Snapalytics: More Cubes on Tuesdays

Have you ever been playing Marvel Snap at an unusual time and experienced unusually good results? It’s happened to me a few times! Was it luck? Was I just unexplainably on point? Or could it be that the competition in my matchmaking region is less fierce in the wee hours of Saturday mornings?

Today’s topic seeks to answer those questions. The idea was suggested by Bishop, a member of the Marvel Snap Zone community on Discord who has experienced a similar phenomenon. He wondered if there was any basis to it.

Read on to learn the impact that time has on your wins (and, more importantly, your cubes).

The Dataset

The data I’m covering today is pretty simple to explain! As always, the data comes from match results recorded by our Marvel Snap Tracker. I looked at matches from both Ranked and Conquest modes. The time covered includes the entirety of last month’s season (Rise of the Phoenix) and the first week-and-a-half of the current season (Big in Japan).

I’ve broken up the data from Ranked mode by player Rank. The most obvious boundary to draw is between Infinite (100+) and the rest. Once you hit rank 100, you can’t go below 100 for the rest of the season and you’ve earned every reward based on rank. The game totally changes. Snapping and retreating aren’t nearly as meaningful. I’ve split the pre-Infinite ranks into two categories: 30–69 and 70–99. This is somewhat arbitrary, but we do see different statistics in the lower grouping versus the higher grouping. A reminder that players lose about 30 ranks between seasons, with players who reached Infinite starting the following season at rank 73.

For Conquest data, I’ve separated data from Proving Grounds for a similar reason that Infinite ranks are pulled from Ranked data: the stakes are just completely different in Proving Grounds. Snapping and retreating don’t carry the same weight when you have nearly nothing to lose.

Time Variance Analytics

Ranked: Day of the Season

Let’s get right into it! First up, a look at how quickly cubes are brought in based on how far into a season we are.

Wild, right? You may already be familiar with this phenomenon if you hang out with the Marvel Snap community on Twitter. Every season, there’s a whole bunch of players who knock out the entire climb to Infinite in less than a week. That effect seems clear here, especially if you look at the pink line representing ranks 70–99 (remember that players who reached Infinite the previous season start the next season at rank 73).

The chart shows high cube income for the first 2–3 weeks (cube rate represents the average number of cubes won per game, by the way). After that, it kind of levels off. It’s worth digging into what’s really happening there!

I’ve seen a lot of players describe their first climb to Infinite as quite challenging. That was definitely true for myself. I didn’t reach Infinite until my fifth season playing the game (which happened to be April’s Hit Monkey season). That climb really took the entire five months to gradually improve my starting rank for each season, and then lots of effort in that final season before finally reaching rank 100 and claiming my prize of a golden card back. Having reached Infinite once didn’t suddenly make it easier to do again. I actually missed Infinite the following month (Nebula season), peaking around 95.

Variants

Hit Monkey

On Reveal: Gain +2 Power for each other card you played this turn.

Series 4 Rare – Collection Level 486+ (Pool 4)

view card details Start a deck Decks with Hit Monkey

But when you hear the first-week Infinite players talk about it, they make it sound easy. In fact, for this latest season, I saw a bunch of people describe it as their easiest Infinite yet, due to all the bots along the way. Meanwhile, there are many players who struggle for every single cube they win. Their climbs feel stuck — plateaued anywhere from rank 10 to 99.

What’s the deal? I’ve seen people speculate that you have to push through the challenging lower levels of matchmaking before your matchmaking rating (MMR) gets high enough to earn an easy bot-alicious climb.

It’s certainly counterintuitive. Is it true? I don’t know. I can tell you what the data shows, and I can tell you more about my own experience.

The chart shows that players using the Marvel Snap Zone tracker win a lot more cubes early in the season. That’s all we know for sure. I think it’s reasonable to assume that players using a tracker are generally more serious and more skilled than the average player (remember a huge portion of the Marvel Snap player base is casual). That’s important to keep in mind as we look at data from our tracker — it’s not representative of an average player.

Based on anecdotal evidence, it also seems reasonable to assume the sharp spike at the beginning of the season comes from that set of players who hit Infinite quickly. At that point, they either move on fully to Conquest mode, or else their efforts in Ranked mode move to the blue line (rank 100+). When the chart levels out, pretty much the only players left in the pre-Infinite ranks are the strugglers and the slow-and-steady climbers. It’s notable they’re still bringing in cubes at a rate that’s low, but meaningfully above 0 (with ranks 70–99 at an average of 0.2 cubes per game by the end of the season).

Let’s take a peek at win rate, too.

This chart shows a similar effect. There’s a spike up front, and then it pretty much levels out after 1–2 weeks (depending on which rank you’re looking at). I imagine the massive initial spike for Infinite represents players who are playing against almost exclusively bots before more players join them in Infinite (since pre-Infinite and Infinite players can’t match into each other any more).

Back to my own experience with Ranked mode. The Nebula season was the last season I missed Infinite. I’ve been three-for-three since. And each has been progressively easier. I don’t attribute that entirely to MMR shenanigans, though. I got some good guidance from our very own PulseGlazer. Between his advice and lessons I learned in Conquest mode, I’ve really changed the way I play.

Variants

Nebula

Each turn your opponent doesn’t play a card here, +2 Power. (except the turn you play this)

Series 5 Ultra Rare – Collection Level 486+ (Pool 5)

view card details Start a deck Decks with Nebula

I think the main key to reaching Infinite is actually learning how to play Ranked mode properly. It took me 7 months to do that. You have to do that in order to get to the point where you can reliably defeat a bot (if you can’t, it’s not helpful to queue into a slew of ‘em). You have to do that in order to reach the maybe-magical MMR that enables a speedy ascent to Rank 100.

I think there probably is something funky happening at high MMR that makes the Infinite climb easier. There’s too much anecdotal evidence to refute that. This season was my fastest Infinite to date, too, taking a little over two days. It was my fifth Infinite, but the first one where I felt like I didn’t even need to employ the skills I’ve been developing.

It’s not really what this article is about, but it feels silly mentioning important changes to my gameplay without saying a word about what those changes were. I consider these critical to enabling me to hit Infinite consistently. Maybe some of them will benefit you!

Climbing Tips

1. You’re probably playing too aggressively. I’ve actually known several people who got their first Infinite after deciding to never snap. Now, I didn’t take it that far, and I wouldn’t advise you to. However, I started retreating if my opponent snapped and I didn’t already have the answer in my hand. Previously, I lost many 4- and 8-cube games because of staying in hoping to get lucky despite terrible odds, or worse — just to see what happens. Take ego out of it. Retreat if you’re probably going to lose.
2. Pick a deck and stick to it. Sure, it’s okay to change your deck every now and then (especially if the meta seems to be stomping on you!). However, your skill with a deck will grow the more time you spend with it. You’ll learn which matchups are in your favor. You’ll learn when your deck should snap, and when it should retreat.
3. If you’re not able to play seriously, go play in the Proving Grounds. Don’t burn your hard-earned Rank by taking a break with a “just for fun” deck or by playing when you can’t focus. Seriously, I can’t stress enough how important silly breaks in Proving Grounds were to my climbs!

Ranked: Day of the Week

Is it possible the day of the week can impact player stats? The game does have a series of events that happen on fixed days. New seasons, new cards, and new patches drop on Tuesdays. Hot locations on Sundays. Featured locations on Wednesdays. OTA updates on Thursdays.

When I first grouped the data by day of week, the results were astounding.

That’s a massive effect on Tuesdays! I thought maybe it was showing the impact of new card releases, and then the meta figuring out how to deal with it afterwards. It was really exciting data, but the impact seemed too good to be true and… it was!

I realized that new seasons start on Tuesdays. Remember the spike we saw at the beginning of seasons? That factor needs to be separated. Look what happens if we separate weeks within the same season. This is for ranks 30–69 only. Each line represents a different week number within a season.

The Tuesday effect gradually gets weaker and weaker until it’s flattened out in week 4 and especially 5. And each week gets lower than the preceding week. So technically the most cubes are earned on Tuesdays, but it looks like the phenomenon almost entirely has to do with how long the season has gone on. Tuesdays end up higher overall because Tuesday is the first day of in-game months and weeks.

A reminder that we’re only sure of this effect existing among players using our tracker. That said, it’s a safe bet it scales to the entire player base because bots exist in ranked mode. And there are always more pre-Infinite players on Tuesdays than on any other day of the week.

The dampening over time is even more pronounced in ranks 70–99. Look how much faster it flattens out! I think what’s actually happening here is the effect of people graduating out of the rank range we’re observing. Take the chart below. Players stop contributing to the chart as soon as they hit rank 100. So, the players raking in the most cubes represent a significant portion of the players reaching Infinite at any given point during the season. Once they hit Infinite, they’re no longer earning cubes for the chart’s rank range, and the average cube income for rank 70–99 constantly decreases.

It’s an effect that totally makes sense, but it’s interesting to see it!

Another thing that makes sense — the players who reach Infinite fastest have the highest cube rates. Whether that effect comes from skill or the existence of bot-heavy MMR’s, the impact is significant. At the beginning of the season, the average cube rate across all ranks 70–99 is 0.9. The top players have cube rates much higher than that — we even recorded a couple players with cube rates over 3.0 across 50+ games! 🤯

Just to be clear, that’s extremely unusual. A player with a cube rate over 1.0 is rare enough, and there are 20 times more of them than those with a cube rate over 2.0.

As stated previously, players who hit Infinite the previous season start the following season at rank 73. After that, a bonus 3 ranks are awarded every time a player reaches a new rank tier (e.g., when you hit 80, you are bumped to 83). That means a player who starts at 73 has only 21 ranks to climb before they hit 100. At 7 cubes per rank, that’s 147 total cubes needed.

For some quick perspective, those 147 cubes take a lot longer for a player earning 0.2 cubes per game (735 games) versus a player who’s earning 1.2 cubes per game (123 games). Slow and steady is a perfectly viable strategy — I won my first two Infinites that way.

Finally, just for fun, the same chart for rank 100+. It takes all of 3 days to flatten out! I imagine the spike comes from the fact that in the first couple days, there aren’t many players to match with in Infinite mode, so those who get there fastest get matched with a lot of bots. Just a guess. Beyond that, I don’t think it’s very interesting or useful to study Ranked mode at the Infinite tier.

Ranked: Hour of the Day

For Day of the Season and Day of the Week, I kept the time relative to when the in-game day resets (think of the time your variant shop resets). For this section, Hour of the Day, I have converted the game time to the player’s local time as accurately as possible (I approximated based on region when time zone was missing).

Is it worth waking up early or staying up late to win extra cubes?

I think it’s most interesting that the biggest spike for ranks 70–99 happens when the other ranks experience their biggest dips. There don’t seem to be any times that are especially good, but whew!— y’all should stay out of Ranked mode between 2 and 5 in the morning! Why are ranks 70–99 seemingly immune to that effect? No idea. It’s weird, right?

For the sake of completion, here’s the same chart for win rate instead of cube rate.

Conquest Mode

I did examine all the same time-based metrics for Conquest mode. It seems like Conquest mode is far more resistant to any time effects, including that big start-of-season edge. Perhaps the absence of that effect is due to the effect being bot-driven (bots exist in Ranked mode to a much greater extent than in Conquest mode)?

I’ll share the cube rate charts. Trust me the win rate charts are very flat and boring.

Conquest: Day of the Season

Conquest: Day of the Week

Conquest: Hour of the Day

Conclusion

Well, it turns out there were no big revelations on this topic — it seems that time is mostly not a significant effect — I still thought it was an interesting exploration! I especially like the fun fact that players win more cubes on Tuesdays (even if it’s not in the way you’d first interpret that statement). My apologies if you ended up feeling suckered in by that title. 😅 I ultimately decided it was fine because it’s technically correct. That’s the best kind of correct!

Ultimately, I’m glad for the results. Wouldn’t it be weird if you could get a competitive advantage just by playing at a certain time? Instead, you can rest easy when it comes to the time factor. Well… you should probably avoid trying to climb at 2–5 o’clock in the morning! 😜

I hope you all have a great week! If you’ve got an idea for a cool data- or math-heavy Marvel Snap analysis, I’d love to hear it! Please share it with me in the comments below or else on Twitter! 💜

Now, to figure out why I was playing so on point that one time I stayed up way too late… 🤔

This analysis was powered by the data from our Marvel Snap Tracker. If you play on PC or Android, consider giving it a download! It has many features including a live overlay, the ability to sync your collection to our site (letting you review your personal stats), and powering useful tools like the Meta Snapshot.