Best Of

Arenas and doing all the EQ/back issues. I also pay for the daily unit ($5 isn’t bad considering it’s less than fast food).

I save up to 1m battlechips and pop them into units. You may be wondering, “gee rev, which crystal is best worth my time for units?” So far I can say it’s the Uncollected because it’ll save you some sanity. Here has been my luck with battlechips so far:

ThroneBreaker per 1m battlechips:
12/20 - 3k units
1/23 - 4,125 units
2/14 - 1,875 units
3/6 - 375 units (this one hurt)
3/27 - 1,875 units
4/15 - 1,125 units
5/6 - zero units. Almost threw my phone against a wall. This is where I shifted to Uncollected.

Uncollected per 1m battlechips:
5/27 - 1,800 units
6/16 - 2,175 units

There you have it. Farm arenas, pop those uncollected crystals.

Wow the fact that @DNA3000 disproved your claims a couple days ago in a different thread not related to this topic before you even posted this is insane 4d chess stuff.

Anyways you are incorrect

I don’t spend time here so my apologies for missing that post. You are entitled to your opinion. However, the crystal yields were so consistent through Mar 2024 that it was palpable. It had flattened so much I almost stopped tracking the data. Then it pivoted - hard! The unit returns dove. The gold remained as consistent as ever. It still is today. The unit returns never returned to a stasis. They are all over. Something happened. Data doesn’t lie.

It is interesting you find the sudden change to be suspicious, and not the fact that the data was "flat" for an extended period of time while buying crystals in only 100k BC batches. That would be only ten TB crystals, which cannot possibly have been generating consistent results Even fifty UC crystals is not enough to generate consistent results: your data should have been wobbling all over the place.

You say about 12 million BC worth of data across about 34 months (August 2022 to present). Assuming roughly even amounts of BC over time, that would mean the data from March 2024 to now would be about four out of the 34 months, or about 1.4 million BC worth of data. Depending on which crystals you're opening, that's not a lot of data. It is 28 TB crystals, 140 UC crystals, or 700 basic crystals. Even 700 is a relatively small sample size. Here's all of my UC crystals, opened from February 2020 to April 2021, in terms of units per batch of one hundred crystals:



The units per batch swing up and down, as you'd expect. On average only 15% of crystals contain units, so even in a batch of a hundred you'd only expect about 15 unit drops. Very small changes to that number due to random chance are going to cause the units per batch to also change by a lot. And if you look at the seven batch moving average, you can also see that the average units per 700 crystals also changes within that data set, from a high of over 2200 to a low of about 1500. This is all statistically normal.

Since I only open TB crystals these days, I've only opened two batches of arena crystals since then, one for 2250 units and one for 375. The average of the two is only about 1300 units, but I wouldn't conclude anything from just 40 crystals. In another thread I posted a chart of all my TB crystals from 2021 to present. Here's that chart with some better annotation:



As far as I can tell, that seems reasonably random, and if anything until the very last batch (May 10) the moving average was trending upward. However, I don't have enough data to completely rule such a thing out. But conversely, without actually seeing data, given the past history of claims about the crystal odds eventually turning out to be nothing, I would need a lot more evidence to convince me there's anything other than random variations going on. In particular, I take note of this statement of yours:

My mathematical output is apples-to-apples, as if all the crystals were base Arena Crystals.

That's suspicious, because there's no way to do that. Statistically speaking the expected average value of units per battlechip is independent of which crystal you open, but the expected statistical variation is highly dependent. To put it another way, seeing an average of 3000 units from a million BC worth of basic arena crystals is statistically significant, but seeing an average of 3000 units from a million BC of TB arena crystals is statistically meaningless. The fact that you're reporting blended averages across different crystal types means there's no way to ascertain whether your results are statistically meaningful.