Testing Preordain: Quantitative Results

The time has (finally) come to actually reveal the results of my latest banlist test. Looking back, testing two different decks made this harder than it needed to be. Focusing just on Storm would have yielded more satisfying data, though not a more significant result. As you will see, it appears that Preordain would not have that much impact on the top-tier metagame for a variety of reasons. Some of these I mentioned previously; a few will be explained here. However, this ultimately doesn’t matter. Other developments since I began this test ensure that Preordain is never being unbanned.

What I’m going to do is reveal the aggregated result. My questions were, “Is Preordain safe for Modern?”, and whether the overall data show if this is true. I’ll then break it down by deck and matchup to show how that result was achieved. What you’ll see is that Preordain did not significantly impact deck performance for either Storm or UW Control. This suggests that an unban is plausible. However, as I will get to later on, this result will not change Wizards’s stance, and I don’t anticipate playing Preordain in Modern in the foreseeable future.

Overall Result

I feel the need to start with this disclaimer: this is not a definitive result. The results I’m reporting are my experimental results and are meant to model the impact of unbanning Preordain on the Modern metagame. It would take many more tests with more decks to give a truly definitive result.

As a reminder, there were 640 total matches, or 320 with each deck. Play/draw alternated with each match regardless of result to ensure fairness and prevent bias. They were all typical matches—best of three with sideboarding. Please refer to the previous article for all the decklists.

  • Total Match Wins: 333
  • Total Win Percentage: 52%
  • Total Control Wins: 165
  • Control Win Percentage: 51.6%
  • Total Test Wins: 168
  • Test Win Percentage: 52.5%

As you can see, I didn’t have very impressive results. I’ll be going into why as I deal with each deck, but having Preordain didn’t feel very special. It was very similar to Sleight of Hand in Storm, and was inconsistently good in UW. I think we all know what the statistical test will show, but I’m going to include it anyway for academic honesty.

Once again, I’m reporting the z-test result because I think more people are familiar with it. As the P-value is greater than 0.05, we accept the null hypothesis and there is statistically no variation between the results. From this we can infer that Preordain had no real impact on my test decks.

Storm Results

Storm was something of an odd test for me as it really didn’t feel like an integral piece of the testing. What I mean is that the matches rarely came down to how I, as the Storm player, played. A few times a poor sequence hurt me, but for the most part the actual combo played itself. I know I wasn’t playing it perfectly, but Gifts Ungiven provided enough forgiveness that I didn’t need to. If that card resolves, you should always win. My losses were either caused by me mulliganing to death or my opponent’s disruption preventing me from comboing in the first place.

A note on sideboarding: Storm cannot afford to exchange many cards without severely harming its odds of comboing. I remember years ago hearing that Jon Finkel never sideboarded at all with Storm if he could help it, and who am I to argue with Johnny Magic? As a result I boarded as little as possible.

Grixis Shadow

I was told that Shadow was a very hard matchup for Storm. They have lots of relevant disruption and a powerful clock, the classic anti-combo recipe. This proved to be true, though Shadow has a hard time actually sticking a clock I found. They don’t have that many threats, so sometimes I was able to play the long game and come back from having my hand shredded.

  • Storm Control Wins: 15
  • Control Win Percentage: 46.9%
  • Storm Test Wins: 16
  • Test Win Percentage: 50%

With only a one-game difference between test and control, there is no chance that the result is statistically significant, which the analysis confirms.

P > 0.05, so accept the null hypothesis, there is no statistical variation in the data.

Sideboarding really didn’t change the matchup. Grixis had a pretty good gameplan pre-board, and it was still great after siding. There wasn’t much that Storm could do to change that other than go for Empty the Warrens more.

Storm’s Sideboarding:

-1 Grapeshot

+1 Empty the Warrens

Grixis Shadow’s Sideboarding:

-2 Lightning Bolt -2 Terminate -2 Kolaghan’s Command -1 Snapcaster Mage

+1 Grafdigger’s Cage +1 Nihil Spellbomb +1 Izzet Staticaster +2 Stubborn Denial +2 Collected Brutality

Eldrazi Tron

I thought this would be a worse matchup than it ended up being. Storm doesn’t fail with just one piece of disruption, so a single Thought-Knot Seer is not that bad. E-Tron sometimes just fails to do anything relevant except make a single big threat. Chalice of the Void was ignorable on one and often the game ended before they could put it on two. But when that did happen, it was game over for me.

  • Storm Control Wins: 16
  • Control Win Percentage: 50%
  • Storm Test Wins: 16
  • Test Win Percentage: 50%

Absolutely no change. Again, I don’t think the statistical analysis is necessary, but here it is anyway.

There’s no statistical difference between the control and the test.

I suspect that sideboarding had a much larger impact on the matchup than expected. E-Tron brings in a lot of great ways to shut down Past in Flames, meaning you’re forced to rely on Empty the Warrens, for which they have All is Dust and lots of creatures. Big Walking Ballistas were a nightmare, as was Wurmcoil Engine. We debated bringing in Shatterstorm for all the artifacts and ultimately decided against it. By the time you’d play it most games, you’ve already lost.

Storm Sideboarding:

-1 Grapeshot

+1 Empty the Warrens

E-Tron Sideboarding:

-4 Matter Reshaper -2 Karn Liberated -1 Endbringer

+2 Grafdigger’s Cage +2 Relic of Progentius +2 Warping Wail +1 Wurmcoil Engine

Counters Company

Counters was a really swingy matchup. Play/draw really mattered because you’re both combo decks and can kill on turn three. Game one was just a straight race, and Storm was more consistent. After boarding it got complicated. Counters has decent answers, and Storm does not, but it might just get locked out without Echoing Truth.

  • Storm Control Wins: 17
  • Control Win Percentage: 53.1%
  • Storm Test Wins: 18
  • Test Win Percentage: 56.3%

There’s a theme with these results. See if you can spot it.

Again, little has changed. Preordain isn’t important in a racing situation. It just ensured that you never fizzled, which is pretty rare anyway.

Since the goal was to win turn three, Storm didn’t sideboard on the play. On the draw you had to be the control deck, relatively speaking, so there was sideboarding then.

Storm on the draw Sideboarding:

-1 Baral, Chief of Compliance -1 Pyretic Ritual -1 Desperate Ritual -3 Remand

+3 Lightning Bolt +1 Echoing Truth +1 Anger of the Gods +1 Pyroclasm

Counters Company Sideboarding:

-1 Qasali Pridemage -1 Kitchen Finks

+1 Eidolon of Rhetoric +1 Orzhov Pontiff

Burn

I thought Burn would be a better matchup than it actually was. I didn’t appreciate how good Searing Blaze actually was against Storm. You’re reliant on your cost reducers to go off, and Blaze kills them efficiently. Burn also reliably goldfishes turn four and can turn three you if your mana cooperates, so they can race you. Also, Eidolon of the Great Revel is lights out. The only way to win with that on the field is to Empty the Warrens. And you’re probably dead anyway. We always played game one as if we didn’t know what we were playing, but for games two and three my Burn pilot aggressively mulliganed for Eidolon.

  • Storm Control Wins: 19
  • Control Win Percentage: 59.4%
  • Storm Test Wins: 17
  • Test Win Percentage: 53.1%

Not a big change again. I believe the difference was the Burn’s aggressive mulligans paid off a few more times against the test deck.

Again, not a significant result. Well within the “noise” of the test. I actually expected this. With a smaller n-value you need really disparate results to achieve statistical significance.

Sideboarding for Storm was hard here. You needed to remove Eidolon and couldn’t rely on Empty. In exploratory testing I found that you could Empty for a lot and still die so we decided to stick to the Grapeshot kill as much as possible. On the draw we decided to add more counters.

Storm Sideboarding:

-1 Desperate Ritual -1 Pyretic Ritual -1 Baral, Chief of Compliance

+3 Lightning Bolt

Additionally On the Draw:

-1 Empty the Warrens -1 Gifts Ungiven

+2 Dispel

For Burn we took out the clunkiest burn spell and Lavamancer for relevant disruption. We debated Kor Firewalker for a while and decided against it.

Burn Sideboarding:

-1 Grim Lavamancer -3 Rift Bolt

+4 Relic of Progenitus

Jeskai Control

Jeskai was another swingy matchup, mostly because their clock was what really mattered. Given the time, I would just sculpt to my heart’s content and win through their permission. The fact that this version didn’t have Geist of Saint Traft helped on that front, but Spell Queller was also a beating combined with all their burn.

  • Storm Control Wins: 14
  • Control Win Percentage: 43.8%
  • Storm Test Wins: 15
  • Test Win Percentage: 46.9%

With a different sideboard on Jeskai’s side I can see this matchup becoming much worse.

It’s very significant how not significant these results are. I really am running out of things to say here; it’s only going to get worse for UW.

For sideboarding we adjusted the counter suite for Storm while Jeskai really had the opportunity to adapt. The Empty plan is meant for disruption-heavy decks, but it really hates it when you prepare and have sweepers.

Storm Sideboarding:

-1 Grapeshot -1 Remand -1 Baral, Chief of Compliance

+1 Empty the Warrens +2 Dispel

Jeskai Sideboarding:

-4 Path to Exile -1 Lightning Helix -2 Electrolyze

+2 Negate +3 Rest in Peace +2 Supreme Verdict

Storm Conclusions

Preordain did not excel in Storm. It was simply too like Sleight of Hand, which it replaced, to have any significant effect. Where it was better was post-sideboard when you were digging for pieces and seeing only junk, but that didn’t happen too often. Most games I cantripped a few times then attempted to go off. Even against Jeskai the games didn’t tend to go very long. The opponent’s disruption and clock mattered more than the power of my cantrips. Therefore, I have no evidence here that Preordain would change anything for Storm.

UW Control Results

Testing UW was far harder. A control deck has more decisions and takes longer to finish a game, which is why this took so long, but it also required more of a play adjustment between the control and test decks. Serum Visions and Preordain are better at different things and expecting one to do the other’s job was disastrous in exploratory testing. As a result I had a harder time with the deck.

Grixis Shadow

UW has a pretty good matchup thanks to its redundancy. You can’t really stop their first few turns, so they will shred you, but you are likely to recover and draw more powerful cards as the game goes on. As long as you don’t just die to beefsticks you’ve got a great shot at out-valuing them.

  • UW Control Wins: 19
  • Control Win Percentage: 59.4%
  • UW Test Wins: 17
  • Test Win Percentage: 53.1%

You know by now where this is going.

We didn’t sideboard very much, both decks are close to where you want them maindeck. The adjustments were based on the assumption the games would go longer.

UW Sideboarding:

-1 Vendilion Clique -1 Spell Snare

+2 Rest in Peace

Grixis Shadow Sideboarding:

-2 Lightning Bolt -2 Terminate -1 Fatal Push

+2 Stubborn Denial +1 Liliana the Last Hope +2 Collective Brutality

Eldrazi Tron

Eldrazi was a weird matchup. Their deck is fairly inconsistent, and when I could use Spreading Seas to capitalize on that, it was easy. Against their good hands and/or Cavern of Souls, it got much harder. Playing only unconditional removal was very good as well. However, sometimes Eldrazi is just Eldrazi, and Chalice can prove backbreaking.

  • UW Control Wins: 15
  • Control Win Percentage: 46.9%
  • UW Test Wins: 16
  • Test Win Percentage: 50%

There’s not much to say really, it was just a slugfest.

UW is almost pre-sideboarded against Etron. I wished it wasn’t, as I would have liked more Detention Spheres for Chalices, but that wasn’t an option. I debated Spell Queller but it didn’t perform in exploratory.

UW Sideboarding:

-1 Spell Snare

+1 Supreme Verdict

Etron Sideboarding:

-2 Dismember -2 All is Dust

+2 Hangerback Walker +2 Relic of Progenitus

Counters Company

This was a weird matchup. Sometimes Company went for the long-game; sometimes it was just jamming the combo. UW never felt safe and it was a really stressful test.

  • UW Control Wins: 17
  • Control Win Percentage: 53.1%
  • UW Test Wins: 15
  • Test Win Percentage: 46.9%

Collected Company is a hell of a card.

I decided to target the Company value plan with my sideboarding, since that was their best card and I didn’t really have more ways to interact with the combo. Spreading Seas is not effective against mana dorks.

Company went for sweeper insurance.

UW Sideboarding:

-4 Spreading Seas -1 Mana Leak -1 Logic Knot

+3 Rest in Peace +2 Dispel +1 Supreme Verdict

Counters Company Sideboarding:

-1 Fiend Hunter -1 Vizier of Remedies -1 Devoted Druid -1 Qasali Pridemage

+3 Voice of Resurgence +1 Selfless Spirit

Burn

This went worse for UW than I thought it would. It plays less lifegain and fewer counterspells so it can be a struggle.

  • UW Control Wins: 17
  • Control Win Percentage: 53.1%
  • UW Test Wins: 18
  • Test Win Percentage: 56.3%

I know the percentage jumps look big but that’s just a quirk of small n samples.

Sideboarding is what you’d expect: dead cards out, counters in. The Quellers were pretty good here as both disruption and a clock. You can’t wait forever against Burn. Spreading Seas is too tempo-negative to play early, and late, it’s not relevant disruption.

UW Sideboarding:

-4 Spreading Seas -3 Supreme Verdict -2 Jace, Architect of Thought

+4 Spell Queller +2 Dispel +2 Timely Reinforcements +1 Negate

Burn Sideboarding:

-4 Searing Blaze

+4 Relic of Progenitus

Jeskai Control

We played this matchup as a control deck against a midrange deck. Neither I nor my Jeskai pilot were sure that’s correct, but nothing else made sense at the time. It was weird because most of their cards aren’t good but can still kill you if unopposed. Blessed Alliance was shockingly good as a result.

  • UW Control Wins: 16
  • Control Win Percentage: 50%
  • UW Test Wins: 20
  • Test Win Percentage: 62.5%

This matchup was the closest to actually significant results I got. I think if I had done the usual 50 it would have been significant for reasons I’ll describe below.

Sideboarding in control mirrors is hard. I decided that his Snapcasters were better than mine and that I wanted to fight on his turn to resolve planeswalkers. I also didn’t want to just lose to burn.

UW Sideboarding:

-4 Spreading Seas -3 Supreme Verdict -2 Condemn -1 Blessed Alliance

+4 Spell Queller +3 Rest in Peace +2 Dispel +1 Timely Reinforcements

Jeskai Sideboarding:

-2 Lightning Helix

+2 Negate

UW Conclusions

The Jeskai test revealed why my results didn’t change much from control to test. Preordain is a mid-game card, and Serum Visions is an early-game card. What I mean is that during mid-game topdeck wars, Preordain is better, because you can find and cast the card you want right way. Visions gets you deeper, but you get a random card. This is great in the early game where you want to hit land drops and set up your turns. When games end quickly, Preordain doesn’t get the chance to shine.

Preordain’s Place

Based on my results and experience playing the card, I do not believe that Preordain is unequivocally better than Serum Visions. During the first few turns, the smoothing power of Visions is far superior, and if you want a card to set you up for the long game, you would always choose that card. However, when you need to find something right now, Preordain will be your go-to. As a result, I don’t believe that they necessarily fight for space, nor would you always play sets of both. A mix is more likely. With this in mind and my lack of significant results, I believe that Preordain could be unbanned.

Some Caveats

I wasn’t testing decks that overload on cantrips. This was a deliberate decision to keep this as scientific as possible. If I start wildly redesigning decks, then the test becomes more about my deckbuilding ability than the actual strength of the cards. As I’ve said from the beginning, it’s better to use an established list and see how the card boosts its power. So I didn’t play 12-cantrip Storm or Serum Visions and Preordain in UW Control.

In Storm, I’m certain this was fine. I’ve played heavy cantrip Storm in Modern before, and the Gifts version feels better. Having a way to search for mana and Past in Flames was very good, and I can’t fathom cutting that package. Players have argued that I should just cut the utility spells for extra cantrips, but I’m skeptical. As noted, Chalice and Eidolon win the game against you, and having a few ways to answer them is necessary. The cantrips are still weaker than Legacy’s; a single Echoing Truth is not going to cut it.

As for UW, I’m not certain. Finding the room for cantrips requires cutting real cards. Modern is faster than Legacy, so you can’t really durdle or fill your deck with air, especially as a control deck. Miracles got away with that thanks to Counterbalance. Maybe it would be correct, but I’m uncertain. In any case, trying to find out adds more variables and is therefore untestable at this time.

It Was All for Opt

The problem is that nothing I’ve just said really matters. This has nothing to do with its value or the process, but with Magic moving on while I’ve been working. See, Opt has effectively killed Preordain‘s chance to be unbanned. The first reason is similarity. Opt is Preordain, adapted for instant speed. The effect is weaker, but it gains speed. Standard Wizards balancing strategy. Yes, I know Opt came first, but it wasn’t appreciated in its time. If you have to nitpick, just flip my statement around; it’s still true.

This feeds into the other problem. Wizards has previously said that too many cantrips is a problem. They’re worried both about consistent combo and overly consistent control (à la Caw Blade). They’re fine with a few weaker cantrips, but add some more power and things get risky. As a result, I think Opt is a definitive statement to the Modern crowd that the cantrips won’t be unbanned. Wizards will not risk cantrips killing variance again.

In the end, that’s my conclusion. Preordain is not necessarily better than Serum Visions, and would be a worthwhile risk to unban. This will not happen because Opt subsumes Preordain‘s theoretical place.

See you next week for the results of fitting Ixalan into Merfolk.

David began playing Magic during Odyssey block, quit playing Magic when Caw Blade ruled the world, and returned to Modern shortly before Deathrite was banned. He’s made an appearance at the Pro Tour, made money at GP Denver, and is constantly grinding and brewing in Modern.

5 thoughts on “Testing Preordain: Quantitative Results

  1. Thanks for taking the time to do this. I’m somewhat surprised by how similar Preordain performed compared to Sleight of Hand – it seems that the only thing to truly hold it back from being unbanned is fear of 12-cantrip Turbo Xerox decks, but that seems unlikely – Opt’s release has come and gone, and Storm pilots collectively shrugged. Preordain is definitely better than Opt in that shell (it digs one card deeper, and Storm can only go off on sorcery speed anyway), but it’s all feeling like the card should be a pretty safe unban.

    1. It’s possible that in different shells I would have had different results. When Preordain was axed it had always been played alongside Ponder so we never had a good sense of it’s actual powerlevel, though we know just how good Ponder is thanks to Legacy. If any banned card is guilty by association, it’s Preordain. I’d like to see if it’s actually worthy on its own, though like I said Opt makes it less likely.

      1. I think the card keeping Preo on the banned list, if any, is actually Sleight of Hand, not Opt. Preo is a strictly better Sleight, and I can’t imagine decks would play both instead of a Serum and Preo combo. So Preo would straight replace Sleight and make that card obsolete, which does not lead to any kind of net diversity gain. Since Modern looks good right now and Sleight is a played card, I doubt Wizards touches Preordain.

Leave a Reply