Pattern Recognition #56 - Fetches
Features Opinion Pattern Recognition
berryjon
15 February 2018
2687 views
15 February 2018
2687 views
Hello everyone! My name is berryjon, and I am TappedOut.net's resident Old Fogey and part time Smart Ass. I write this series, Pattern Recognition, as a means to entertain, educate and something else that starts with an E. Apparently, that's supposed to be enlightening, but I'm not sold on that. Oh well, I'll see what's next in my thesaurus for next week.
Today's subject has been one that has been debated endlessly, and I assure you, my little spiel here will probably not change people's minds, save those undecided.
So, let me get straight to the point.
Fetch lands don't work.
Or rather, they don't do what a lot of players assume they do.
When people play Fetch Lands, there is, I have found, a rather large segment of the player base that concludes that taking a land out of their deck is a good thing.
And it is! However, the method by which they are going about doing it is not.
Before I go any further, I would like to read the following article by Garret Johnson, in which he goes through the math of how effective the Fetch Lands really are.
So, while some people are reading that, let me put the problem and its perceived solution into simpler language.
How many of you have drawn a card, only for it to be a land, and you don't need any more lands? Show of hands? Everyone? Good! Except you in the back. You with your hand down. You're either a liar, or you've never played the game before.
What a Fetch Land does, is it takes one additional land out of your deck. So you get not only the Fetch Land you removed by drawing it, but also one additional land taken away! That's one less land that you may or may not have drawn in the future, gone!
Let me include a graphic here, to better demonstrate what I'm talking about. In this example, every "L" is a Land, and ever C is, well, every other card.
C|C|L|C|C|L|C|C|L|C|C|L|C|C|L|C|C|L|C|C|L|
Now, in this case, there are 21 cards left in the deck, with seven of them being lands. Now, without knowing where in the deck those lands are, you would have about a one in three chance of drawing a land. This is simple statistics, and beyond a doubt. Now, what happens if you were to Fetch a land? Well, in theory, your deck would now look like this, after shuffling:
|C|C|L|C|C|C|L|C|C|L|C|C|L|C|C|C|L|C|C|L|
Now, this is a rough approximation, as the odds of the top card in your library move from, in this case, from 7/21 to 6/20, a change from 30% to 33.33%.
Yes, you read that correctly. Here is the first erroneous assumption made when people fetch a land. You won't always be improving the odds of making the next card not be a land. This becomes more and more pronounced as you get fewer and fewer cards in your deck. When there are more cards - say, moving from 20 Lands out of 50 cards to 19 lands out of 49, you go from 40% to 38.78%!
The exact median where you start to not help your chances is something I can't tell you - it wholly depends on the number of fetchable lands in your deck, and how many cards are left in the deck at the time.
Another problem, one that might not seem to be there, but it is one, is the paid price in life. Now, let's compare a Fetchland with a Shockland and a Painland.
I choose these three types because they are the most recognizable (I hope), and they all share a similar costs for their utility. And I realized that I haven't actually defined what a Fetch Land is yet, and I'm, like, halfway through the article! Let's fix that.
A Fetch Land is a land that, when activated, allows you to search your library for a land card that meets certain criteria, and to put it into play. However, while there are lands like Thawing Glaciers, for the purposes of this article, I have and will be defining a Fetch Land as one of the cycle that first appeared in Odyssey and was finished in the Zendikar block - where for the cost of one life, and to sacrifice the Fetchland, you could search your library for a land with one of two chosen basic land types, and put it into play tapped. For example, Arid Mesa. Now, this price is required. It's not an option. You want your land? Pay the life.
A Pain Land is a land that requires that you take a point of damage, or lose a point of life for each activation. City of Brass is one of the most iconic, but there are versions like Battlefield Forge which allow you to put into your mana pool without paying a cost.
Lastly, a Shock Land is a land that comes into play tapped, unless you pay a cost, namely two life, and it comes into play untapped instead. Sacred Foundry is an example of this, having started out in the Ravnica blocks. They are names as such because playing them untapped is like casting a Shock on yourself for free.
The question between these two becomes a matter of preference. Do you accept the two life loss as inevitable? Play a Fetch. Want the option to lose life, or take your time? Play a Shock land. Want to gamble that you can find (or already have) enough coloured mana? Go for a Painland.
Now, you may have noticed that most Fetch lands don't specify that they have to go for basic lands. Which means you can pull cards like Scattered Groves or Mistveil Plains. So I want to warn you against popping an Arid Mesa for a Sacred Foundry and paying 3 life in the process for one land. That's all kinds of me wondering just what is up with your thought process to think that is even remotely a good idea!
But lets get back to the Math for a moment here. Why is it that it takes until 20 turns before there is a measurable and definite card advantage for Fetching just once?
Well, there are a couple of answers. The first, and most unprovable for either side, is confirmation bias. People play the Fetchland, and they know that there is one less land in their deck, that the Fetch land effectively reduced their deck size by one. And because perception is more important than objective fact, people believe that Fetchlands will provide a measurable effect that they can be affected by.
Second, if you go back to the link I gave at the start of the article, is that this is also over the course of twenty turns. That's not card draws, that's turns, as you will have drawn 26/27 cards (7 initial plus 19 or 20 draws depending if you go first or not). With nearly hald your deck having gone into your hand by this point, there's no guarantee that you're even going to draw one of your Fetches at all! Spates of bad luck can mean that even when you do, you've already got all the mana you need, which leads into the real point two - or perhaps it's two point five?
Fetchlands are also dead draws after a certain point in the game.
What are you going to do with your third or fourth Fetchland, assuming even, that you have the life to pay for them?
Nothing, bluff, or hope you're also playing a deck with Landfall. One of the under-appreciated benefits to the second Zendikar block, I think, was the combination of Fetches and Landfall for double triggers on a turn.
But what I think is the most damning indictment of the true utility of Fetchlands is the assumption that you're pulling the next land out of your deck. You're not! You're just redistributing the lands in your deck after you shuffle, and on average, you don't gain or lose anything! Yes, there is a small percentage, but by the time that has an actual effect that isn't a random up or down, the game is most likely over!
What Fetchs are meant to do, what they should be doing in a deck, is to help fix your mana base. With the predominance of dual lands, using a Fetch to pull down an off colour dual land can only help. Boza, one of my contributors, pointed out that you can use Arid Mesa to pull a Steam Vents in a / deck, which can allow you to cast Lightning Bolt on Turn 1, then help cast Cryptic Command on Turn 4 without shorting out your mana curve or base. Fixing Mana screw is more important that getting another land out of your deck.
It's also a great way to enable Revolt and Fatal Push, so there's that.
When these cards were first printed in Odyssey, the only options for what they could target were basic lands - which was a point so obvious, I think it slipped Wizards' minds that this could backfire horribly, and the Alpha Dual lands, which were long out of print.
But the addition of new dual lands with basic land types - Shocks (already mentioned), Cycling (such as Fetid Pools), Slow (such as Canopy Vista, and require that you have two or more basic lands already in play) - that also had basic land types, meant that the viability of these cards increased massively. Enough so, that when Wizards started printing other fetches, like Terramorphic Expanse or Evolving Wilds, they specified Basic lands to help mitigate the abuse they could pull off. And lowered the rarity to make them more viable and accessible in limited formats.
Now, this is not to say that Fetches are bad, per-se. Rather, they are simply a case of a card whose use has greatly exceeded what was intended for it. It happens all the time, except because these are lands we are talking about, they tend to show up everywhere, and that ubiquity has exploded all facets of their usefulness and uselessness.
So yeah, personal opinion? I would rather have a basic land than a Fetch land. The extra card out of the deck isn't something that is provably good, and I value the two life more. Your opinion may, of course, vary. I would like to hear it in the comments below!
Join me next time when I talk about the most eye-raising aspect of Dominaria that I am glad Wizards is tackling again. What does it mean to be Legendary?
Until then, please consider donating to my Pattern Recognition Patreon. Yeah, I have a job, but more income is always better. I still have plans to do a audio Pattern Recognition at some point, or perhaps a Twitch stream, and you can bribe your way to the front of the line to have your questions, comments and observations answered!
LordMithos says... #3
Well written article as always, but what about in the case of EDH where you're likely running far fewer fetch lands, you have a much larger margin for error with your life totals, and mana fixing is arguably harder in multicolour decks?
February 15, 2018 1:52 p.m.
The statistics of playing one fetch one time might not be impressive butt say you play 10 fetches over the course of 2 out of 3 game the satistics become much more impressive
February 15, 2018 1:58 p.m.
Tyrant-Thanatos says... #5
shadow63 not when you're lopping off 10 bloody life for a smidge chance of not pulling another land they don't.
Anyway, I agree, Fetches are decent cards and have their uses, but for some reason they've become massively overrated. A term I don't often like to use, but for once I feel is appropriate.
The only deck I personally run Fetchlands in is a Dredge/Life from the Loam deck that can repeatedly exploit one fetchland to churn out land after land. Even there I only run 2 of them.
February 15, 2018 3:08 p.m.
Continuing off pheardeamon's point, I avidly play grixus death's shadow and run 12 fetchlands with 6 Mana sorces. The deck also has a lot of draw power with cantrips and the use of Mishra's Bauble. I tend to think of when I should draw and fetch based on if I need lands or not. While the effects may be minor percentages turn 1, with several draw a turn and several fetches the math adds up quickly. Granted this is a deck that mitigates the life loss, decks like burn may want those minor percentage increases. Also with the unban of Jace, the Mind Sculptor the fact that fetches shuffle seems relivent
February 15, 2018 6:19 p.m.
Fetches have a place, and I will be the first to admit it. But putting them in every deck for a meager potential advantage whose probability resets every game is not one of them.
I tend to not play fetches. That's my choice, and I don't think my decks or I are worse off for it.
February 15, 2018 8:23 p.m.
Tyrant-Thanatos says... #9
I feel like people too often brush off the life cost as being insignificant. 1 life isn't a large price to pay, but it doesn't need to be. The potential advantage you gain from using fetchlands to thin your deck is even less significant.
They can be useful for color fixing, they can be useful in decks that synergize with them, be it something like Death's Shadow which prefers to be at lower life, or decks that play with their grave a lot and can make repeated use of a single fetchland. But the rampant fad of putting them in every single multicolor deck anybody ever runs is nonsense.
February 15, 2018 8:35 p.m.
hikerdude5 says... #10
What you said about fetches not always reducing the odds of drawing land is not correct.
For n>m,
m-n<0
m-n-1<0
mn>mn+m-n-1
mn>(m-1)(n+1)
m/(n+1)>(m-1)/n
Setting m equal to the number of lands in our library before fetching and n+1 equal to the number of cards in our library before fetching, it follows that the ratio of land cards to total cards will always be lower after fetching, assuming we have at least one fetchable land in our library. Therefore, the odds of drawing a land will also decrease.
February 15, 2018 8:59 p.m.
While I understand and agree with the general point you're trying to make, there are some errors with your points on the probability of drawing lands.
First off, 7/21 = 33.333% and 6/20 = 30%, not the other way around. Your chance of drawing a land has still decreased. Secondly, this isn't taking into account the spot that the fetch land took up.
Using your example of a 21 card deck with 7 lands in it:
On your first turn draw, you do have a 7/21 (33.33%) of drawing a land. Now, you make your draw and it's a fetch land. You now have 6 lands out of 20 cards. You then play and activate this fetch land and remove another land from your deck. Now you have 5 lands out of 19 cards in your deck. At the start of your second turn draw, your chance of drawing a land is now 5/19 or 26.32%, down from the 33.33% chance on your first turn.
Now, whether or not this 7.01% decrease is worth the 2 life to you is still debatable, but it's still better value than what you seem to be suggesting.
Once again, I agree with the point you're making and I don't think every deck should run fetches purely for "deck thinning." The loss of life and potential to be dead draws is still fairly substantial and should be taken into account. However, they're use as a "deck thinning" tool is still a little more viable than your example suggests.
Other than that, well-written article! Pleasure to read!
February 15, 2018 11:20 p.m.
CharonSquared says... #12
Minor quibble: the fetchlands were originally printed in Onslaught, not Odyssey
Bigger quibble: I run Grixis Delver in modern. Delver decks, mine included, generally run 19 lands. If I ran all basics, that would be about 6 mana sources of each color. Even with all the cantrips in a delver deck, you'd get color screwed almost every game. The combination of fetch and shocklands lets a high number of your lands get you any two out of your three colors. Although the deck thinning isn't a huge deal, the deck simply wouldn't function without fetches.
February 16, 2018 2:24 a.m.
While I agree with the basic premise of "not every deck should be running fetch lands," I could spend quite some time assailing the flawed conceptual details and the lack of mathematical details included in this specific argument. It's late, so I'm not going to spend the time, but I'll simply say Magic players could learn a great deal about statistical probability from poker players, rather than attempting to take a macro-analysis approach by utilizing a hypergeometric function or running a Monte Carlo simulation, as random statistical probability is far more important than simulated statistical certainty when playing Magic (as well as poker).
As noted in the attached article, which is quite dated but mathematically accurate, telling players the mean statistical number of lands in your opening hand is 2.33 based on a 60 card 40/20 spell to land distribution is much less meaningful (in my opinion) than telling players you'll start the game with either 2 or 3 lands in your opening hand in roughly 60% of your games. Cards don't come in fractions in actual gameplay, so understanding the potential benefits of the random statistics, especially in early turns, becomes much more relevant than knowing you'll be nearly mathematically certain to draw an additional spell by turn 20, which based on a 1,000,000 simulation scenario becomes valueless.
Ultimately, the greatest value of the fetch lands has been and always will be color fixing, as having the ability to play the correct spells in your early turns means more than anything in Magic, especially in the deeper formats of Modern and Legacy. Fetch lands supersede even original dual lands in effectiveness for color fixing, as an Arid Mesa can become any dual or shock with mountain or plains in their super type (7 of the 10 dual/shock versions available to fetch). Meanwhile, a Plateau is just that, and a Sacred Foundry is just that, still minus 2-life to use immediately. Having the exact color(s) you need the turn you play the fetch is almost always worth the 1-life paid. It's why the value of fetches has held fairly steady despite the significant reprint in MM3 - they'll always be in demand, as they should be for competitive players.
Finally, as pointed out by several other commenters, the deck thinning effect, while modest in many cases, always results in an improvement in statistical odds to draw a non-land card...ALWAYS - no matter how small. Any insinuation that it does not is simply incorrect; there may be diminishing returns as the number of lands declines, but there is still always a statistical decrease in the likelihood of land draw when using a fetch land. If that's your only reason for playing them, then I'd agree - the juice isn't worth the squeeze, so to speak, but it's the all-in-one utility of the fetch lands that make them so important to so many decks. Saying they are overrated is simply erroneous.
February 16, 2018 3:37 a.m.
Confused me a little with the probability until I realised you'd put the percentages the wrong way round. I agree that they aren't necessarily worth the life loss but they will always be improving your chances of making your next draw not a land, even if it's only a fraction of a percent.
February 16, 2018 7:22 a.m.
Tyrant-Thanatos says... #15
Azdranax Your argument seems to hinge almost entirely on decks that run more than 2 colors. Arid Mesa is in no way better at color fixing than Plateau or Sacred Foundry in a strictly RW deck. And yet... people run them anyways. And they claim, for the thinning. Overrated is exactly what they are. Useful in a lot of decks? Yes. But I've literally watched people run then in fucking mono-color decks for the thinning effect. What else can I call that but overrated?
February 16, 2018 2:04 p.m.
Rhadamanthus says... #16
@Tyrant-Thanatos: In 2-color decks they make your draws more consistent by acting as additional copies of the lands they fetch. Would you rather have 4 Plateaus in your deck or 8+? If your deck is built such that it wants to be able to consistently generate any of , , or on turns 2 and 3 or risk conking out, then the difference could matter significantly. Losing 3-4 life is better than losing to mana screw.
But I definitely agree that the "thinning" effect is a myth. The value of fetches is how they allow for consistent color generation from your land draws.
February 16, 2018 5 p.m.
Tyrant-Thanatos says... #17
Rhadamanthus: Battlefield Forge, Clifftop Retreat, Inspiring Vantage, Rugged Prairie There's plenty of dual lands I'd take over Arid Mesa in RW 7 days of the week, and most of them don't cost $30+.
And this doesn't change how often I see it slotted into mono-red.
February 16, 2018 5:17 p.m.
Tyrant-Thanatos I planned to respond to your previous response, but considering your latest response as well, I think I'll simply leave it as I agree with you in terms of mono-chromatic decks using fetch lands to achieve a minimal thinning effect, which is a waste of one-life, as in the vast majority of games having one less land in your library will not have any impact on your actual draws.
Beyond that, we'll have to agree to disagree, as I think we are talking about different things in many ways, as monetary card value is not a thought in my consideration, only color-fixing based on competitive Magic builds for formats requiring the best possible consistency and speed for the builds. That simply would not apply to kitchen table and semi-casual decks, or even EDH decks (outside of the most optimized cEDH options).
That said, in Magic, especially competitive Magic, the worst play, turn for turn, is not being able to play the best option in your hand, and fetches solve that problem as well or better than every other land choice in the game. If they didn't, they wouldn't be an absolute staple of the top multi-colored decks in the modern and legacy formats.
February 16, 2018 9:42 p.m.
Tyrant-Thanatos says... #19
Azdranax Honestly I agree with you on much of that though. They are good useful cards in tuned Legacy and some Modern competitive decks. But they are a prime example of an issue I see all too commonly in mtg: competitive success skews how even casual players see cards. People hear about how many of the high end competitve decks run something, in this case fetchlands, and then proceed to buy mass quantities of them and shove them in every single deck they physically can, whether it belongs there or not. That is, so far as I'm concerned, definitively overrated.
I've managed to outright use a single Lava Axe as a wincon in a control deck because people are so prone to chew through so much of their life by fetching untapped shock lands. "The only point of life that matters is the last one" mentality has leagues of fools throwing away their entire life pools because they perceive multiple points of life as being worth less than one mana.
Fetchlands are completely valid, and in fact very strong in the right deck.
February 16, 2018 10:07 p.m.
As a guy who has played serious mono- control, and one who has gone on record here in PR as saying exactly that, the idea that the only point of life that matters being the last one is only true when you realize that you have, and may need those other ones to protect the last.
On another forum, this article has spawned a not-quite-as-interesting discussion, but one where one of the participants noted that Fetch Lands only really work under the , which posits that you only get reliable results after a truly absurd number of iterations.
That, and there's something of a confirmation bias in the use of Fetches that's hard to overcome.
February 16, 2018 11:46 p.m.
People appear to be totally missing the point.
Yeah, guys, fetchlands have their use. They're obviously good when you're running a lot of colors, when you want to lose life, or when you're manipulating your graveyard!
But the post is about how they're not always good, like some people seem to think. There are mono-colored decks where people run fetchlands just for the sake of thinning, even though they don't want to lose life or mess with their graveyards. The point of this post is to warn against choosing to use them solely for the sake of thinning - which I agree with 100%. The thinning benefits are not worth running a fetchland for on their own, even if they MIGHT, maybe be a nice thing that happens when you're running them for some other reason.
February 17, 2018 1:32 p.m.
Wauw. What conclusion: thinning one card from your deck in the early turns aren't worth it, therefore fetchlands are bad.
Along the way, they fix your mana. A lot of decks, at least in modern, have 2 main colors and then splash a third, and it's the splash color that's hard to do, if you don't have fetches. Apart from this, there's regular manafixing - hard to play triplecolor cost cards, if you don't have fetches to get exactly the right lands for that (as long as you play 2+ colors, no problem for monocolored decks, obviously). You want that Siege Rhino on turn 4? It's not impossible out of dual colored lands and basics, but it's fetches that makes it reliable.
Incidental advantages: Shuffling after playing Brainstorm or activating Jace, the Mind Sculptor. As far as I know (and I'm no legacy player at all), this is one of the strongest things to do in that format. Brainstorming without reshuffling is nowhere near the strength of brainstorming and reshuffling.
Triggering landfall (already mentioned). Not something you see that often, but in modern both burn (because Searing Blaze - try and get a basic land to enable landfall in your opponents turn) and decks with Tireless Tracker use this a lot.
Another reshuffle: Anything dealing with the topcard on your library, like Courser of Kruphix or as a play against lantern-control. Useful enough on its own - enough fetches can actually be the only way to play against lanterncontrol before sideboard.
Fetches go the graveyard, which is also an advantage to many decks - anything with Tarmogoyf, or cards that use delirium like Traverse the Ulvenwald and Grim Flayer, or cards that replay from the graveyard like Sun Titan, Ramunap Excavator or Renegade Rallier. Or anything so classic it uses threshold (speaking of classics; fetches can fetch the original ABUR dual lands - they've been able to do that since the first fetches from onslaught). Just count the advantage on this - a card in the graveyard, the land you really wanted for whatever hand you have, and a little bit of late-game thinning.
And only THEN, after you've gotten a good way into the game, the thinning becomes relevant. Is there seriously anybody thinking that removing one card in 52 (opening hand + 1 drawn in your first turn) is anything but microscopically relevant for your turn 2 draw on its own, or is that perhaps a strawman who has learned playing mtg? How often have a deck that didn't really need fetches played them? Are there any examples that could be linked? Monoblue merfolk with fetches whether legacy or modern? Monogreen stompy (without any revolt or landfall)?
February 19, 2018 5:17 a.m.
And yet there are decks that do none of those things that still ran Fetches? The article that I linked at the beginning pointed out that they started doing their math when they saw decks in even a single colour that ran 8 Fetches just for that one 'advantage'.
Also, I have to wonder if you read the comments above your post. A lot of people, including me, whom you seem to be putting words into my mouth here, are saying that Fetches aren't bad, they're just over-hyped, where people will, such as you have, jump on any possible advantage they offer and assume that Fetches are the perfect response.
They have their place. In every damned deck isn't one of them.
February 19, 2018 10:48 a.m.
Modern dredge had a fetch and a fetch-less version. I think the fetch version ended up dominant, but I don't know how the mirror match played out between them.
February 19, 2018 12:46 p.m.
Funkydiscogod: When it loses to Burn, or non-infinite sources of damage? When it gets punished either through Aven Mindcensor, Blood Moon, or in rare cases, Ob Nixilis, Unshackled; or the opponent casts Stifle or Disallow?
Fetchlands aren't perfect. They can't be in every deck. Stop trying to move goalposts!
February 19, 2018 1:13 p.m.
The problem with the article (and this has been pointed out all over the place over the years every time it gets posted) is that it assumes everything in a vacuum. It ignores that life is a resource.
For example, suppose I lose next turn to my opponent's infinite combo and the only way I can win is if I do so on my current turn. In that case, paying all but 1 of my life total to gain a even a 1% chance of drawing the card I need to execute my own combo is worth it. If I'm playing Storm and sitting at 4 life, paying 3 of that life to go grab a Steam Vents before I try to draw a card is worth it.
This is done at multiple instances throughout the game because I can only play one land a turn. So with the example of Storm in Modern, if I'm going up against another fast deck like Death's Shadow, Burn, or another combo deck, I can map out the number of cards my opponent can play.
So let's continue by taking a look at Burn. It can do 4 damage with Boros Charm. 2 damage a turn with Goblin Guide is one of its more efficient means of damage. Monastery Swiftspear does a lot of damage but is limited by the mana available. Atarka's Command likely does the most damage. A resolved Eidolon of the Great Revel is enough to kill me on its own. What's the fastest way that a burn deck can kill me, as a Storm player?
Hitting a land drop each turn is ideal for the Burn player. If they hit 4 lands and nothing but spells, they should be able to do 20 damage by turn 4.
So with 7 starting cards and then 4 turns being on the draw, they have 11 cards total. 4 of those are lands, meaning they get 7 of their burn spells. The absolute nuts is if they get multiple Goblin Guides and Atarka's Command. So:
Turn one: Goblin Guide, attack for 2 (18 life remaining).
Turn two: Cast another Goblin Guide, attack for 4 (14 life remaining). Cast Lightning Bolt (11 life remaining).
Turn three: Monastery Swiftspear -> Atarka's Command for 3 to the face (8 life remaining), pump the team, attack for 9 (0 life remaining).
That's 21 damage by turn 3, with the Burn player getting a strong draw. So if I'm on the play, and I know I take 21 damage by the Burn player's turn 3, that means the most life I can pay of my own is 10. That's because it's what keeps me alive at the end of turn 2. If I pay even 1 life, I'm dead on turn 3 regardless of what else I play. Because of this, any life I pay by turn 3 is irrelevant.
So if my opening hand has only 2 Misty Rainforest, but I need , then I'm going to pay at least 2 life to fetch a pair of Steam Vents. If I have to fetch and shock twice, I go to 14. I'm dead when my opponent gets their turn three, so I decide to combo off. Part way through my combo, I draw Scalding Tarn and Island. I can still play a land and I only need to get , so I can play the Island. However, since I know I'm dead the following turn, I may as well pay the life to use Scalding Tarn, even if it only improves my draw by 0.000000000000001%.
This is a long, convoluted example, but I hope it illustrates my point. Life is a resource. Games more often that not depend on certain life total breakpoints (in our Burn example, it goes 18 -> 13 -> 1). Using a fetch to thin out your deck in those cases makes the smallest percent worth it.
The times where the article is correct is when looking at a meta that does random amounts of damage each turn and you're playing Mono-Red Burn. It's better off to play nothing but basic Mountains, even if the option of fetches is there. That's because you'll never be in a situation where paying life to improve your chances will be relevant.
Even then, sometimes the shuffling itself becomes a relevant resource. Or you want to fill your graveyard (Grim Lavamancer being the go to example). But most of the time, it comes down to mana fixing. Because fetchlands can grab dual lands, they open up all 5 colors without any issues. If you rely just on dual lands, you may draw all of your spells, but only your lands. If those lands were Verdant Catacombs? It's no longer an issue because you can go get your Taiga or Badlands.
This becomes even more obvious against decks with few actual win conditions. Sneak and Show is going to attack me with either Griselbrand, Emrakul, the Aeons Torn, Ashen Rider, or all of the above. Scapeshift is going to kill me with Valakut, the Molten Pinnacle. Once you know the threshold for when these decks can kill you, crossing that threshold means any life you pay from fetchlands makes the deck thinning worth it.
Edit: I should mention, my point is that "Fetches are good for thinning your deck" is a correct statement. This is only clear if you take the time to think about why. But once you do, it's obvious and becomes second nature to most players.
February 20, 2018 9:20 a.m. Edited.
The article that you linked and says that there are monocolored decks is about 15 years old.Runnig Fetchlands only because of deck thining is not worth it thats absolutely correct but most modern or legacy deck are not mono colored and most Mono colored decklists ive seen (Merfolk, goblins, 8-rack for example) dont run fetches so its simply wrong that they are overrated.
Every 2+ colored deck needs them because it gives you additional copies of your shock or o-duals.
The problem with fetchlands is the "deck thinning myth" that gives you only very small advantages in reality (ive actually never heard anyone saying that those effects were huge so i dont even know if thats a problem)
Pheardemons says... #2
Can you go into the statistics of multiple fetchlands? An avid player of modern here, I have played in, and against, games where 4 or 5 fetchlands have been used. Namely, the new Death's Shadow decks that want to lose life, and play somewhere in the range of 8 to 10 fetches if not more. One fetch may not be viable, but more could increase those chances correct? Are those statistics worthwhile?
Also, something you didn't mention was that, for decks that don't have many turn one plays, it is simply paying 1 life for a dual land at their opponents end step. Would you say that a dual land isn't worth one life? With the small percentage difference of the fetch itself.
February 15, 2018 1:37 p.m.