Christmas Chests opening trick data (with Screenshots)

Introduction

What is the trick?

It is a method to get better rewards per snowflake via the following steps:

1. Close the app entirely
2. Reopen the app
3. Claim one Christmas Chest (10 Snowflakes)
4. Repeat steps 1-3

Are you sure this works? Maybe because you believe the rewards are better so you only observe the good ones and ignore the bad pulls?

I have data here in the next section, see for yourself and decide.


Methodology and Results

My Setup

• 2300 Snowflakes
• 4 sets of 50 Chests each (because Prize Log shows up to 50)
• Close app as soon as snowflakes are subtracted (to save time, whether or not you watch the spin does not seem to affect the trick)
• Ran some extra experiments with the remaining 30 Chests (see Further Questions section)
  

Data

For each set of 50 runs, the pictures shown are in the order:

1. Bottom 6 rewards (note my blue slider is all the way at the bottom)
   
2. Next worst 7-12
3. Best 6 rewards
   

I took only one screenshot (Bottom 6) for my multiple-chest opening, but the trend certainly holds from past experience, and if you've done any you can take a look at your own worst rewards by scrolling to the bottom of your Prize Log.

Mass Opening


First 50


Second 50



Third 50


Fourth 50


Further Questions

Huh. Maybe you got really lucky 200 times in a row?

The statistics suggest that that is very very unlikely.

Let's define a trash spin as anything that gives below 50K Ag or 15 Au.
Now if you look at your own multi-opening, you'll see that trash spins are pretty common, I'd say about 20% at least. Now let's be very generous and assume an actual drop rate of trash spins of 10%. Using WolframAlpha,

(layman: how likely is it that I get less than 1 trash spin in a sample of 200 spins, if the chance per spin to get trash is 10%?)

(layman: for every 10 billion people who do 200 spins, you would expect 7 of them to get zero trash spins simply by chance)

Is it likely that somehow the many forum users reporting this phenomenon are simply experiencing a 7-out-of-10 billion chance? You decide.

Does this work if I don't close the app, but just use the 'Open Chest' function for 1 Chest at a time? Maybe the differential in the algorithm is between the single chest and multiple chest opening.

Nope, doesn't work, I tried it. On the third spin using this method I got 7 Au, whereas I had nothing below 15 Au for the previous 200 spins. You need to restart the app.

What about if I leave the Chest screen and return to Hangar, then back into Chest opening screen?

Nope, doesn't work, I tried it. On the second spin using this method I got 1 Au, whereas I had nothing below 15 Au for the experiment's 200 spins. You need to restart the app.

What if I play a game in between Chest openings?

To be honest I didn't try that, but I suspect you need to restart the app. Let me know the results if you try.
Jan Update: eh, it seems that playing a game between spins is akin to restarting and DOES work.

So how much Au and Ag did you earn from spending 2500 Au worth of $ on Snowflakes instead?

Since I did not pull any Gold items at all in 200 tries, the return is purely in Au and Ag.

After selling some of my weapons (I'm still selling them, it's taking a long while) and bots,
I gained ~1650 Au and ~10.7m Ag

These are the sell prices of the following:

• 58K Light Weapons (Pinata, Pin, Noricum)
• 87K Medium Weapons (Tulu)
• 116K Heavy Weapons (Thunder, Kang-Dae, Zenit)
• 122K Patton
  
• 126K Boa
• 128K Vityaz
• 130K Golem
  

And how long did it take you to spin everything individually and sell all your junk?

A few hours so far. Granted, I was multi-tasking, but you're limited by app startup time and sell time is a bit laggy too.

So is it worth it to buy Snowflakes for Chests?

No. Just buy Gold and get the items you want. Use this trick only for Snowflakes earned from battles.