The Platinum Wall: When Playing On Gets Expensive

Yesterday I played a Draft of  The Hobbit (HOB) , my first of the September season after taking a few days off from MTG Arena. Truth be told, I was feeling a little burned out after the poor results at the end of August.

I started the season really well: both my last few  Marvel Super-Heroes (MSH) Drafts and my first Hobbit Drafts went pretty well. In fact, I even wrote a slightly euphoric post about it: Having a Dwarven Good Time!. I climbed the ranks fairly quickly, all the way to Platinum 4 — and that’s when the problems, and the gem burning, began…

An uncomfortable hypothesis

After going through my 17Lands history, a fairly simple idea occurred to me: maybe once I hit my current rank ceiling, continuing to play Premier Draft stops being a good decision if my goal is to preserve gems.

That doesn’t necessarily mean I should stop playing. If I want to improve, at some point I need to learn how to compete against better players.

But maybe I was mixing up two different things: playing for results and playing to practice.

To test the idea, I gathered every Premier Draft I have recorded on 17Lands since May: 45 Drafts and 252 games.

Here are the numbers by starting rank:

It isn’t a perfect ladder. In fact, my worst result after Platinum is Bronze, while my best rank is Silver, so it would be far too easy to claim that every step up the ladder simply costs me a fixed number of percentage points. On the other hand, although I can’t prove it, I also have a theory for why Silver looks better than Bronze: when a season starts, ranks reset, so you can run into very strong players who are still climbing…

Either way, one thing does stand out: Platinum is clearly different from the rest. Below Platinum, my record is 93–81 (53.4%). At Platinum, it drops to 30–48 (38.5%).

That’s a difference of almost 15 percentage points.

August, before and after the wall

If we look only at the August season — by far the month in which I’ve played the most — everything lines up. I started at Bronze 2 and, after ten Drafts, on August 13 I went 5–3, which took me from Gold 2 to Platinum 4.

Up to that point, my season record was:

31–27 — a 53.4% win rate.

Nothing spectacular, but it was positive. More importantly, I had already reached what was probably a realistic goal for that season.

I could have stopped there.

I didn’t.

I played 16 more Drafts, all of them starting in Platinum, and the result was:

30–48 — a 38.5% win rate.

Of those sixteen Drafts, five ended 0–3, and twelve ended with two wins or fewer.

The cumulative graph tells the story better than I can:

After reaching Platinum I played 78 more games, lost a considerable amount of gems and, by the end of the season, I was still exactly at Platinum 4.

From a rank and bankroll perspective, it’s hard to argue that this was a good investment.

But correlation is not causation

This is where I should slow down before declaring that I’ve discovered a new universal law of Magic Arena.

The sample is still small and, more importantly, there are several variables mixed together.

For example, every Draft I started in Platinum was The Hobbit (HOB). Maybe rank wasn’t the only problem; maybe I still didn’t understand the format well enough, maybe I drafted worse during that stretch, or maybe I simply ran badly.

There’s also a difference between looking at aggregate results and looking at the actual quality of play. An 0–3 can contain three extremely close games, while a 7–0 can include opponents who never drew their fourth land.

So I don’t think these numbers prove that Platinum makes me lose.

What I do think is reasonable to say is that Platinum is currently the point where my mistakes stop being free.

In Silver, I can have a mediocre Draft and still scrape together a 4–3 or a 5–3. In Platinum, it feels as though a bad drafting decision, a questionable mana base or a suboptimal line of play turns into a loss much more often.

And that fits pretty well with what happened after the reset.

Yesterday, September 10, back at Silver 1, I played HOB again and went 7–2.

One Draft proves absolutely nothing, but it’s hard to look at that result without thinking: “maybe the set wasn’t the only problem”.

Playing to win, or paying to learn?

The tempting conclusion would be very simple:

When I reach Platinum, I stop playing until the reset.

If my only goal were to maximize gems, that would probably be a fairly sensible strategy. But I’m not convinced by it either. If I avoid Platinum because I lose there, I’ll never stop losing in Platinum. At some point I need to play against opponents who punish mistakes that go unnoticed in Silver or Gold.

So maybe the solution isn’t to stop playing, but to change the reason why I’m playing.

Until I reach Platinum, Drafts are part of my bankroll: I’m trying to compete, climb the ladder, and have the prizes pay for the next entry.

Once I reach my ceiling, Drafts become practice. And practice has a cost. What doesn’t make much sense is pretending those two kinds of Draft are the same thing.

A budget for making mistakes

Next season I want to try a very simple rule: when I reach Platinum, I’ll set aside a small practice budget. Maybe three Drafts. Maybe five. Enough to keep playing against stronger opponents and identify what isn’t working, but small enough that a bad run doesn’t turn into sixteen consecutive Drafts spent trying to win back the gems I just lost.

More importantly, the goal of those Drafts will need to be different.

Not just looking at whether I finish 2–3 or 5–3, but reviewing the picks, rewatching close games and trying to identify recurring mistakes. If Platinum really is my current ceiling, that’s where there is the most to learn.

In the end, the August summary can tell two completely different stories.

The first is fairly depressing: 26 Drafts, a 44.9% win rate and 13,130 gems lost.

The second is a little more interesting: I started at Bronze 2, reached Platinum 4 and, once I got there, discovered that the next jump in skill isn’t simply about playing more.

Now I need to learn to play better.