How Important is the Die Roll?

[VegasJake]

Let me preface this by saying that I am just getting back into T1 and Magic in general after a 1.5 year hiatus and so don't have much recent play experience, which limits my ability to begin this discussion with much data of my own.  I am aware of the rule for not making "vacuous" posts here, and I apologize if this qualifies, but I think this is an important issue I haven't seen discussed here lately that others may be able to flesh out.

In my recent testing against the Workshop-based decks that run Trinisphere and/or Crucible, I have discovered that playing first makes a huge swing in my win percentages.  This is due to getting a chance to play out Moxen before the 'Sphere drops, or being able to play a threat or discenchant effect for the Crucible before it gets active with Wasteland recursion.

In general, what is the effect of playing first on the chances for winning in today's Type 1?  Is it a hugely signficant factor in most/all or some particular matchups, or is the effect usually not so dramatic (i.e., still present but only around maybe 10-20 percent swing in the matchup percentages)?

[mrieff]

I (and others as well) have asked for statistical evidence to measure this advantage. In practical terms this means counting wins on a few very large tournies with strong metagames.

This has been done once by Wotc at an extended Pro Tour, where playing first only gave you a slightly above 50% chance to take the match.

I am still very much in favor to repeat this experiment worldwide. It should provide very interesting info, and should confirm or destroy the prejudice that T1 is all about the die roll

[VegasJake]

Interesting.  I imagine that the particular decks/cards in an environment, those that lend themselves to very quick "swingy" plays, would increase the significance of playing first.  I.e., Workshop + Crucible + Trinisphere.  In much slower environments the effect would be diminished and for this reason the results from the Extended study would be less relevant to T1.

[Necrologia]

Actually mrieff, I think they found that going first actually hurt your chances of winning. It was pretty negligable, but the person drawing won 50-55% of the time.

It's also good to note that the test occured when Extended was in crazy Tinker combo mode. It's not like the format was much slower than Type 1 at the time, and in fact, it could have been argued to be even faster due to the lack of Force and StP.

[Gandalf_The_White_1]

Playing first is simply amazing.  Timwalk is restricted for a reason- it's undercosted and thus broken.  Well, playing first is effectivly getting a free timewalk over your opponent.  It gives you SOOOO MUCH TEMPO.  

And, as you said, the die roll is more important in macthups involving workshop decks because of their mana denial lock components.  Sometimes you get stuck with a hand fullo of moxen locked under a sph3re, and other times they can only watch helplessly as you bust out 4-5 mana sources 1st turn and hit them with R&R or w/e.

With control decks they get drain mana on their sencond turn and thus if they play 1st they will have it on their opponent's turn 2.  If they play second they can only have turn 2 drain mana up with a mox saphire or black lotus.

Aggro/control decks thrive on tempo so going first is a huge boon to their strategy.

Combo decks love to be a turn ahead of their opponent in mana development as they have more resources to combo out and the opponet has fewer resources to stop them.

There are many games in which the die roll determines who wins and who loses.  I often I look at the board and at my hand and think, damn if I went first I would be winning this now.

Of course, most of this is just stating the obvious, but I hope it helped you out.

[WildWillieWonderboy]

The importance of playing first is almost negligible, simply because of the presence of Force of Will. In the early stages of the game, if something threatening happens, FoW is a Time Stop. The "broken" plays involve dropping moxen and then dropping some threat. That means that the player whose turn it isn't gets to use two cards to tap you down and leave P1 sitting there with a shit-eating grin and a grip of three cards, half of what he's going to have when he takes his turn. What's that? The going first "broken" player has FoW backup? Lucky him, he gets to go down to one card. Supposing he was playing a worthwhile draw spell like ancestral, he now has four cards, one more than he would have had otherwise, and he is still down in cards. These plays are highly uncommon and the greatest advantage in going first right now is getting the first chance to initiate this exchange if it is favorable, sort of like chess.

The decks that make the best use of the first turn are the ones with duress and/or draw sevens.

In the case of draw sevens, you get to give your opponent, who has nothing on the board, a new set of seven cards, and he doesn't get to choose whether or not to keep them! This means that you can accelerate fearlessly into a new hand -- provided you had duress or FoW backup -- and just pass the turn without fear of the backlash from refilling your opponent's hand.

[bebe]

Not that I have statistical evidence but the last two tournaments I only lost the matches where I  won the die roll - both matches against Shock Wave's Workshop deck. I won all the other matches or drew. That is a total of 13 matches. Now I play Forces and Dazes. I find that the most common opening was fetch and artifact mana and pass on my opponents' first turns. I would likely play it the same way breaking the fetch on my opponents end step with Ancestral or Brainstorm or breaking it to counter on my opponents turn. In a deck sporting twelve counters and lots of draw spells going second worked fine when playing control combo.

[Shaman]

The die roll seems to be rather important nowadays if your opponent can start with WS+trinisphere, getting a virtual at least double time walk, or with a huge lock on the game. As we Italian are discussing in our forums, T1 is getting more and more ForceOfWill-dependant, and this is strictly related to the greater importance the die roll has achieved.

P.S: obviously I am talking as a player in a "heavy-artifact" metagame.

[Xman]

I think how important the die roll is depends on what you have in your hand, the matchup, and the deck.  If I am running something like Belcher, and my opponent is playing Workshop, you better believe I want to go first.  IF I am playing Hulk, or any other control deck, I want to go first in order to get the Drain Mana up and running.  If I am playing Combo like TPS or Draw7, I want to go first.  But if I am running against something that does run FoW, I fully expect them to have it in their first hand.

Now, it does depend on the situation.  After recent testing on this in the past few hours on MWS, I found it really didn't make a difference, unless I had a god draw, then I could combo out first turn.  but usually it took a few turns to get it up and running, so comboing out against someone running 4cc was risky, knowing they had drain mana up, as well as enough cards to have a drain and a FoW in hand.  but it was worth the shot anyway, and usually paid off.

I think it does give a SLIGHT edge, depending on what deck your playing and the matchup.  But as stated before, it does give a tempo boost to the player who plays first (Workshop, 3sphere, go anyone?).

[Smmenen]

Honestly, I wouldn't be surprised to find that the exact same results obtained for T1 - that is, people on the draw won 50-55%.  I also wouldn't be surprised to see a stat that said people in t1 won 50-55% of the time playing first.  

I think going first is very important, but not enough to consider changing the system.

[Mixing Mike]

I think going first is very important, but not enough to consider changing the system.

Agreed.  The system as it stands makes sure each player gets to play first if they choose to.  This way, somebody has to win on the draw to take the match.  It's fair enough for me.

[ELD]

[arrogance]When I go first, I win because I go first.  When I go second I win cause I'm ELD[/arrogance]

Seriously though, going first is huge when looking at the more abusive decks.  When Combo or Workshop go first, they are at a huge advantage.  Workshop decks running 3sphere are, with out question, in a huge advantage when going first.

[Covetous]

This question could be re-stated as:
If you win the die roll, would you ever choose to draw, and if so, when?

If you would never draw, then it is obvious that the die roll (and thus the ability to play first) is very important.  If you would draw sometimes, then this decreases the importance of the die roll.

I personally feel that the die roll is critical in some matchups and nearly irrelevant in others, depending on your deck.  If you are not playing FoW, then the die roll becomes more critical because you want to have your chance to take a turn before something broken happens (workshop-->something stupid, etc.)  If you run FoW, then it is less critical because an opening hand FoW on the draw keeps card advantage even (they are one card down but FoW uses two cards to counter their one threat), while preventing brokenness.  

The problem I see with an analysis of this type is that there are many confounding factors such as specific matchup, deck type and opening draws/mulligans.  In some matchups, it doesn't matter who plays first, you will almost always win or lose, so those wins on-the-draw potentially detract from the statistically-detectable advantage of going first (if it exists).  (Does that make any sense?)  On the contrary, having to mulligan to 6 when going first puts you at a large card disadvantage (25% fewer cards than your opponent), potentially preventing you from winning, and thus detracting from the potentially detectable benefits of going first.  I think that these confounders may be impossible to control for, but then again, they may be irrelevant.  

I'm not sure what to conclude although I think that currently going first is more critical than at certain times in the past.  However, the powerful, game-swinging cards present in type 1 may negate the potential advantage of going first.  (Think of my example of mulling to six then playing A Recall, eliminating the card disadvantage you were at while retaining the tempo of going first)  The powerful cards in type one make our format different from other formats, where the die roll is a trade-off between tempo and card advantage--in type 1, either player may utilize cards which cause a large swing in either tempo (i.e. moxen, 3 sphere) or card advantage (many examples), potentially reducing the relevance of going first.  Alternatively, having tempo-boosters when going first or card advantage-generators when going second can increase the disparity between going first or second.

In decks packing 3sphere (or when playing against such decks), playing first may be more important because when you play first and drop turn 1 sphere, you can generate a huge tempo swing that may not be present if you were to drop a turn 1 3sphere when playing second.

[SliverKing]

I think going first is very important, but not enough to consider changing the system.

Agreed.  The system as it stands makes sure each player gets to play first if they choose to.  This way, somebody has to win on the draw to take the match.  It's fair enough for me.

Um.... not unless I've been playing wrong ...


Example:

You win the coin flip, elect to play first.

Win on your first turn.

I elect to play first game 2.

I win on my first turn.

You elect to play first game 3.

You win on your first turn.


You won the match, without having to win a game where you were drawing first.

(Clearly this is ridiculous hyperbole, but the point is the same.  The play/draw rule is probably as close to fair as we can get it, but it does create unfair situations from time to time).

[Comrade]

I was thinking about the die roll also as I happened to lose it a lot in Waterbury. Let's suppose the following course of events happen:

Player A & B, A wins the coin flip. A wins the game.

Ok, so A has won one game, and he also was granted his choice of playing or drawing by luck only (no skill involved).

Player B must win the next game or he has lost the round. Note, the chips are sitting with A right now even though B is playing first because A has already won one game out of 3.

Let's suppose Player B wins game 2. Then Player A gets to decide to play or draw AGAIN! Not only did he or she get granted extra luck for game#1, but A also gets a bonus decision for game 3!  Clearly, I think the coin flip awards the winner too much. I would propose the following "quick fix":

Round procedure
1. Flip to determine who goes first game 1
2. If Player A wins round 1, skip to step 4.
3. Player A wins round 2. (despite not having decision of play||draw)
    Player A may have won, but at least Player B got to decide game #2.
    Round determined.
4. Player B wins game 2.
5. Flip to determine who goes first game 3
6. Round determined. Note that player A doesn't get a free decision here.

[Ishi]

In chess, at the grandmaster level, the person who plays first wins about 55% of the time*.  To combat this, the tournament directors determine who gets white and black, and the tournament is paired such that players alternate between getting white and black (as much as possible).

What would people think about using a system like this in magic?  If people think the dice roll is critical in Vintage, this seems like it would be a good improvement.

* To be precise, it's more like he wins 35% of the time, loses 25% and draws 40%.

[Dr. Sylvan]

It's also good to note that the test occured when Extended was in crazy Tinker combo mode. It's not like the format was much slower than Type 1 at the time, and in fact, it could have been argued to be even faster due to the lack of Force and StP.

I was under the impression that they did that experiment pre-Mirrodin, at a GP in the season following Justin Gary's Oath win, but I am too busy at work to dig up the URL.

I have tried to encourage this be added to matchslips and reported, but it hasn't really gone over well. I guess it's not considered to be that important to winning/losing. (The only time when it is important is when combo is too good, meaning it's time to break out the BANHAMMAR anyway.)

[Machinus]

Honestly, I wouldn't be surprised to find that the exact same results obtained for T1 - that is, people on the draw won 50-55%.  I also wouldn't be surprised to see a stat that said people in t1 won 50-55% of the time playing first.  

I think going first is very important, but not enough to consider changing the system.

Steve, I don't think that was the implied intention of this post, or of posts or questions like this. I think if there is a statisical correlation, then it stands to benefit players to play decks that take advantage of this. For example, if three 100+ player tournies in MA, NY, and OH found correlations close to 60%, then it would be in the best interest of any player in those metagames to play shops, or whatever deck was winning, if they want to win themselves.

I don't think very many people want to change the way things are done with regard to going first. The mulligan rules along with the built-in tradeoff between the tempo and additional card are very good, and it is only in an examination of deck potential that this sort of discussion comes up.

I personally don't think it is really a significant factor when talking about workshop.dec. Storm decks exploit this to a much higher degree, and I think we already know there is an acute correlation there.

[Azhrei]

You can't look at individual duels so much as whole matches. I think that looking at the match as a whole, going first initially in Vintage is HUGE. With the ability we have to literally end the game on the first turn or to create an insurmountable lead on the first turn, going first is HUGE. It's not everything, but it's probably a good 25-33% of the match result.