3 Types of successful Goblin decklists
Quote:
Originally Posted by
Pee-Dee-2
ID against Zombardment and became 4th.
Drawing after only 5 rounds of Magic? And even worse: drawing against Zombardement? You are getting old, Pascal :tongue:
---------------------------------------------------------------------------------------------------------------------------------
@ ALL: Here's the article, as promised.
Hey there fellow Warchiefs.
Last month I presented a summary of some data that I collected (click here for reference). I went on collecting data because I was a little disappointed that some important question could not be answered because of lack of data. Today I’m presenting you an analysis on a larger dataset, which might solve some, but not all of the questions raised last time:
Here is what I wanted to find out this time (all questions relate to Vial Goblin decklist in Legacy, in case this isn’t obvious):
(1) Do splash color(s) affect the overall performance?
(2) Which number of lands-total is most related to success?
(3) What is the “correct” number of Rishadan Ports?
(4) Are lists with Mogg War Marshal strictly better than lists with Warren Instigator?
(5) Piledriver an indicator of success? (is it true that: the more PD the better the performance?)
(6) The more MWM, the better the performance?
(7) The more cmc2 spells the better?
(8) Do Goblin Sharpshooter and Skirk Porspector affect performance?
(9) How does the total number of non-Goblin cards in a decklist relate to success?
Quite a bunch of problems to solve, eh? Luckily I did find answers to most of the questions! Here are the facts:
I analyzed 272 matches of 45 different decklists. Again, most of the decklists were winning decklists which reflects a certain report-bias for good results. So, the overall performance of this particular set of decklists was quite good to begin with. That means: if I’m stating that a certain card “does not positivey affect the performance of the decklists” then that actually means the card in question is frequently found in winning decklists and doesn’t make the performance better than “above average”. The average performance of the whole dataset is expressed in a Win-% of 62.85
(1) Do splash color(s) affect the overall performance?
Short answer: No. Overall mono-red decklists were just as successful as multicolored ones.
(2) Which number of lands-total is most related to success?
Decklists with 22 Lands were most successful and slightly above average(65.05% Wins), while other lists’ success ranged from 53.00% to 62.07%
(3) What is the “correct” number of Rishadan Ports?
This question can’t be easily answered. Read more about that in the latter part of this article!
(4) Are lists with Mogg War Marshal strictly better than lists with Warren Instigator?
Short answer: No. However, don’t jump to the conclusion that it wouldn’t matter whether you run MWM or Winstigator. Further analysis revealed that it’s a matter of configuration of the whole decklist.
(5) Piledriver an indicator of success? (is it true that: the more PD the better the performance?)
Short answer: No. However, don’t jump to the conclusion that the number of Piledriver doesn’t matter! Further analysis revealed that it’s a matter of configuration of the whole decklist.
(6) The more MWM, the better the performance?
Yes, that is true for decklists that do NOT run Warren Instigator. That is, if you decide not to run a Winstitigator decklist, you should run as many MWMs as possible.
(7) The more cmc2 spells the better?
This analysis was the most fruitful of them all. I will discuss the results of this analysis in detail later on. For now let’s just leave it at the conclusion that ”No, there is no linear relation between the number of cmc2 spells and performance. Read the latter part of the article for a detailed explanation.”
(8) Do Goblin Sharpshooter and Skirk Porspector affect performance?
Yes, they do. Lists with Sharpshooter performed slightly above average (63.84%), while lists without him did worse than average (60.84%). The same pattern was found for Skirk Prospector: 1 Skirk Prospector resulted in 63.95% wins; Prospector-less lists win 60.79% of their games.
(9) How does the total number of non-Goblin cards in a decklist relate to success?
This question can’t be answered easily. Read the discussion below to get a better impression of this relationship.
Some of these results really surprised me, so I further explored the dataset to see if I could see a pattern in it (I want to spare you with the details. Let’s just say that I did a bunch of analyses to actually reveal a certain pattern at the end). After looking more closely at the data I found that one variable seemed to be most defining for success, namely “the number of cmc2 spells”. The pattern I found was such that decklists with either 10, 13 or 14 cards with converted manacost of 2 (a sum of: Piledriver, WInstigator, MWM, Gempalm Incinerator, Stingscourger, Tin Street Hooligan) were more successful than those with either 11 or 12 cmc2-cards (decklists with 4-9 cards with cmc2 were excluded from analysis due to lack of data). Here’s an overview of the respective WIN-%s:
10 cmc2 cards: 67.47% (30 records)
11 cmc2 cards: 60.24% (46 records)
12 cmc2 cards: 56.87% (70 records)
13 cmc2 cards: 67.21% (43 records)
14 cmc2 cards: 69.95% (38 records)
As you can clearly see, the decklist with 10, 13 and 14 sms2-spells score quite some points above the average of 62.85, while the other two score below it. And now comes the interesting part! The analysis revealed that most of the decklist with 10 cmc-2 spells are very similar to each other, not only in their cardchoices, but also in the NUMBERs of the keycards they run. E.g. Almost all lists with only 10 cmc2 spells were Winstigator-lists, and not only that, most of them (90%) were mono-Red. The same is true for the “13-“ and the “14-category”. The decklists seem to agree not in particular on the number of cmc2-spells, but rather on many cardchoices throughout the deck, which, by incidence, result in a certain number of cmc-2 spells. That is, they did not, in advance, decide that “I’m gonna have precisely 13 cmc2-spells in my deck, let’s look what they will be.”, but the other way round: “I’m gonna design my deck in such a way…Oh, this results in having 13 cmc2-spells. I wonder if that number matters at all…”.
Interestingly it was not only the case that I had found 3 distinct, successful decklists. It was also the case that those 2 other categories (the “11” and “12”) had similar features too. So what I’m now presenting to you are 4 stereotype decklists, 3 of which are successful and 1 of which is awful, so to say. It’s no surprise that you might recognize those…
http://magiccards.info/scans/en/mi/152.jpghttp://magiccards.info/scans/en/zen/154.jpghttp://magiccards.info/scans/en/m10/139.jpg
STEREOTYPE #1: The Winstigator list (the 10-cmc2-lists)
Features:
* WIN-%: 67.47
* most lists of this data-subset had 3 to 4 Warren Instigators
* 90% were Mono-R
* 75% did not run MWM
* most lists had only 2 Piledrivers
* higher average number of Stingscourgers and Chieftains
Stereotype decklist
Mana (22)
4 Cavern of Souls, 4 Wasteland,
2 Rishadan Port
10 Mountain
2 Chrome Mox
CMC2 (10)
2 Goblin Piledriver
3 Warren Instigator
3 Gempalm Incinerator
2 Stingscourger
Rest 38
4 Aether Vial, 4 Goblin Lackey, 4 Goblin Matron, 4 Goblin Ringleader
3 Goblin Warchief,
3 Goblin Chieftain
1 Krenko, Mob Boss
2 Siege-Gang Commander
2 Others
The first list is a classic Winstigator decklist. I averaged the number of the most frequent cardchoices to get the picture of the decklist you can see above. Note, that this is not necessarily the best decklist, and not even how I would build it. The decklist above reflects the average (the “stereotype”) decklist of the successful “10 CMCM2 spells-category” that has proven to be successful according to the data. It’s important to remember that I asked the question “How does this mysterious 10 cmc2 data subset look like?” and not “What does the average Winstigator list look like?”.
http://magiccards.info/scans/en/mm/324.jpghttp://magiccards.info/scans/en/on/205.jpghttp://magiccards.info/scans/en/ts/170.jpg
STEREOTYPE #2: 13 CMC2 spells – The classic Vial Goblin deck
Features:
* WIN-%: 67.21
* no list of this subset had fewer than 3 Piledrivers
* majority had 3 or 4 Rishadan Port
* 70% had 4 Gempalm Incinerator
* no single list had Thalia or Karakas
* 90% of the lists had only 4 non-Goblin spells (namely Aether Vial)
Stereotype decklist
Mana (22)
4 Cavern of Souls, 4 Wasteland,
3 Rishadan Port
5 Fetchland
1 Taiga
5 Mountain
CMC2 (13)
4 Goblin Piledriver
3 Mogg War Marshal
4 Gempalm Incinerator
1 Stingscourger
1 Tin Street Hooligan
Rest 38
4 Aether Vial, 4 Goblin Lackey, 4 Goblin Matron, 4 Goblin Ringleader
4 Goblin Warchief,
1 Goblin Chieftain
0-1 Krenko, Mob Boss
1-2 Siege-Gang Commander
1 Skirk Prospector
1 Goblin Sharpshooter
This decklist might be familiar to you, and it should be. This list is basically the one that B-rad, Max Tietze and Jim Davis are playing again and again and for probably eternity. Note that in this data-subset, about 40% of the decklists were mono-Red, while making nearly the same cardchoices (Tuktuk Scrapper instead of TSH, 1 more Rishadan Port instead of the Dualland and 5 Basic Mountains instead of 5 Fetchlands). However, the picture remains the same: the analysis of the “13 CMC2 spells”-data-subset spit out this averaged decklist, with very straight-forward cardchoices, such as a maximum number of manadenial, Piledriver, MWM and Gempalm Incinerators, with the classic triumvirate of Sharpshooter, Skirk Prospector and a Shatter-Goblin.
http://magiccards.info/scans/en/dka/24.jpghttp://magiccards.info/scans/en/lg/248.jpghttp://magiccards.info/scans/en/on/330.jpg
STEREOTYPE #3: 14 CMC2 spells – The multicolored Thalia-build
Features:
* WIN-%: 69.95
* 100% of the decklists in this category were multicolored
* 75% of the decklists played Thalia and Karakas
* 2-4 Piledrivers
* 3-4 MWM
* 85% had both Sharpshooter
* 100% had 1 Skirk Prospector
* 100% had 4 Goblin Warchiefs
* 100% played without Goblin Chieftain
Stereotype decklist
Mana (22)
4 Cavern of Souls, 4 Wasteland
6 Fetchland
3 Duallands
1 Karakas
4 Mountain
CMC2 (14)
3 Thalia, Guardian of Thraben
2 Goblin Piledriver
3 Mogg War Marshal
4 Gempalm Incinerator
1 Stingscourger
1 Tin Street Hooligan
Rest (38)
4 Aether Vial, 4 Goblin Lackey, 4 Goblin Matron, 4 Goblin Ringleader
4 Goblin Warchief,
1 Krenko, Mob Boss
1 Siege-Gang Commander
1 Skirk Prospector
1 Goblin Sharpshooter
This decklists seems like a configuration of the CLASSIC decklist, since a majority of cardchoices are overlapping. However, by having played both versions I know that the most defining features are different: The THALIA decklist does usually not play Rishadan Ports at all (or at least very few of them) and basically trades 3 slots of Piledrivers and Goblin Chieftains for 3 Thalias. Another defining feature is the heavy dedication to splash colors. To open up more effective SB choices. In comparison of CLASSIC vs. THALIA decklists you could say that the first is more straightforward in its cardchoices, while the second has a bigger accent on non-Goblin cards to improve the deck (Thalia, SB-cards).
http://www.7cornersexporting.com/Joo...50050c166b.jpg
STEREOTYPE #4: 12 and 13 CMC2 spells – The “awful” list?
Features:
* WIN-%: 60.24 and 56.87
* 2 Rishadan Port
* some play Karakas, even without Thalia
* many Fetch- and Duallands, although most decklists were monoR
* only few lists with 4 Piledrivers
* many lists without Sharpshooter and Skirk Prospector
* some lists with WInstigator, but without Chrome Mox
Stereotype decklist
Mana (22-24)
4 Cavern of Souls, 4 Wasteland
2 Rishadan Port
6-7 Fetchland
2 Duallands
0-1 Karakas
3-5 Mountain
CMC2 (12 or 13)
1-3 Goblin Piledriver
2-3 Mogg War Marshal
3-4 Gempalm Incinerator
1-2 Stingscourger
0 or 3 Thalia
Rest (38)
4 Aether Vial, 4 Goblin Lackey, 4 Goblin Matron, 4 Goblin Ringleader
4 Goblin Warchief,
0-2 Goblin Chieftain
0-1 Krenko, Mob Boss
0-2 Siege-Gang Commander
As you might have noticed it is really hard to provide a clear picture of this category, which is mostly because the data isn’t clear there. My interpretation would be that this category’s most defining feature is that it doesn’t make any dedications at all. It isn’t maxing out anything, neither Piledrivers, nor MWMs, nor Manadenial, nor number of basiclands. This category appears to squeeze everything into one decklist, that has proven to be successful elsewhere. On thinking about that magicmerl’s spaghetti-sauces parable came into my mind: you have 3 excellent spaghetti sauces, packed in 3 different glasses. Every sauce tastes wonderful on its own. However, if you open the glasses and mix the sauces up into 1 sauce, it will taste like regular spiced ketchup, maybe even worse.
Lessons learned
* to my surprise there are 3, not 2 different types of successful Goblin decklists
* I want to suggest a terminology for them: either (1) Winstigator, (2) Classic and (3) Thalia or (1) Type 1, (2) Type 2 and (3) Type 3.
* Goblin Sharpshooter and Skirk Prospector make the difference, albeit a small one
* extremes pay off, which is evident in the dedicating choices of the 3 successful decklists
* do not forget that those decklists are stereotypes and that they will always need one or two tweaks to make them fit the local metagame
* Playing good and smart Magic is more important than having a decklist that is 100% “correct”. This last point is not a result of my today’s analysis, but rather a reminder for people like me that tend to lose themselves in the idea that there is a perfect decklist and forget about the beauty of the GAME you are playing. There is more to Magic than only deck building.
I hope you liked my article. Comments are, as always, appreciated.
Cheers,
GL