Haven't voted yet, but I was planning to put it in the teens. There are a number of black cards that are powerful in more decks that I like more, but you're absolutely right that there is a massive difference between an entomb reanimation deck and one w/o. Like, there are few plays scarier than EOT entomb if you can't answer something large and in charge.
I don't even run Entomb in my cube. We did run it for a while, but it rode a lot of sideboards and felt a little too narrow. It's a solid reanimator enabler and it's cute with flashback spells, but I don't feel like reanimator needs it to be competitive and the other applications felt kind of niche.
If the metric is straightforward, then you should be able to state it in more precise terms so that people are not interpreting it differently. (Getting everyone to try to adhere to it may be a different matter.)
As far as I can tell, there are three possible metrics we could be evaluating: (a) increase in match win% conditioned on seeing the card in a pack, (b) increase in match win% conditioned on picking the card, or (c) how much the card feels like it carries its weight.
Combining two or more of those metrics does not make sense as you don't end up with something meaningful. Metric (c) is not very useful and I hope that's not what we're talking about. In the case of Lore Seeker, it might crack top 20 on metric (b), and it's probably the worst card in a cube on metric (a). If you combine those and say Lore Seeker is kind of average, when there is no sense in which Lore Seeker is an average card, you're just not saying anything meaningful.
Most cards do nothing to affect your win % if you don't pick them, so in nearly every case you can evaluate the card entirely based on (a). Picking Lore Seeker turns a later pick in the draft into an additional first pick, which more often than not will increase your win %, I'm not sure why you'd call it the worst card in cube on that basis. If I understand you correctly, it sounds like you're suggesting that the drafters who don't pick Lore Seeker benefit more from the extra cards than the person who picks it and gets that extra first pick, so it's the worst possible card you can pick. I think the value of getting that extra first pick more than makes up for that, and I tended to pick Lore Seeker fairly highly when I ran it in my cube, but yeah it sure was nice to get 2nd pick from that extra pack for free when it happened. All the same, I plan to evaluate it based on how good it is when picked, just like any other card. Frankly, I don't expect Lore Seeker to crack the Top 20 of many people's lists including my own, so I feel like I've already spent more time evaluating it for the purposes of this project than necessary.
Are there other cards that you're having difficulty evaluating in the context of how effectively it contributes to game wins, or is Lore Seeker the only one?
On the Entomb discussion:
Hopefully this is the right thread for this?
Feel free. Once the vote is over, though, it's probably better to discuss the vote on any given category in its Results thread.
Private Mod Note
():
Rollback Post to RevisionRollBack
465 card Unpowered cube thread. Draft it here and I'll be happy to return the favor.
450 card Peasant cube thread. Draft it here.
But... wanted to air my opinion on the black voting. People are comically underrating entomb. I've seen no person rate in their top 20, which is absurd in my eyes.
The difference in power level between a reanimator deck that has entomb, and one that doesn't is night and day.
Reanimator is effectively a 3 peice combo deck that requires the following:
An expensive creature in your hand to be able to discard to your graveyard
A way to get the fatty from your hand into your graveyard
A reanimation spell
Entomb allows you to merge 2 of those steps into 1, tutoring for ANY creature in your deck (some decks require different threats) that's not in your hand for the cost of 1 mana.
There's no on color card that comes CLOSE to that. Speed and consistency is critical for combo decks, and entomb is both the fastest and the most consistent way to set up the combo (by a significant margin).
Very early in my duplicate cube I tested 2 copies of entomb (underrating its power), and it was IMMEDIATELY apparent how broken reanimator became when I did. Much more so than a second copy of animate dead/reanimate.
I also had the same thought when I was first browsing through everybody's vote (I do a lot of browsing / posting at work but don't do big posts like that since I'm on my phone).
My analysis of Lore Seeker mostly comes down to this:
If you're deciding whether to pick Lore Seeker or X, assuming the player after you will take Lore Seeker if you don't, your choices is between (a) X + an 8th pick or (b) colorless bear + a 1st pick + a 9th pick. Given that the difference between an 8th pick and a 9th pick is negligible and the bear is not likely to be played, the choice is between X and a random 1st pick (where, recall, X is the best of the other 14 cards). For purposes of evaluating win% conditioned on picking the card, that means LS is about as good as an average first pick (including on-color cards). That might put it into top 20 colorless by that metric.
To evaluate win% conditioned on seeing the card, we have to think about a few things. First of all, what is the effect when you see a LS, compared to seeing a blank? If you see it pick 1-2 and take it, you upgraded a second pick or a below average first pick to an average first pick. If you see it and pass it, it probably gave you an extra 8th pick. Lore Seeker should pick 1st picked just about 50% of the time, and probably second picked close to 100% of the time, so we'll use that as an estimate. On the other hand, what is the effect when you don't see LS? Either it's not in the draft (about 1/3 chance given an 8-man 540), or it is. If it is in the draft and you don't see it, it gives you an extra 2nd-8th pick at no cost (8th pick being less likely).
Skipping a bit of math, we're comparing:
(a) 2/3 upgrading a second pick or below average first pick (equal chance) to an average first pick + 1/3 getting an 8th pick
against
(b) 2/3 chance of a free 2nd-8th pick, with 8th pick being less likely.
And yes, I think (b) is better.
That means your expected win% actually goes down for having seen LS, which would make it the worst card by that metric. If you think (a) is better, it's still a much closer comparison than for other cards, which would put it close to the bottom.
I haven't thought as much about how other cards would be measured because LS is the most interesting in that respect, and I thought hard about this when deciding whether to include LS in my cube. There are some similar considerations for Paliano Vanguard and other cards that may reveals colors during the draft, as both you and your neighbors benefit from improved signaling, but I think those considerations are going to be secondary and/or end up neutral, especially since you come out ahead by seeing or taking the card.
Just a heads up, I have been slammed at work since the blue thread closed, but I am working on it.
I hope to have the results thread up by tomorrow night
Entomb didn't make my list either. It's by far the most narrow and high variance of all reanimator cards. It's great when you're playing a reanimator deck, but even then it can be a blank card (when you have no reanimation spells), and it's a bad top deck. There're plenty of discard outlets to enable the deck, and those cards are never dead. I don't think about reanimator as pure combo decks, rather try to combo early or get good value anyway, where both cards like Reanimate and Wild Mongrel are valueable and playable on their own.
I didn't know that post #2 in this thread had all the voting results compiled in it. Could a mod sticky this thread (and maybe have "and results" added to the thread name) so it can all be in one convenient and easy-to-find location? Thanks!
I find that cards like Smokestack, Tangle Wire, Braids, etc. that require breaking a symmetrical effect are naturally trumped by enough of the decks in cube that they often underperform for us. Braids is the best of them, but even then I've seen Braids be worse for the owner enough in a deck/board well suited for the Braids to go off that it's hard for me to vote on them. (Although I did vote on Braids, because it has brought enough people to frown town to make the top 20. Very few cards punish people on the draw or starts that might stumble than Braids does.) Extra points go against Smokestack since your opponent has a full turn to prepare for it, it's such a slow engine. I understand that YMMV, but the cards haven't done it for us in the many years we have run them. (We took out Smokestack a couple years ago because it was horrendously bad for us. Tangle wire still makes some decks, but less than before.)
Relic has been a fixing Worn Powerstone, it's been incredible for us. It's great for supporting a splash, and it acts as a key cog in any 2+ colored deck.
Tangle Wire is about as symmetric as the war between the US and ISIS.
I kinda agree on Smokestack being not that strong, as it is quite difficult to make it work (but it's awesome when it works). Maybe I rated it a bit too high, because I just like the archetype.
Coalition Relic is just great. A mana rock for 3 that comes into play untapped and makes 1-2 colored mana every turn is gud.
Maybe some kind of 11th hour scandal will crop up involving fraudulent donations to the Library of Alexandria or the environmental destruction caused by Strip Mine and Wasteland to drive people to the polls to vote against them.
I made some box and whiskers plots with R from the raw data that you provided. I think they're pretty neat.
Since the central tendency measure for box plots is the median, I ordered/ranked the cards by median instead of mean, using the mean only as a tie-breaker for cards with the same median.
Using the median reduces the influence of outliers on the ranking.
For example, in the colorless results:
Mana Vault has a median of 17 and a mean of 14,652.
Jitte has a median of 16 and a mean of 16,130.
The difference between those two cards is that Jitte received votes from everybody, in the range 13 to 19, with most votes around 16.
Whereas Mana Vault received better votes from the majority of voters, in the range 15 to 18, with most votes around 17, but there are 2 voters who did not vote for Mana Vault at all, on one voter who ranked it last.
Another example where using the median makes a difference is at the bottom of the colorless ranking:
Sword of Light and Shadow has a median of 2, and a mean of 2,348.
Crucible has a median of 0, and a mean of 2,479.
Here, a majority of voters did vote for Sword of LaS (with a low rating), whereas a majority did not vote for Crucible at all, but those who did vote for Crucible ranking it higher.
Can someone explain why Horizon Canopy is voted on here?
Horizon Canopy has "Land" in the type-line, it makes more than 1 color of mana, it's Vintage legal, and it contains no colored mana costs, so I'm not sure where else it might be categorized.
Horizon Canopy is a GW dual land, but not part of a cycle. It therefore feels a bit out of place in the general land vote and seems to belong in the Selesnya section of the guild card vote. It kinda bothered me, too.
(I wish we had a full cycle of this type of dual land. That would certainly help categorizing them. It would also be pretty awesome for cube.)
Uril, the Miststalker RGW -- Ulamog, the Infinite Gyre C -- Vhati il-Dal BG -- Jor Kadeen, the Prevailer RW -- Animar, Soul of Elements URG
Kiki-Jiki, Mirror Breaker R -- Maga, Traitor to Mortals B -- Ghave, Guru of Spores BGW -- Sliver Hivelord WUBRG
Also, follow us on twitter! @TurnOneMagic
My Tribal cube
My 93/94 old school cube
My Artifact cube
My Hearthstone Quiz App for iOS
MTGS Average Peasant Cube 2023 Edition
Follow me. I tweet.
Most cards do nothing to affect your win % if you don't pick them, so in nearly every case you can evaluate the card entirely based on (a). Picking Lore Seeker turns a later pick in the draft into an additional first pick, which more often than not will increase your win %, I'm not sure why you'd call it the worst card in cube on that basis. If I understand you correctly, it sounds like you're suggesting that the drafters who don't pick Lore Seeker benefit more from the extra cards than the person who picks it and gets that extra first pick, so it's the worst possible card you can pick. I think the value of getting that extra first pick more than makes up for that, and I tended to pick Lore Seeker fairly highly when I ran it in my cube, but yeah it sure was nice to get 2nd pick from that extra pack for free when it happened. All the same, I plan to evaluate it based on how good it is when picked, just like any other card. Frankly, I don't expect Lore Seeker to crack the Top 20 of many people's lists including my own, so I feel like I've already spent more time evaluating it for the purposes of this project than necessary.
Are there other cards that you're having difficulty evaluating in the context of how effectively it contributes to game wins, or is Lore Seeker the only one?
On the Entomb discussion:
Feel free. Once the vote is over, though, it's probably better to discuss the vote on any given category in its Results thread.
450 card Peasant cube thread. Draft it here.
I also had the same thought when I was first browsing through everybody's vote (I do a lot of browsing / posting at work but don't do big posts like that since I'm on my phone).
My High Octane Unpowered Cube on CubeCobra
If you're deciding whether to pick Lore Seeker or X, assuming the player after you will take Lore Seeker if you don't, your choices is between (a) X + an 8th pick or (b) colorless bear + a 1st pick + a 9th pick. Given that the difference between an 8th pick and a 9th pick is negligible and the bear is not likely to be played, the choice is between X and a random 1st pick (where, recall, X is the best of the other 14 cards). For purposes of evaluating win% conditioned on picking the card, that means LS is about as good as an average first pick (including on-color cards). That might put it into top 20 colorless by that metric.
To evaluate win% conditioned on seeing the card, we have to think about a few things. First of all, what is the effect when you see a LS, compared to seeing a blank? If you see it pick 1-2 and take it, you upgraded a second pick or a below average first pick to an average first pick. If you see it and pass it, it probably gave you an extra 8th pick. Lore Seeker should pick 1st picked just about 50% of the time, and probably second picked close to 100% of the time, so we'll use that as an estimate. On the other hand, what is the effect when you don't see LS? Either it's not in the draft (about 1/3 chance given an 8-man 540), or it is. If it is in the draft and you don't see it, it gives you an extra 2nd-8th pick at no cost (8th pick being less likely).
Skipping a bit of math, we're comparing:
(a) 2/3 upgrading a second pick or below average first pick (equal chance) to an average first pick + 1/3 getting an 8th pick
against
(b) 2/3 chance of a free 2nd-8th pick, with 8th pick being less likely.
And yes, I think (b) is better.
That means your expected win% actually goes down for having seen LS, which would make it the worst card by that metric. If you think (a) is better, it's still a much closer comparison than for other cards, which would put it close to the bottom.
I haven't thought as much about how other cards would be measured because LS is the most interesting in that respect, and I thought hard about this when deciding whether to include LS in my cube. There are some similar considerations for Paliano Vanguard and other cards that may reveals colors during the draft, as both you and your neighbors benefit from improved signaling, but I think those considerations are going to be secondary and/or end up neutral, especially since you come out ahead by seeing or taking the card.
I hope to have the results thread up by tomorrow night
http://www.cubetutor.com/cubeblog/63569
450 card Peasant cube thread. Draft it here.
The results of Straw Poll #3 are also up:
The community has voted 24-4 to vote on Wildfire and Burning of Xinye as 1 card in this year's Power Rankings.
The community has also voted 19-7 to vote on Incinerate and Lightning Strike/Searing Spear as 1 card in this year's Power Rankings.
450 card Peasant cube thread. Draft it here.
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!
450 card Peasant cube thread. Draft it here.
Cheers,
rant
My Cube
CubeCobra: https://cubecobra.com/cube/overview/5f5d0310ed602310515d4c32
Cube Tutor: http://cubetutor.com/viewcube/1963
http://www.cubetutor.com/cubeblog/63569
I find that cards like Smokestack, Tangle Wire, Braids, etc. that require breaking a symmetrical effect are naturally trumped by enough of the decks in cube that they often underperform for us. Braids is the best of them, but even then I've seen Braids be worse for the owner enough in a deck/board well suited for the Braids to go off that it's hard for me to vote on them. (Although I did vote on Braids, because it has brought enough people to frown town to make the top 20. Very few cards punish people on the draw or starts that might stumble than Braids does.) Extra points go against Smokestack since your opponent has a full turn to prepare for it, it's such a slow engine. I understand that YMMV, but the cards haven't done it for us in the many years we have run them. (We took out Smokestack a couple years ago because it was horrendously bad for us. Tangle wire still makes some decks, but less than before.)
Relic has been a fixing Worn Powerstone, it's been incredible for us. It's great for supporting a splash, and it acts as a key cog in any 2+ colored deck.
Also, follow us on twitter! @TurnOneMagic
I kinda agree on Smokestack being not that strong, as it is quite difficult to make it work (but it's awesome when it works). Maybe I rated it a bit too high, because I just like the archetype.
Coalition Relic is just great. A mana rock for 3 that comes into play untapped and makes 1-2 colored mana every turn is gud.
Smokestack is a deck centerpiece, and a great grindy engine. Love it. It's hugely powerful in the right decks, and we've had tremendous success.
I think both cards belong in the top 20 for sure.
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!
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!
450 card Peasant cube thread. Draft it here.
My High Octane Unpowered Cube on CubeCobra
450 card Peasant cube thread. Draft it here.
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!
Since the central tendency measure for box plots is the median, I ordered/ranked the cards by median instead of mean, using the mean only as a tie-breaker for cards with the same median.
Using the median reduces the influence of outliers on the ranking.
For example, in the colorless results:
Mana Vault has a median of 17 and a mean of 14,652.
Jitte has a median of 16 and a mean of 16,130.
The difference between those two cards is that Jitte received votes from everybody, in the range 13 to 19, with most votes around 16.
Whereas Mana Vault received better votes from the majority of voters, in the range 15 to 18, with most votes around 17, but there are 2 voters who did not vote for Mana Vault at all, on one voter who ranked it last.
Another example where using the median makes a difference is at the bottom of the colorless ranking:
Sword of Light and Shadow has a median of 2, and a mean of 2,348.
Crucible has a median of 0, and a mean of 2,479.
Here, a majority of voters did vote for Sword of LaS (with a low rating), whereas a majority did not vote for Crucible at all, but those who did vote for Crucible ranking it higher.
Horizon Canopy is a GW dual land, but not part of a cycle. It therefore feels a bit out of place in the general land vote and seems to belong in the Selesnya section of the guild card vote. It kinda bothered me, too.
(I wish we had a full cycle of this type of dual land. That would certainly help categorizing them. It would also be pretty awesome for cube.)
Uril, the Miststalker RGW -- Ulamog, the Infinite Gyre C -- Vhati il-Dal BG -- Jor Kadeen, the Prevailer RW -- Animar, Soul of Elements URG
Kiki-Jiki, Mirror Breaker R -- Maga, Traitor to Mortals B -- Ghave, Guru of Spores BGW -- Sliver Hivelord WUBRG