Collecting MTG Arena: Ikoria Edition

---

With the release of Ikoria on April 16th, MTG Arena drafting underwent a complete metamorphosis. Person draft finally, and ahead of schedule! 3-match Traditional draft! One-off reduced price Sealed events! New names for formats! Also, Godzilla! Also, you can't turn off Godzilla! (When drafting or browsing your collection, at least.)

I'm here to cut through all the hype and figure out the most important question: what should I be doing to collect this set most efficiently? In this time of a) financial uncertainty and b) everyone being stuck at home playing Arena, this is what I really want to share with all of you. As such, I've redone my MTG Arena spreadsheet with some new features.

1) Updated spreadsheet

Link for updated MTG Arena spreadsheet v2.0 (google docs)

Link for updated MTG Arena spreadsheet v2.0 (excel)

-One single sheet now allows you to keep a complete log of all of your limited events, whether Premier draft, Traditional draft, Quick draft, or Sealed. It also has improved predictions for how many more events you need to complete the set, now separate for each type of event and Rares/Mythics.

-There's a space to tell it whether you paid with Gold or a Token, and it uses that to break down "how much you spent" into 3 categories: Net Gems, Net Gems ignoring the cost of Gold, and Net Gems ignoring the cost of Gold and Tokens.

-You can set your own personal Raredrafting percentage now! You can set it at 95% (usually for Quick drafts) or 50% (usually for Traditional drafts) or anything between 0% and 100%. You can play around with it and see very accurately how many more events you'll need to do based on how much you Raredraft, and see directly how much of an effect Raredrafting has. It's nifty!

-There's a sheet to record all of the Mythics you have in standard. If you've been following this advice, you'll already have all of the Rares, so it's useful to track which Mythics you're still missing.

2) Critical point equation updates

Last year, we introduced the critical point equations for predicting how many events you need to get a complete set of Rares (and Mythics).

R = Total number of Rares of that set already in your collection
P = Total number of reward packs of that set already in your collection
N = Number of "new" Rares you pull from a draft on average (Higher earlier, lesser later, but an average across the set is fine.)
W = Average number of reward packs from doing the draft.
D = Number of drafts you still need to do.

The formula for the Critical Point (solving for D) is now:

D = (53*4 - P*106/121*14/15 - R)/(N+W*106/121*14/15)

Previously I used 7/8, but the correct value to use for "how often do you open a Rare in a pack" is 106/121. Although it's super close to 7/8, we should be as precise as we can. Also, Arena updated how often you see wildcards in reward packs. It used to be 1/24 packs for both Rare and Mythic wildcards, now it's 1/30 each. So that changes the old 11/12 to 14/15. These combine for a ~2% increase in how many Rares you get from reward packs. (1.94%)

Same thing for Mythics, except the effect is smaller:

M = Total number of Mythics of that set already in your collection
P = Total number of reward packs of that set already in your collection
N = Number of "new" Mythics you pull from a draft on average (Higher earlier, lesser later, but an average across the set is fine.)
W = Average number of reward packs from doing the draft.
D = Number of drafts you need to do.

The formula for the Mythic Point (solving for D) is now:

D = (15*4 - P*15/121*14/15 - M)/(N+W*15/121*14/15)

Here, the change to the more precise 15/121 actually works against you. But the change to wildcards more than balances it out, and overall you'll see ~1% more Mythics from reward packs. (0.98%)

All of these updates have already been integrated into the new collecting spreadsheet, but it's worth pointing this out for others who use these formulas.

3) Graphs! (Or: The Meaty Part)

Here, we're going to examine how all 4 event types stack up against each other in 4 important categories, using the best data we have. Please note that the x-axis for all graphs is identical; it's your personal game win %, spanning from 45% to 75%.

a) Length of Event (Games Played)

When we examined this last time, Traditional was 5-matches, and took an insanely longer amount of time to play vs. Quick draft (then called Ranked draft). At 60% wins, you needed to play 11 games on average vs 6.5 games for Quick. Plus sideboarding time! On average you played 65% more games. Prominent streamers were just playing the first 3 matches and then dropping from the event, because it really was that bad. Thankfully, things are much better now.

Now on average you only play 10% more games in Traditional. That's a huge difference from 65%. Including sideboarding time, it may still take you a little bit longer, but now it's very comparable. In fact, as you can see above if you have 67%+ wins, you'll actually play LESS games overall than in any single game event! (Quick, Premier, or Sealed.)

b) Rares Seen

While the data for how many Rares you see in a draft is noisy, the average converges pretty quickly. After watching a LOT of streamer past broadcasts for Ikoria drafts, the values below are the average amount of Rares you SEE in a draft, regardless of whether you take that Rare or already have 4 copies of it.

Premier Drafts - you see 6.7 Rares and 0.47 Mythics per draft

Traditional Drafts - you see 7.6 Rares and 0.60 Mythics per draft

Quick Drafts - you see 3.66 Rares and 0.42 Mythics per draft

I'm not sure how well these values will hold up for other sets, but these are the values for Ikoria, assuming this article or something else doesn't radically alter player or bot behavior. To these numbers we add the number of reward packs you earn per win % to get the following graph:

c) Net Gems

This is a pretty straightforward update to what we did last time.

As you increase in win %, you obviously start netting gems instead of spending them. For Quick draft, the event is "free" at 74.8% wins. For Premier, the event is "free" at 67.7% wins. For Traditional, the event is free at 64.2% wins. (Note: We only use game win %, not match win %.) I was surprised that Traditional seems to be clearly the best choice if you're an above average drafter and want to play as much limited as possible for as cheap as possible.

d) Gems Spent to Acquire one Rare <4

All 3 graphs above contain no significant assumptions. However, what we are most interested in is slightly different - how many Rares <4 you acquired during the event (or from its reward packs). So for this final graph, Rares Seen needs to be reduced in 2 ways, and we need to make some assumptions. The first reduction is your Raredraft % - even the most hardcore Raredrafter occasionally passes a Rare, and with the new draft types, it can be correct to pass much more often. The second reduction is to account for seeing 5th copies of Rares as you near collecting the set.

For the graph below, I have the following assumptions, but it may vary for you.

Rares 5+ : For the past 3 sets, I saw on average 12% of my Rares were 5th copies when I drafted all the way to a complete set. So the factor I use is 88%.

Quick Draft Raredraft: 80% (This means 80% of the time you will take a Rare if you have less than 4 of it in your collection)

Premier Draft Raredraft: 60%

Traditional Draft Raredraft: 50%

With the above assumptions, if you have 54% wins, you'll look to do Traditional drafts, and Quick drafts if you're below that. But the assumptions can make a big difference, and you may want to alter your Raredrafting in the set based on several factors.

e) The data! More spreadsheets!

Don't like my assumptions? Like messing around with data? If you're interested, you can mess around with it yourself:

Link for updated Calculate Arena Events spreadsheet v2.0 (google docs)

Link for updated Calculate Arena Events spreadsheet v2.0 (excel)

The data on gem and pack wins for each win % is fixed, but you can vary the Raredrafting to create graphs of your own.

4) Summary! (Or Showdown: Quick vs Premier vs Traditional vs their inbred cousin, Sealed)

That's a lot of data! And there's several other factors to consider. So let's break it all down.

-Quick drafts are only available for 2 weeks a month, and of course have bots passing to you. The bots this set are similar to every set since War of the Spark - they Raredraft like crazy, and on average you see 1 passed Rare each draft.

-Traditional drafts take 10% more time, plus sideboarding time. Maybe 15 extra minutes total? You also have to commit to playing an entire match at once, which often is the larger issue.

-Quick and Premier drafts use the ranked ladder, so theoretically your opponents start out easier and will get more difficult over time. Once you're Platinum, the difficulty noticeably increases.

-Premier and Traditional are pretty well-balanced against each other regardless of your win %, assuming you don't Raredraft significantly different between them. But Traditional offers better value if your win % is moderate to high, and you see more Rares and Mythics. And note that the advice from last time is unchanged: you should always take a Mythic <4 if you're a set collector! No exceptions.

-Premier seems the most popular format, and the queue times for it are noticeably shorter than Quick draft and Traditional draft. This is despite it not being the "best" choice at any particular win %, but it has a good mix of value if you Raredraft little and do well, and giving you a bit of gems back if you Raredrafted heavily and do poorly.

-If your win % is about 60%, then each Rare you acquire costs you 50 gems. Assuming you don't have a ton of Rares from other sources, that's ~$58 in Gem value to get a set, which is in the ballpark of what we calculated last time. Considering you get $0.85 a day in Gold if you get to 4 wins, it will take you 70 days to get $58 worth. So f2p collecting the Rare set is definitely possible in theory. (It's also possible in practice, because I know people who have.) If your win% is a lot lower, it's harder, of course. The balance between Raredrafting and "trying to win" is something the spreadsheet can help you with, but it's ultimately up to you.

-Sealed is absolutely terrible value. In April, as part of a Daily Deal they had Sealed tokens for 1800 Gems (instead of 2000). I did run some calculations with Sealed costing 1800 Gems, and it puts it where it should probably be already - very close to Quick draft on graphs c) and d), instead of languishing at the bottom. Basically, 1800 is the price they should be charging all the time if they want players to actually consider Sealed. I bought the one token, but I'm not sure I would have bought more than one, even if allowed to. At 1800 Gems, it's no longer shamefully lighting gems on fire, though. Bottom Line: Never play Sealed unless it's on sale, and even then, it's just a maybe.

5) How I approach set collecting now

Now that there are 3 kinds of drafts, and Raredrafting % is highly relevant, it's not easy to state a single path that works for everyone. It's no longer obviously "wait for Quick draft and Raredraft 100%" like it was previously. Honestly, it's very apparent that someone smart (Lee Sharpe?) was in charge of the changes here. They carefully balanced things so that no matter which event you play, you can't go TOO wrong. Except if you play Sealed at full price - again, don't ever do that.

If we could define a precise relationship between your Raredrafting % and your win %, we could write down a specific program that would be guaranteed to be most efficient way to collect the set. While there obviously is some relationship between Raredrafting more and winning less, I believe there's no easy way to collect data that would allow us to calculate it. All I can imagine is having the same person draft a set 50-100 times with 100% Raredrafting, and then do it again 50-100 times again with 75% Raredrafting, then do it again with 50% Raredrafting. And then a bunch of other players of other skill levels would have to do that too. I don't think that's going to happen anytime soon.

But here's what I'm planning on doing for future sets.

A) Do Premier drafts until I become Platinum again, Raredrafting heavily (80%+) the first few drafts when I'm lowest in rank, and gradually easing off to 60%ish. It's hard to really aim for any specific percentage, but 60% is roughly "take a Rare over a medium playable for your deck". If Quick draft becomes available before I hit Platinum and my win % is terrible, I'll switch to Quick drafts and increase my Raredrafting to 100%.

B) Once I reach Platinum (or at least by Diamond, when it really heats up), switch to Traditional drafts. Just take the Rares I need when I can, hopefully 50-60% of them. Don't try and Raredraft it too hard, despite the high amounts of passed Rares. Pick non-crap Rares a bit higher than crap that you'll eventually see 5th-6th pick. Once it says I'm at 100% on the spreadsheet, I stop and crack my 100ish packs, and start saving up for next set.

C) That is, unless they run an Omniscience draft for the set. I really, really love this ridiculous coin flip of a format. I drafted M20 Omniscience 35 times in a single weekend, even though I already had every Rare in M20. That's how much I like it. I've really been missing Omni for ELD, THB, and now IKO, and I really hope they bring it back. Wizards, if you'd like me to spend Gems beyond set collection, that's all you need to do. Card styles and oddball constructed events just don't interest me the way Omniscience draft does. 

I recommend using a tracker that can tell you while you're drafting how many copies of a card you already have. I personally use Untapped.gg, but there's several others. Trackers are super useful for tracking Common and Uncommon collection, which is really not worth the hassle of recording on a spreadsheet. It also saves you from checking during the draft if you already have 4 of a Rare. However, note that this feature currently doesn't work with Premier or Traditional drafts because Wizards changed the output log (or rather, didn't add it in for person drafts). They did say they're working on it, so I assume it'll be back sometime before next set. Until then spreadsheets open while drafting!

Speaking of trackers, there are some out there that can do some of the things my collecting spreadsheet can do, but I still like spending the time to create and use the collecting sheet. You get to take control of what you see and what you record. If you don't care about a feature, you can delete it or hide it. Plus, I haven't seen one yet with all of these features, that can scale predictions with Raredrafting.

Anyway, Ikoria is a fascinating set, and there's a ton I could talk about. How much I truly enjoy the clever way they executed the Godzilla crossover (even though it annoys me a bit to see it while scrolling through my collection). How Mutate plays in some really neat rules spaces that have never been explored, and I thought probably never would be. How Companion pretty obviously should have had "When you take a mulligan, or if you decline to take any mulligans this game, put an additional card from your hand on the bottom of your library" as part of the ability. But there's a crack team of Seth and other contributors here at MTGGoldfish to handle all of that, so I'll stick to my deep numerical analysis and goofy humor articles. Til next time!