they shouldn't look at others lists, but I will encourage people to try to create their lists independently. Spoiler tags may help even if they are an imperfect solution--at the very least if there is an early trend among the first voters, people don't have to see it.
@ Metamind: Ordering is necessary (also makes it a lot more fun for people). Try not to think of it as you have but by the "Would I take Card A over Card B, all else equal" presented earlier in the thread.
I realize these solutions may not be pleasing to all, but I have to do the best I can and then get on with it. I'm really excited to see what out results look like 3 years later. I know a lot of people loved this project then and I'm hoping it will be just as fun and interesting today. I'm definitely going to include fum summaries of the results as Spike Rogue did because that was one of my favorite parts and I'm sure others loved that too.
Thanks a lot for organising this, it’s been a long time since we’ve done this on the forum and I’d be very happy to participate.
On the format, count me, squarely, in the power level camp. Power means different things to different people, yet we all follow wtwlf’s new set reviews religiously, we go to Cubetutor to see draft percentages, and we go to SCD threads to see other people’s views on the playability of cards. A debate can be had on which card is more powerful between Isamaru, Hound of Konda, which is played in any white aggro deck but is not especially impactful, and Monastery Mentor, which sees less play but has a far bigger impact. But anyone trying to tell me that Crystal Shard is more powerful than Tinker will have a very hard time of it and save for the means to prove it, that person would be objectively wrong. Power level is, to a very large extent, an objective notion. That it is difficult to define does not make it any less so.
Rankings cards by how much you like them is a different beast altogether. What makes people like a card? Art? Flavour text? The kind of game states it creates? How much skill it takes? How much it contributes to balance in your cube? An experience you’ve had with it in the past? I care a lot more about random strangers’ professional opinion (if I believe they are qualified – as I do in this case) than I do their personal opinion. For this reason I personally have zero interest in a favourites list and would not participate (on that note, I object to the notion that such a format would be more inclusive). That is my personal opinion though and I can imagine that other people, justifiably, may disagree.
However, there is a second, objective, problem with a favourites list. How can I possibly know why you specifically like a particular card? Maybe you like it because you think it is powerful, which is a valuable opinion to me. Maybe you like it because the card takes a lot of skill to play right, which is sort-of valuable to me. Maybe you like the card because when you were a kid you built a deck around it and it’s a nostalgia monster. This last fact is not valuable to me at all. The problem is that because I don’t know which of the three (or other) scenarios applies, I can’t interpret the result in any meaningful way. Even if you were ranking the cards by power level, your list would be 100% useless to me.
In my personal case, I love Sakura-Tribe Elder because I cracked up my cousin once calling it STEve, and I think it’s a funny card because of that. Blazing Archon is one of my favourite white cards because it has amazing flavour. And I love Daze because I played with it for years in a different format. How is any of this useful to anyone? Especially if you don’t have this background information!
As for how to define power level, I think any number of criteria can apply. Let’s not get hung up on the fact that few cards are “P1P1” worthy, as if for that reason Swords to Plowshares would be as powerful as Path to Exile. Here are some possibilities:
• The priority within the colour in which you would generally draft the card in a first pack
• The likelihood of you playing the card in an on-colour deck in a sealed format
• The size of a power level-based cube where you would start including the card
• The impact the card will have on a game – e.g. how much it contributes to your odds of winning
• Your gut feeling, which will be a combination of the above. I don’t need a definition to know that Dark Confidant is more powerful than Asylum Visitor.
I’m not even sure if it matters so much to converge on a specific definition of power level. I’m sure I personally give above-average weight to the second category above, and I think for example that Isamaru, Hound of Konda is a pretty powerful card. Does that make me wrong? Why should we not be able to compare my opinion alongside someone who attaches more importance to the first criterion, and who would go rather for a card like Monastery Mentor? I’m not sure any one definition of power level is the correct one. Power is power. (Side note, I think Mentor is more powerful than the Hound!)
I will also throw in my 2 cents on the debate for classifying Noble Hierarch, although I really don’t care where you end up going with this. I care only for sport. I would generally classify it as green, and not Bant. The argument that you should classify a card in the colours where it is strongest is strange to me. If this was my approach I would classify Wild Mongrel as B/G, Sphinx of the Steel Wind as mono U and Bonesplitter as R/W. I think mana cost plays a big role, with a nod to the colours that correlate strongly with where the card sees play. You will play Hierarch in Bant, and it will shine, but it costs G and you will also play it in any Gx deck. G is the common factor, not B/U/G. You won’t play Kird Ape in anything other than an R/G deck, so it belongs to that guild. Scrapheap Scrounger should only see play in Bx decks, so it goes to black. And so on.
Lastly, a quick note on spoiler tags. I think this is good practice. Anchoring is an extremely powerful bias and I would anyway approach something like this by actively ignoring previous posts. There’s very few utility lost in having to click open such tags. Not a fan of private submissions, because that process fails to generate hype and community involvement. Plus, once you’ve finished your submission it’s great fun to look at all the other ones.
"These are the first twenty cards I'd include in X section in order of importance"?
The problem with this prompt is curve. A 20-card section wouldn't contain the 20 "best" white cards.
Quote from Person_On_MTGS »
But anyone trying to tell me that Crystal Shard is more powerful than Tinker will have a very hard time of it and save for the means to prove it, that person would be objectively wrong.
This would be exactly true for a cube that supports a blink/bounce/ETB trigger abuse deck and has no support to make Tinker powerful. Those kinds of cubes exist. Now, you're telling voters that they're objectively wrong because their notion of "best" isn't the same as yours ...when in fact they're 100% accurate and the only thing that changed was context.
The bigger problem is assuming that an amalgamation of the data is useful, when in fact, the opposite is true. If you combine the data from the Team Tinker's votes and the votes from team Shard into one dataset, and team Tinker has 20x the voting members ...what happens when you sit down to draft a cube with team Shard? The data is useless to you. And that's the problem. Cube powerlevel is too context-based to provide meaningful data with a small dataset. 25 people voting or whatever isn't anywhere near statistically significant enough to provide meaningful data. But if you ask a question that everyone can honestly answer using the same metrics, your dataset can be smaller and still produce worthwhile data.
tl;dr - If you ask the question "what are your 20 favorite white cube cards" people can look at that combined data and get a feel of what kinds of cards they could include in their cubes to make people happy. If you ask the question "what are the 20 best white cube cards" you would have to qualify that information with about 60 disclaimers before everybody's even voting with the same metrics.
Quote from Person_On_MTGS »
The argument that you should classify a card in the colours where it is strongest is strange to me. If this was my approach I would classify Wild Mongrel as B/G...
Not strongest. Intrinsically best. There's no intrinsic color identifiers that make Wild Mongrel anything but green. Your Golgari classification would be contextual organization, not intrinsic organization.
Organizing cards where they're intrinsically best is the reason why Vedalkan Shackles is blue for most cube managers and not colorless.
Echoing this because I agree with all of it (that Wild Mongrel point is making me reconsider my classification strategy), and also to wholeheartedly agree with the use of spoiler tags. Anchoring is a real thing, people!
So far these are the voting criteria definitions I am most persuaded by--again my goal is to have a short series of guideposts and not a definition that is too rigid nor too loose.
Rank your top 20 cards in terms of:
- How critical this card is to winning.
- How frequently you would maindeck this card in a deck playing this color/guild/shard/wedge (or just in general if colorless).
- What cards are most important within your cube.
The 1st two I believe are more power-level focused metrics, whereas the last measure encourages participation by those who may not share this general cube design philosophy. I really just can't justify the word favorite because I think it leads to connotations that could certainly produce wildly different and largely unhelpful results that we can't use as a standard of comparison to previous similar ranking projects. Also, I think the third definition is somewhat similar to favorite, but doesn't carry the connotations that I'm concerned about. Hopefully, that is an agreeable compromise.
Spoiler tags will be a thing. At least, I'll be asking that everyone try to do this for the sake of avoiding anchoring bias and encouraging personal reflection rather than following any trends.
Thank you guys, this has been a thought-provoking discussion already and we haven't even gotten going yet.
You may want to request that voters use a certain format or something, last time i helped do a few of the calculations, and one thing that took more time than you would think is making the data consistent so you can put itnto an excel format or slmething wothout spending to much time edditing it.
What do you recommend? Don't want to make formatting a barrier to participation--the only extra thing I will ask is that people use spoiler tags. Beyond that, I was just going to have them number 1-20 and name the cards:
What do you recommend? Don't want to make formatting a barrier to participation--the only extra thing I will ask is that people use spoiler tags. Beyond that, I was just going to have them number 1-20 and name the cards:
I recommend not having people use numbers in and just the name of the card, and ask each voter to make sure the image pop's up for the card if they move mouse over (saves issues with slight differences in spellings in cards like Jace, the Mind Sculptor where there are commas, or other cards with ' etc.
"These are the first twenty cards I'd include in X section in order of importance"?
The problem with this prompt is curve. A 20-card section wouldn't contain the 20 "best" white cards.
I mean, yeah, but how many people are building cubes with 20 card sections? If I were building a 360 card cube with 50 per section, the first 20 cards I'd include would likely be the same first 20 cards I'd include if I were building a 540 card cube with 70 cards per section.
But I feel like we're all just arguing semantics here. Ultimately, while I do find the data we get from these top 20 votes to be interesting to read through, I don't think it has much value in terms of the lists of cards we end up with. I don't mean to be rude if that comes off that way.
The type of data that I think is more valuable is either a top X for each casting cost in each color or just a cube comparison that results in how many cubes run specific cards. Looking a list of the top X White 5 drops is helpful for new cubers looking to get into the format and for veteran cubers who may just be looking for something fresh to swap in over Baneslayer. Alternatively, a cube comparison that tells me 30 cubers are running a card that I've been overlooking is helpful for me to find new cards for my cube and also helpful for new cubers looking for a starting point for a new cube. Both of these data finding methods are better, imo, than a Top 20 for each color. It's hard to agree on a metric and then get everyone voting to stick to that metric. And, who is this data for? It's unlikely that I'll find something new for my own cube on these lists. If a new cuber looks at this and just says, Oh I need to include these twenty cards, that doesn't give them any sort of context for where they need to go from there.
I think the data has a lot of value. Often the value is less in the actual rankings and more in the variety of other cards that show up on the spreadsheet that people may not have thought to try before. Also, the guilds section rankings tends to be a resource I go back to again and again. Lastly, the rankings sometimes challenge me to question my assumptions on how good a card is when I find the average is much lower or higher than I would have thought. I'm sure there's more too, but this is what I've gotten from this project in the past. Also, its just a lot of fun (probably less fun for the organizer--which is now me).
Allred, I think I'd prefer people keep the numbers just to ensure they didn't skip a number, list the wrong amount of cards, etc. Might make a little more work for me, but I'll deal.
This would be exactly true for a cube that supports a blink/bounce/ETB trigger abuse deck and has no support to make Tinker powerful. Those kinds of cubes exist. Now, you're telling voters that they're objectively wrong because their notion of "best" isn't the same as yours ...when in fact they're 100% accurate and the only thing that changed was context.
The bigger problem is assuming that an amalgamation of the data is useful, when in fact, the opposite is true. If you combine the data from the Team Tinker's votes and the votes from team Shard into one dataset, and team Tinker has 20x the voting members ...what happens when you sit down to draft a cube with team Shard? The data is useless to you. And that's the problem. Cube powerlevel is too context-based to provide meaningful data with a small dataset. 25 people voting or whatever isn't anywhere near statistically significant enough to provide meaningful data. But if you ask a question that everyone can honestly answer using the same metrics, your dataset can be smaller and still produce worthwhile data.
tl;dr - If you ask the question "what are your 20 favorite white cube cards" people can look at that combined data and get a feel of what kinds of cards they could include in their cubes to make people happy. If you ask the question "what are the 20 best white cube cards" you would have to qualify that information with about 60 disclaimers before everybody's even voting with the same metrics.
I have to say I haven't seen many cubes where Crystal Shard is more powerful than Tinker. But that's not really the point.
You can very well build a cube where Shard is more powerful than Tinker, and such cubes may very well exist. But any cube manager behind such a cube must surely be aware of the fact that his design philosophy has not been to create the most powerful cube. That doesn't make his cube better or worse (you misquoted me using those terms) but it does make it less powerful. If the cube manager argues that Crystal Shard is more powerful than Tinker in his cube, he would be right. If he argues that it is more powerful in cube in general, he would be wrong. And the latter is what is being asked in this project. And I'd be very interested in this cube manager's opinion on that, because he has added a layer of complexity to his cube that suggests an above-average experience level.
So your amalgamation logic is based on the false premise that a user would provide the cards that are the most powerful in HIS cube. If I'm playing a Legacy cube and am asked about the power level of blue cards, why would I not rank Ancestral Recall and Time Walk? Is my opinion invalid because I don't play those cards in my cube? I don't see why anyone would interpret the question so narrowly as to restrict his answers to cards that are most powerful in his cube specifically, even if this cube imposes design restrictions that reduce power level. But I'm sure that a disclaimer could be added to the original post to clarify the purpose of the project, if people think it is necessary.
Now let's assume we ask people for their favourite cards instead. Then the above cube manager may very well list Crystal Shard above Tinker. Let's assume he does. Only now do we have Team Shard and Team Tinker. Let's say team Shard manages to get their pet card listed. Me browsing the forum and seeing the thread on favourites cards sees it listed and decides to try it. I just shove it in without any context, as we don't have any. But my cube wasn't built to make Shard a great card, so after some bad experiences I cut it from my cube. What happens when I sit down for a cube with team Shard? First of all, this is another false premise - I will never sit down with team Shard: I've built my cube on power level, and have not sacrificed power to add a blink/ETB/other theme to the point where Shard would be more powerful than Tinker (by the way, it is my understanding that the overwhelming majority of cube managers on this forum build their cube on power level). Anyway; what happens is I thought it was a mediocre card from my experience and, perhaps surprised to see it in my pack, I pass the pack on without considering even for a second to pick it.
So picking favourites provides useless data even if the favourite is picked on cube-specific power level. What happens if favourites are picked for art, flavour or other things that do not translate into anything without context?
(Note on the above: I think Shard is actually a decent card and the above text paints it a bit negatively; for the sake of the argument what matters is that it is weaker than Tinker, which I think is not controversial. If it helps, think of a weaker card than Shard).
Lists based on power level measure an objective notion imperfectly. Lists based on favourites measure a subjective notion perfectly. It is my view as I have explained in my previous post that the first is clearly preferable. You prefer the second, which I think is fair. But you seem to be arguing that the second is objectively better, and I think you are very much wrong in that regard.
It appears the voting periods are largely agreeable so I'm going to go ahead and state that we will be doing M-F voting periods for each of the sections.
Guild voting will go 10 deep and split ally/enemy on different weeks. Shards/Wedges will only go 2 deep and will be all 10 on the same week.
We can go with the intrinsically best standard for color classification. So basically, I think we are mirroring 2016's classifications with a few notable changes:
1) Noble Hierarch is Bant again. (I'm not keen on the idea of having a card voted upon in multiple sections and the arguments here overwhelmingly favor classifying Hierarch as Bant).
2) DOJ and *** are grouped together this time.
3) Bomat Courier and Scrapheap Scrounger will be red and black respectively.
I'm still lacking a concise definition to use for our voting criteria. I've read through the comments and these are my thoughts:
@Karshtakavaar: I think you're right in that this project is somewhat specific to the MTGsalvation audience which is generally more interested in what is strongest and not about diversity when it means sacrificing power level. I don't want to disinclude anyone, but I also would expect someone with a top 20 of mostly budget or fun options may not see much, if any, of their list reflected in the top 20--though it would be on the spreadsheet. If you're okay with that then by all means give us your top 20! I think that is preferable to a lot guesswork as to where cards you don't play or have experience with might fit.
As we get more options cube diversity is expanding, but you're presenting power vs diversity as if they are an either/or choice and they aren't necessarily. We're lucky enough to be able to achieve both goals with greater ease now. Many cubes are upping their size due to the amount of powerful choices that have become available to us. In short, I don't think the power axis is being edged out so much as the number of options on this power level has increased. Archetype design is very important, but there's also a lot of powerful cards that don't fit cleanly into an archetype but are definitely worth cubing. I think a healthy cube has some clearly defined archetypes, but isn't beholden to them. We should see aggro, midrange, and control supported in every cube, but when a cube is running, say, Hardened Scales to support a +1/+1 counter theme in lieu of stronger generic cards, I personally feel that is a mistake.
@wtwlf: I don't agree with your statement about the 2016 definition using an "arbitrary/inconsistent powerlevel metric that nobody can possibly agree on". In fact, if anything there was a little bit too much agreement which is partially why I want to incorporate the use of spoiler tags this time. I strongly prefer less sentimental criteria such as "favorite" or "most loved" in favor of "most played" "most impactful" or "contributes most to a win". I want to keep hearing feedback and hammer down our definition for voting criteria in the next couple days. Thanks again for your feedback.
I also don't think we'll be able to come to a true unanimous agreement on the voting criteria. Even if you tell folks to post their top twenty "most impactful" white cards and everyone agrees on that, you'll still gets posters who interpret that differently. What does impactful really mean in terms of cube? Is Armageddon the white card that's had the most impact on games for me? Does that make it rank higher than Moat or Balance? Maybe I'll just shortcut impactful to mean P1P1 or favorite since these are easier for me to interpret. I don't believe there's really a way to avoid this metric being mostly arbitrary based on the individual posters.
MTGS Average Peasant Cube 2023 Edition
Follow me. I tweet.
I really have two favorite cards lists if we go that route - one as a player and one as a designer. The first includes cards that I like to play personally, and is biased and influenced heavily by the decks I love playing. The second factors more heavily issues like archetype support, playability and balance. For example as I player I value Searing Blaze over Abrade, but as a cube designer it is the opposite.
I suggest that if we do decide to use favorite cards as a criteria, we should be after the second type of list - favorite cards as a cube owner/manager/designer and not as a cube player/drafter. The second is a lot more valuable. Of course our biases as players affects our judgement as designers so it isn't a cut and dry separation. But at a cube manager's perspective we need to cater to many more needs - the likes of several players and the entire environment, which is less prone to outliers.
I still think P1P1 is the better criteria as it is well-defined. Of course it will be different from cube to cube, that is why we do this project. The individual vote doesn't matter that much, it's the aggregates that count.
The list on cube cobra
Read my blog on cube - Latest post June 2nd 2022
(BW, this is your project and you're going to be doing the leg work on it. You've already been generously solicitous of everyone's opinion, and we're all dead split on it so you're not going to find consensus. Just tell us what to do with Hierarch and let's all move on.)
I also think a P1P1 (or most powerful, or best) makes the most sense, especially since maximizing power level is the overriding philosophy on this forum. If people want to have an opportunity to highlight their favorite cards or whatever, just add a "Favorite Cards" section of like 5 or something at the end of the real rankings.
My Cube on Cube Tutor
If we want to do P1P1, we should do far less than 20 cards per section. How many white cards are honestly P1P1-quality cards? Like 5 maybe? Also, P1P1 value is split into two very separate camps. There are those that think that safe, universally playable cards make the best P1P1 choices, and there are those that think that snagging the most powerful card for the right deck is worth gambling on. As much as I like Path to Exile, I don't honestly know how awful my pack would need to be in order to P1P1 that card. Best is subjective. Most powerful is subjective. ...I just don't think those metrics provide useful data anymore. Cubes are too diverse and varying powerlevels are too contextual.
Am I seriously the only person that thinks that a list of peoples favorite cube cards wouldn't be more valuable? It can be the list of "the 20 best white cards to include in cubes" ...which is data everyone can accurately vote on, everyone's opinion is equally valid, and you can use whatever metrics make you excited to cube with certain cards. BlackWaltz3, this is YOUR project that you're championing. If you want to pick a metric and have people try to conform to it when voting, just do it. But I can't imagine a powerlevel definition being outlined that is both accurate and inclusive.
My 630 Card Powered Cube
My Article - "Cube Design Philosophy"
My Article - "Mana Short: A study in limited resource management."
My 50th Set (P)review - Discusses my top 20 Cube cards from OTJ!
MTGS Average Peasant Cube 2023 Edition
Follow me. I tweet.
Also, if Noble Hierarch is Bant, then the Mox should probably be in there respective colors and not in colorless? This would also make an argument that there could possibly be a [mana]WURBGa] section, and Black Lotus, City of Brass and the like would go there since they are intrinsically best in 5 color decks. (not arguing that we should have a 5 color section, but simply pointing out the flaw in using the "intrinsically best" isn't quite the line that makes sense to me.
http://www.cubetutor.com/cubeblog/63569
I mean, "favorite cards" or whatever is way more subjective, since by definition it's nothing but personal opinion. At least P1P1 has some basis in objectivity w/r/t perceived power level.
More to the point, I don't know what the point would be of the project you're proposing. Like, if it's just "I want everyone to know about this cool card", that's fine, but why not just have a big thread about people's pet cards? Or hell, just make an SCD about it. God knows this place could use some more activity outside of spoiler season. But what difference is it to anyone that Parallax Wave and Porphyry Nodes are two of my favorite cards? Who does that help? And why do we need to have a vote on it?
Anyway, if P1P1 is such a difficult thing for people to grok (I really don't think it is, but whatever), it's not hard to come up with a different way of thinking about it. Something like this:
Edit: Calibretto said it better. Would I take Card A over Card B, all else equal.
My Cube on Cube Tutor
Correct. But if the goal is obtaining good data, we can get that with asking people their opinions on their favorite cube cards. It's better to ask a subjective question with the goal of obtaining data that can genuinely be answered by it than to ask subjective questions if the goal of obtaining objectively useful data. The voting sample size won't be even remotely large enough for the variance to be ironed out by averages. Not by thousands and thousands of voters. But with peoples favorite cards, the information is honest and equal (and honestly? about a million times more valuable).
Then we're asking the wrong questions. Because if that's making a list of "top P1P1 cards", the data we're getting is pretty meaningless.
Me too. Everybody can vote, it doesn't matter what metric(s) you use to determine which cards make the list for you, and it doesn't matter what kind of cube you prefer to draft. I'd rather do this and err on the side of inclusivity.
My 630 Card Powered Cube
My Article - "Cube Design Philosophy"
My Article - "Mana Short: A study in limited resource management."
My 50th Set (P)review - Discusses my top 20 Cube cards from OTJ!
I agree with the notion that an agglomeration of favorite cards is not very useful. I also don't think it would lead to more sensical results necessarily - if Goblin Welder is the only artifact matters card in the list, it would be ill-advised to jam it without support. You still need to know the context of the cards on the list, and it will still be subjective, there is no getting around this issue.
The list on cube cobra
Read my blog on cube - Latest post June 2nd 2022
My 630 Card Powered Cube
My Article - "Cube Design Philosophy"
My Article - "Mana Short: A study in limited resource management."
My 50th Set (P)review - Discusses my top 20 Cube cards from OTJ!
Agreed. That was my point. It's rare that I'd ever take Path at P1P1, so if it's on a list of the top twenty P1P1 white cards it raises a lot of questions about how useful the data actually is.
Why not just say "These are the first twenty cards I'd include in X section in order of importance"? These lists should be here to support new cubers looking for a resource when building their cubes.
MTGS Average Peasant Cube 2023 Edition
Follow me. I tweet.
As far as voting criteria, my inclination is to have not just a one factor definition such as "best" "p1p1" or something like "favorite" but rather a short series of factors (maybe two-three sentences) to consider as guideposts for people's lists. I'm thinking the most important measures are "how critical this card is to winning (not upon resolution)", "how often would you main-deck this card", and "most important to my cube". I don't claim these aren't subjective, but they aren't as widespanning as favorite/most loved or as narrow or uninclusive as "p1p1" or "best" I think they are good middle-ground guideposts. It improves upon the definitions used last time, but doesn't change them so radically that this can no longer be validly compared to our past results.
Noble Hierarch is going to be whatever there are the most comments in support of. I started with Green and then all the Pro-Bant people commented, so I said "Fine, Bant" and then the pro-green people chimed in....lol.
I count myself, Hoodwink, and allred for team Green and a few others have commented for team Bant. I'll just let the numbers tell me what to do and spare me the headache.
Okay now some user specific feedback:
@calibretto: PM's probably would accomplish my goal of limiting influence from others lists, but I am NOT going through a bunch of PM's to compile the data! Hehe. I don't want people to feel they shouldn't look at others lists, but I will encourage people to try to create their lists independently. Spoiler tags may help even if they are an imperfect solution--at the very least if there is an early trend among the first voters, people don't have to see it.
@ Metamind: Ordering is necessary (also makes it a lot more fun for people). Try not to think of it as you have but by the "Would I take Card A over Card B, all else equal" presented earlier in the thread.
I realize these solutions may not be pleasing to all, but I have to do the best I can and then get on with it. I'm really excited to see what out results look like 3 years later. I know a lot of people loved this project then and I'm hoping it will be just as fun and interesting today. I'm definitely going to include fum summaries of the results as Spike Rogue did because that was one of my favorite parts and I'm sure others loved that too.
Keep the feedback coming! Cheers =D
On the format, count me, squarely, in the power level camp. Power means different things to different people, yet we all follow wtwlf’s new set reviews religiously, we go to Cubetutor to see draft percentages, and we go to SCD threads to see other people’s views on the playability of cards. A debate can be had on which card is more powerful between Isamaru, Hound of Konda, which is played in any white aggro deck but is not especially impactful, and Monastery Mentor, which sees less play but has a far bigger impact. But anyone trying to tell me that Crystal Shard is more powerful than Tinker will have a very hard time of it and save for the means to prove it, that person would be objectively wrong. Power level is, to a very large extent, an objective notion. That it is difficult to define does not make it any less so.
Rankings cards by how much you like them is a different beast altogether. What makes people like a card? Art? Flavour text? The kind of game states it creates? How much skill it takes? How much it contributes to balance in your cube? An experience you’ve had with it in the past? I care a lot more about random strangers’ professional opinion (if I believe they are qualified – as I do in this case) than I do their personal opinion. For this reason I personally have zero interest in a favourites list and would not participate (on that note, I object to the notion that such a format would be more inclusive). That is my personal opinion though and I can imagine that other people, justifiably, may disagree.
However, there is a second, objective, problem with a favourites list. How can I possibly know why you specifically like a particular card? Maybe you like it because you think it is powerful, which is a valuable opinion to me. Maybe you like it because the card takes a lot of skill to play right, which is sort-of valuable to me. Maybe you like the card because when you were a kid you built a deck around it and it’s a nostalgia monster. This last fact is not valuable to me at all. The problem is that because I don’t know which of the three (or other) scenarios applies, I can’t interpret the result in any meaningful way. Even if you were ranking the cards by power level, your list would be 100% useless to me.
In my personal case, I love Sakura-Tribe Elder because I cracked up my cousin once calling it STEve, and I think it’s a funny card because of that. Blazing Archon is one of my favourite white cards because it has amazing flavour. And I love Daze because I played with it for years in a different format. How is any of this useful to anyone? Especially if you don’t have this background information!
As for how to define power level, I think any number of criteria can apply. Let’s not get hung up on the fact that few cards are “P1P1” worthy, as if for that reason Swords to Plowshares would be as powerful as Path to Exile. Here are some possibilities:
• The priority within the colour in which you would generally draft the card in a first pack
• The likelihood of you playing the card in an on-colour deck in a sealed format
• The size of a power level-based cube where you would start including the card
• The impact the card will have on a game – e.g. how much it contributes to your odds of winning
• Your gut feeling, which will be a combination of the above. I don’t need a definition to know that Dark Confidant is more powerful than Asylum Visitor.
I’m not even sure if it matters so much to converge on a specific definition of power level. I’m sure I personally give above-average weight to the second category above, and I think for example that Isamaru, Hound of Konda is a pretty powerful card. Does that make me wrong? Why should we not be able to compare my opinion alongside someone who attaches more importance to the first criterion, and who would go rather for a card like Monastery Mentor? I’m not sure any one definition of power level is the correct one. Power is power. (Side note, I think Mentor is more powerful than the Hound!)
I will also throw in my 2 cents on the debate for classifying Noble Hierarch, although I really don’t care where you end up going with this. I care only for sport. I would generally classify it as green, and not Bant. The argument that you should classify a card in the colours where it is strongest is strange to me. If this was my approach I would classify Wild Mongrel as B/G, Sphinx of the Steel Wind as mono U and Bonesplitter as R/W. I think mana cost plays a big role, with a nod to the colours that correlate strongly with where the card sees play. You will play Hierarch in Bant, and it will shine, but it costs G and you will also play it in any Gx deck. G is the common factor, not B/U/G. You won’t play Kird Ape in anything other than an R/G deck, so it belongs to that guild. Scrapheap Scrounger should only see play in Bx decks, so it goes to black. And so on.
Lastly, a quick note on spoiler tags. I think this is good practice. Anchoring is an extremely powerful bias and I would anyway approach something like this by actively ignoring previous posts. There’s very few utility lost in having to click open such tags. Not a fan of private submissions, because that process fails to generate hype and community involvement. Plus, once you’ve finished your submission it’s great fun to look at all the other ones.
The problem with this prompt is curve. A 20-card section wouldn't contain the 20 "best" white cards.
This would be exactly true for a cube that supports a blink/bounce/ETB trigger abuse deck and has no support to make Tinker powerful. Those kinds of cubes exist. Now, you're telling voters that they're objectively wrong because their notion of "best" isn't the same as yours ...when in fact they're 100% accurate and the only thing that changed was context.
The bigger problem is assuming that an amalgamation of the data is useful, when in fact, the opposite is true. If you combine the data from the Team Tinker's votes and the votes from team Shard into one dataset, and team Tinker has 20x the voting members ...what happens when you sit down to draft a cube with team Shard? The data is useless to you. And that's the problem. Cube powerlevel is too context-based to provide meaningful data with a small dataset. 25 people voting or whatever isn't anywhere near statistically significant enough to provide meaningful data. But if you ask a question that everyone can honestly answer using the same metrics, your dataset can be smaller and still produce worthwhile data.
tl;dr - If you ask the question "what are your 20 favorite white cube cards" people can look at that combined data and get a feel of what kinds of cards they could include in their cubes to make people happy. If you ask the question "what are the 20 best white cube cards" you would have to qualify that information with about 60 disclaimers before everybody's even voting with the same metrics.
Not strongest. Intrinsically best. There's no intrinsic color identifiers that make Wild Mongrel anything but green. Your Golgari classification would be contextual organization, not intrinsic organization.
Organizing cards where they're intrinsically best is the reason why Vedalkan Shackles is blue for most cube managers and not colorless.
My 630 Card Powered Cube
My Article - "Cube Design Philosophy"
My Article - "Mana Short: A study in limited resource management."
My 50th Set (P)review - Discusses my top 20 Cube cards from OTJ!
Echoing this because I agree with all of it (that Wild Mongrel point is making me reconsider my classification strategy), and also to wholeheartedly agree with the use of spoiler tags. Anchoring is a real thing, people!
My Cube on Cube Tutor
Rank your top 20 cards in terms of:
- How critical this card is to winning.
- How frequently you would maindeck this card in a deck playing this color/guild/shard/wedge (or just in general if colorless).
- What cards are most important within your cube.
The 1st two I believe are more power-level focused metrics, whereas the last measure encourages participation by those who may not share this general cube design philosophy. I really just can't justify the word favorite because I think it leads to connotations that could certainly produce wildly different and largely unhelpful results that we can't use as a standard of comparison to previous similar ranking projects. Also, I think the third definition is somewhat similar to favorite, but doesn't carry the connotations that I'm concerned about. Hopefully, that is an agreeable compromise.
Spoiler tags will be a thing. At least, I'll be asking that everyone try to do this for the sake of avoiding anchoring bias and encouraging personal reflection rather than following any trends.
Thank you guys, this has been a thought-provoking discussion already and we haven't even gotten going yet.
http://www.cubetutor.com/cubeblog/63569
2. Card B ....
20. Card C
I would recommend something along the lines of:
I recommend not having people use numbers in and just the name of the card, and ask each voter to make sure the image pop's up for the card if they move mouse over (saves issues with slight differences in spellings in cards like Jace, the Mind Sculptor where there are commas, or other cards with ' etc.
http://www.cubetutor.com/cubeblog/63569
I mean, yeah, but how many people are building cubes with 20 card sections? If I were building a 360 card cube with 50 per section, the first 20 cards I'd include would likely be the same first 20 cards I'd include if I were building a 540 card cube with 70 cards per section.
But I feel like we're all just arguing semantics here. Ultimately, while I do find the data we get from these top 20 votes to be interesting to read through, I don't think it has much value in terms of the lists of cards we end up with. I don't mean to be rude if that comes off that way.
The type of data that I think is more valuable is either a top X for each casting cost in each color or just a cube comparison that results in how many cubes run specific cards. Looking a list of the top X White 5 drops is helpful for new cubers looking to get into the format and for veteran cubers who may just be looking for something fresh to swap in over Baneslayer. Alternatively, a cube comparison that tells me 30 cubers are running a card that I've been overlooking is helpful for me to find new cards for my cube and also helpful for new cubers looking for a starting point for a new cube. Both of these data finding methods are better, imo, than a Top 20 for each color. It's hard to agree on a metric and then get everyone voting to stick to that metric. And, who is this data for? It's unlikely that I'll find something new for my own cube on these lists. If a new cuber looks at this and just says, Oh I need to include these twenty cards, that doesn't give them any sort of context for where they need to go from there.
MTGS Average Peasant Cube 2023 Edition
Follow me. I tweet.
Allred, I think I'd prefer people keep the numbers just to ensure they didn't skip a number, list the wrong amount of cards, etc. Might make a little more work for me, but I'll deal.
I have to say I haven't seen many cubes where Crystal Shard is more powerful than Tinker. But that's not really the point.
You can very well build a cube where Shard is more powerful than Tinker, and such cubes may very well exist. But any cube manager behind such a cube must surely be aware of the fact that his design philosophy has not been to create the most powerful cube. That doesn't make his cube better or worse (you misquoted me using those terms) but it does make it less powerful. If the cube manager argues that Crystal Shard is more powerful than Tinker in his cube, he would be right. If he argues that it is more powerful in cube in general, he would be wrong. And the latter is what is being asked in this project. And I'd be very interested in this cube manager's opinion on that, because he has added a layer of complexity to his cube that suggests an above-average experience level.
So your amalgamation logic is based on the false premise that a user would provide the cards that are the most powerful in HIS cube. If I'm playing a Legacy cube and am asked about the power level of blue cards, why would I not rank Ancestral Recall and Time Walk? Is my opinion invalid because I don't play those cards in my cube? I don't see why anyone would interpret the question so narrowly as to restrict his answers to cards that are most powerful in his cube specifically, even if this cube imposes design restrictions that reduce power level. But I'm sure that a disclaimer could be added to the original post to clarify the purpose of the project, if people think it is necessary.
Now let's assume we ask people for their favourite cards instead. Then the above cube manager may very well list Crystal Shard above Tinker. Let's assume he does. Only now do we have Team Shard and Team Tinker. Let's say team Shard manages to get their pet card listed. Me browsing the forum and seeing the thread on favourites cards sees it listed and decides to try it. I just shove it in without any context, as we don't have any. But my cube wasn't built to make Shard a great card, so after some bad experiences I cut it from my cube. What happens when I sit down for a cube with team Shard? First of all, this is another false premise - I will never sit down with team Shard: I've built my cube on power level, and have not sacrificed power to add a blink/ETB/other theme to the point where Shard would be more powerful than Tinker (by the way, it is my understanding that the overwhelming majority of cube managers on this forum build their cube on power level). Anyway; what happens is I thought it was a mediocre card from my experience and, perhaps surprised to see it in my pack, I pass the pack on without considering even for a second to pick it.
So picking favourites provides useless data even if the favourite is picked on cube-specific power level. What happens if favourites are picked for art, flavour or other things that do not translate into anything without context?
(Note on the above: I think Shard is actually a decent card and the above text paints it a bit negatively; for the sake of the argument what matters is that it is weaker than Tinker, which I think is not controversial. If it helps, think of a weaker card than Shard).
Lists based on power level measure an objective notion imperfectly. Lists based on favourites measure a subjective notion perfectly. It is my view as I have explained in my previous post that the first is clearly preferable. You prefer the second, which I think is fair. But you seem to be arguing that the second is objectively better, and I think you are very much wrong in that regard.