jamuraa's picture
By: jamuraa, Marie Janssen
Jun 23 2008 1:01am
Login to post comments

I have to apologize to you, I've been lax in my writing. Normally I'm here every week around the same time, but I have been slacking off because of the different way of compiling the data for the Premier Events in MTGO 3. Because of the fact that I don't really have complete data I'm a lot less excited about writing every week. There are also some other issues which I won't really go into. But that's neither here nor there for the moment, we have some news about how Premier Events are being implemented.

Last week I complained loudly about the lack of Intentional Draws in Premier Events, something which was very much an annoyance for almost all of the players that I have chatted with in my gathering of metagame information. Even the people who got in on tiebreakers and didn't think they would have made it in thought that Intentional Draws should come back. However it was announcedneed to bring back IDs, so they won't. There are a couple of implications of this, which have been hotly debated over in the extensive thread on the MTGO forums. the other day that they wouldn't be coming back at all. The reasoning in the announcement is clear: they don't

The first change is the obvious one - there are no more intentional draws. This means that everyone will need to play 5 rounds in order to finish and get to the top 8. I previously went over some swiss math which advised that if you won every round and were two rounds away from the end of the tournament, you could be guaranteed a spot in the top eight by drawing twice. I will have to revise that assessment. Let's look at the math again, assuming that everyone is forced to play out every round. Who will get into the top 8 at the end of the tournament? Let's take the simple case of the 24 player Premier Event - the smallest number of players you can make MTGO start a 2x event with, something which is becoming increasingly more common nowadays. After three rounds, there are three players who are 3-0: 12 players will win the first round and they will all play each other. 6 of those players will win the second round. In the third round, these 6 players will pair against each other equally and 3 will have the previously guaranteed-in 3-0. The question I have is simple: Will these players still make the top 8, even if they lose all of their remaining matches?

Well, if they all lose their matches, they will have a record of 3 wins and 2 losses. This isn't the best position to be in, because they will be in a group with 8 other people at max. This means that they will need to be worried about beating the other people in that set of people with 9 swiss points. In order to answer this question, we have to go back to our old friend, the first tiebreaker: Opponent's Match Win Percentage (OMW for short). As a refresher, this breaker is made up of the percentage of matches that your opponents have won, divided by the total matches that they have played. It is important to remember that it is also never below 0.33 for any opponent. So let's take the worst case for the 3-0 to 3-2 player. This is the case in which they were so good that they completely demoralized their opponents, making them lose ALL of their remaining matches:

Round Match Result
1 Play against a 0-0 player. This player goes 0-5.
2 Play against a 1-0 player. This player goes 1-4.
3 Play against a 2-0 player. This player goes 2-3.
4 Play against a 3-0 player. We lose this match, and our opponent loses his last round, making him 4-1.
5 Play against a 2-2 player (we were paired down!). We lose this match too. Our opponent is 3-2 after the last round.

This makes our opponent's match win percentage:

0/5 1/5 2/5 4/5 3/5 Total OMW%
0.33 0.33 0.4 0.8 0.6 2.46 / 5 0.4920

If we didn't get paired down, our match win percentage would be better (0.5320, actually). In order to determine if this will make us into the top 8, we need to look into the rest of the 3-2 players. This gets more complicated, because we need to find out how the other 6 or 7 3-2 results got to their records. Let's look at how the pairings are for Round 4. Remember that after round 3, we have 3 players that are 3-0. This means that one of the 3-0 players will be paired down.

Matches Player A   Player B Results
1 3-0 vs. 3-0 1 4-0, 1 3-1
1 3-0 vs. 2-1 1 4-0, 1 2-2 OR 2 3-1
4 2-1 vs. 2-1 4 3-1, 4 2-2
4 1-2 vs. 1-2 4 2-2, 4 1-3
1 1-2 vs. 0-3 1 2-2, 1 0-4 OR 2 1-3
1 0-3 vs. 0-3 1 1-3, 1 0-4

In order to get to the horrible tiebreaker scenario, the red match above must be won by the 3-0 player, and the blue match must be won by the 1-2 player. This means our horrible tiebreaker man is playing the green match. This scenario produces these round 5 matches:

Matches Player A   Player B Results
1 4-0 vs. 4-0 1 5-0, 1 4-1
2 3-1 vs. 3-1 2 4-1, 2 3-2
1 3-1 vs. 2-2 1 4-1, 1 2-3 OR 2 3-2
4 2-2 vs. 2-2 4 3-2, 4 2-3
1 2-2 vs. 1-3 1 3-2, 1 1-4 OR 2 2-3
2 1-3 vs. 1-3 2 2-3, 2 1-4
1 0-4 vs. 0-4 1 1-4, 1 0-5

We either have 7 players at 3-2, in the case when the green match is won by the 2-2, or 6 players at 3-2 in the case where they lose. So what's the best that one of these competitors can do in the crucial first tiebreaker? Well, there are many paths to 3 wins and 2 losses:

Okay, so there are only ten. Remember, we took the top path, so we know the worst tiebreaker for that one, but we're worried more about the best tiebreaker now.

Round Match Result
1 Play against a 0-0 player. This player goes 4-1. (they're in top 8)
2 Play against a 1-0 player. This player goes 4-1. (they're in top 8 too!)
3 Play against a 2-0 player. This player goes 4-1 (is everyone we played making top 8!?)
4 Play against a 3-0 player. We lose, This player goes 5-0 (what the heck!)
5 Play against a 3-1 player. We lose, This player is 4-1. (yep, everyone we played made top 8)

Our tiebreakers are then:

4/5 4/5 4/5 5/5 4/5 Total OMW%
0.8 0.8 0.8 1.0 0.8 4.2 / 5 0.8400

Luckily, while this is theoretically possible, no one else has actually taken this path in our event - we're the only one who went 3-0 and then 0-2. I've gone through the rest of the alternatives, and here's the breakdown of the best theoretical tiebreakers for each path:

Wins / Losses Max Theoretical Tiebreaker
W W L W L 0.8
W W L L W 0.8
W L W W L 0.76
W L W L W 0.72
L W W W L 0.72
L W W L W 0.68
W L L W W 0.64
L W L W W 0.64
L L W W W 0.6

Oh oh. Things aren't looking good for us. Or are they? The thing is, it isn't actually very likely that these tiebreakers will happen. I wrote a small simulation running a 5 round tournament with 24 players. Each match was chosen by a coin flip (don't try this at a tournament!). I then ran a few million simulated tournaments, and noted how many people who were 3-0 at the third round made it into the top 8, and how many people who went 3-0 and then 0-2 made it into the top 8. Here's the results of that:

Total Simulations 3272294
3-0 in round 3 makes top 8 99.48%
3-0 then 0-2 makes top 8 95.93%

Now there are a lot of ways for this to roll out, but the overwhelming majority of the time, if you make 3-0, you will be making top 8. This is because anyone who makes 4-1 will make top 8, and some of the people who are 3-2 will make top 8 as well. Even if you lose both of the games, or just decide to concede because you don't have time to play those rounds, you will make top 8 nineteen out of twenty times. Considering this, I don't think that this decision will affect the players that much. It means more games played for the decision of the top 8, which I approve of, and also makes 0-2 not the drop-point that it used to be. Because of this, players who are at 0-2 still have a small sliver of hope (although not a large one)

Statistics for Standard PEs: 6/05/2008 to 6/08/2008

Well, last week was special and we had more data than we would normally have, but this week we have a lot less than we would normally get, because Shadowmoor events started and we therefore didn't have any events firing anymore. We only have nine events to take a look at this week, and as a cruel bonus, we don't have more than 15% of the metagame due to the fact that I couldn't contact the players in the tournaments. Even so, I think we have some useful data here to look at. 72 total top eight spots in these 9 events, which is one of the smallest that I've included here. Of those 72, we are missing 11. There was an average of 28 players when the events fired this week, but even more didn't fire this week - it's a sad state when you start seeing standard events cancelled due to lack of participation.

Winner - Split - Second - Semifinals - Quarterfinals

Colors Deck Name Placings Percentage
Elves 29% (+2%)
Faeries 16% (-5%)
Merfolk 13% (+6%)
Token Elves 5% (+2%)
Doran 5% (+4%)
Burn 4%
Big Mana 2% (-4%)
Tokens 1% (-2%)
Rock 1% (-2%)
Reveillark 1% (-4%)
Aggro 1% (new)
  Unknown 16% (+7%)

Well, this week we still have elves on top, but the Fae are gaining on them quickly. Merfolk is still hanging in there, but I was seeing progressively less of it in the top tables as the week went by, because it doesn't do so well against the newer Elves decks. If we consider the token-generating Elves decks as well, the crazy critters are taking even more of the pie. Shadowmoor will hopefully shake up this metagame with some new decks. We'll look at a couple of those decks today, because looking at the decks of the past aren't really something that's useful, is it?

This week's top gainer is Doran, which has started to prey on the popularity of Elves with it's possibility to cast Damnation at the right time, and to put a creature that attacks for 5 on turn 3. I really like the prospects for Doran in the coming months, because the manabases are going to be a lot easier to get a handle on with the coming of Reflecting Pool, making turn three big beaters much more easy to come by. The big loser this week is Reveillark, which I am actually surprised at. Reveillark is a great card, and I expect it to be played for as long as it's standard legal for sure, and possibly in extended as well. Block decks are going crazy for it, and there are lots of new variants with the new cards in Shadowmoor which are coming, so maybe we will see a resurgence.

This week we don't have a dedicated slot for "The Others", but there are some which probably have some explaining. The green-black tokens deck that was played in the week previously that I reported has morphed into a green-black-white tokens deck, in order to include the craziness that is Mirror Entity. Green-red Aggro is just what it sounds like, it's red and green efficient creatures in order to beat down. Rock is not being played much, but it is basically just Elves with less elves and more creatures that are hard to kill like Troll Ascetic and Tarmogoyf.

Mode of the Week: Fury Husk

Fury Husk
Standard-Shadowmoor deck piloted by Martin Goldman-Kirst
4 Greater Gargadon
4 Magus of the Moon
4 Marsh Flitter
4 Mogg Fanatic
4 Mogg War Marshal
4 Nantuko Husk
4 Shadow Guildmage

Other Spells
4 Furystoke Giant
4 Bitterblossom
4 Auntie's Hovel
2 Graven Cairns
1 Kher Keep
6 Mountain
1 Pendelhaven
4 Sulfurous Springs
6 Swamp

Marsh Flitter

Welcome to the new hotness of the goblin decks - tapping down your tokens for lots of damage! This deck was featured in a similar form in a WotC deck check video during Pro Tour Hollywood. It didn't place well there, but the deck has been tearing it up at regionals lately and I expect to see a version online quickly after Shadowmoor comes out. There are lots of different ways to win with this deck, which is one of the reasons that it is winning games I think. Let's count the ways - You can win with the classic Furystoke Giant and tapping your guys super-pinging at the opponent's guys head. You can also use Nantuko Husk with the huge amount of tokens that you create with Bitterblossom and Marsh Flitter to attack for the win. But what's that? There's a Greater Gargadon suspended that you can bring out by saccing all of the creatures in play as well, hasty and ready to attack for nine damage.

This deck will also have quite the game against Faeries, which are playing less and less basic lands nowadays. If they don't have a Murderous Redcap when Magus of the Moon comes down, you can expect the game to go in the favor of the Fury player very quickly. It also has Shadow Guildmage to take out all of the x/1 Fae that are flying around. The board might need a little tweaking if Elves persists through the Shadowmoor change, you could use a little more creature kill I think. Expect to see a bit more of the red-black in the meta after the change. It's a good thing.

Outlier of the Week: Green-White Elves

GW Elves
Standard-Shadowmoor deck piloted by Rafael Tejo
3 Gaddock Teeg
4 Imperious Perfect
4 Llanowar Elves
4 Wren's Run Vanquisher

Other Spells
4 Kitchen Finks
4 Safehold Elite
4 Wilt-Leaf Cavaliers
4 Wilt-Leaf Liege
3 Overrun
3 Shield of the Oversoul
1 Wooded Bastion
4 Brushland
9 Forest
2 Pendelhaven
5 Plains
2 Treetop Village

Gaddock Teeg

This deck may look like a block deck, and that's because it is almost exactly that. The only things that are really changed is the inclusion of Overrun and some of the lands which are here. This deck won Nationals for Rafael, and it holds a special place for me because I played this archetype at the PTQ which I was at last weekend. The deck is quite powerful and I am happy to see that it can possibly do well in Standard as well as block. The way to win with this deck is simple: attack with creatures. Of course, that's something that most of the decks do for winning nowadays, so your creatures need to be bigger.

Green-white elves has quite the set of creatures that can be bigger, and it's helped a lot by the main pumper Wilt-Leaf Liege. This guy pumps every single creature in the deck at least +1/+1, and a majority of the creatures by 2. It's just a house, and the ability that it has isn't something to be scoffed at either - just wait until you get a Thoughtseize coming at your face and remember that you can't lose the Liege from your hand.

Tech that I wasn't smart enough to run in my block deck is the sideboard of Faerie Macabre. If there is a card that was tailor made against Reveillark, this is it. It will keep the white flyer in check by making it impossible to actually make it die and work. Remember, the effect from the 'lark is targetted, so you can just wait for the targets to happen and discard the Macabre in order to remove those cards from the grave. Dead Reveillark and happy player.

Well, that's it for the week which was harder than any other to determine the metagame of. I pray for the re-inclusion of the replay feature, which has been announced that it will come back, but it will not include being able to watch games from a Premier Event in progress, only after the top 8 has finished. This means that you can't go scouting the other tables, and you can count out any other full metagame analysis of a single event like I've had a couple times in the article before. Next week I hope to talk a bit about the PTQ that I attended, because there won't be PEs talk about. Also I might join a queue or something and try to get some sense of the metagame that way. I'd like to thank the people who helped me out with the meta this week, including sandydogmtg, Aluisio_Cs, jhonpotter, jerry_cn, _sprite_, wiks, BeatzMachine, RDWFTW, The Great Dweamhebe.PVDDR, Kangoo_Magic, millennium9999, ckayes, flower, Scryb, DmitrichBear, and Anssi A. If I've forgotten you, sorry, I forgot to write you down. Until Next Week, Good luck in the Constructed Queues!


by jamuraa at Mon, 06/23/2008 - 14:37
jamuraa's picture

This is GREAT news.  Thanks for pointing that out to me.  I haven't been able to follow the forums like I normally used to lately, mostly because of the forums finking out every time I try to get to reading a whole one.  I would be thrilled to do full-week full-event replays.  I'll have to think up a new way to do analysis though, as it should be obvious that someone who 0-2 drops may not be a huge input. 

Here's hoping that it comes sooner rather than later - it's really difficult to gather meta as it is. 

by iceage4life at Mon, 06/23/2008 - 18:38
iceage4life's picture

I liked the no t8 PE I was in where after six rounds of swiss people with 12 points had slots #7-17.  Silly as all hell hope WOTC changes things back to matching v2&paper.

Fantastic Article, as always! by hamtastic at Mon, 06/23/2008 - 12:16
hamtastic's picture

Great jorb there, don't ya know.

I'd like to point out one thing though, regarding your statement:

"This means that you can't go scouting the other tables, and you can count out any other full metagame analysis of a single event like I've had a couple times in the article before."

That's actually not true.  While the in event replays are going away they are slated to be replaced with full event replays (Per what the PM Gordon Culp says here: http://forums.gleemax.com/showpost.php?p=16164426&postcount=27). 

Gordon -
"We will implement replays for matches in completed PEs. There needs to be some UI work to allow this. I will be submitting a proposal to add round number buttons to the lower pane in the Event Info scene (something like the page number buttons in the collection binder probably). Then the list of matches will be available for each round, and the user would be able to select a match to replay via a right click context menu. This will go into the hopper of items that the usability group prioritizes and implements."

So, whenever they get all that worked out, there will be full event round by round replays.  :)  Someday.