mythic/rare rate decreased since booster crafting update?

Llorence

Ive been getting 3000 orbs every 2 weeks or so, and i get a mythic when i do that, but in-between times ive been opening alot of packs from spending gold, events, etc, but no mythics. Before the booster crafting update i was getting a mythic every 2 weeks from pack, now im getting a mythic every 2 weeks because of booster crafting.

Gilesclone

Since booster crafting arrived, I’ve drawn more new mythics than I did in months before it happened.  It’s very random.

James13

Data aggregators (collecting actual drop data from many users) have shown no real change as far as I'm aware.

As noted, personal variance can make individual experience quite different from the average.  But it all washes out in the end.

Llorence

hmm i see, well i guess ill keep grinding and hoping for a mythic from a pack i get

James13

I should note, that there was a coalition theory for a while that the Random Number Generator could get on a bad "seed" or good "seed".  As in if you got a good pack it was better to open a stored set of them then and there and if you were opening poorly then stop for a while.  I'm not convinced that it's actually true at all, but I did abide by that superstition for a while (and still do occasionally).

Thuran

In most games, loot boxes are programmed to give better loot if you have been gone for a while, in order to entice you to start playing again.

There are hints of such hidden mechanics also being at play here, as well as other hidden mechanics that ensure your booster packs are NOT random, but instead look at a number of factors before determining the cards you will get.

Which is why we need the droprates disclosed. Once they settle on a fixed droprates, it is much easier to keep them honest, or at least spot intentional skewing of the droprates.

I'm sure most people are aware that your chance of a dupe is fixed as higher than a nondupe, but seemed to improve shortly after BC was introduced. 

What we call "superstition", is often just guesses at how the system works, based on certain patterns, and the devs have done nothing to alleviate these theories.

I haven't seen the " keep opening" theory much. Rather I have seen a theory that once you get a mythic or a rare, you should stop opening packs immediately, as you won't get another right away.

To;Dr: if you don't want us to guess at what you are trying to hide from us, if you are honest, release your tax retur- I mean droprates!

James13

Thuran said:

I haven't seen the " keep opening" theory much. Rather I have seen a theory that once you get a mythic or a rare, you should stop opening packs immediately, as you won't get another right away.

My experience was always colloquially the opposite.  If I opened a mythic in a booster I would almost always find one in my stored monthly pack if opened next right away.  I got three of my masterpeices in the middle of a "hot streak" when HOU released.

Divergent experiences are why it's probably truly random.  Heh.  Gambler's fallacy.  Reading patterns into where there are none.  Etc, etc.

ertaii

I got more mythic recently than ever, so I must disagree

IM_CARLOS

Depends... I got so good days when 3 rares are in one reward booster and a mystic in the next day a mystic as free card booster. But such days don't come very often. Last rare I found (beside rewards) is a month ago. 

Volrak

Thuran said:

Which is why we need the droprates disclosed. Once they settle on a fixed droprates, it is much easier to keep them honest, or at least spot intentional skewing of the droprates.

Agreed!

Thuran said:

I'm sure most people are aware that your chance of a dupe is fixed as higher than a nondupe, but seemed to improve shortly after BC was introduced.

Unless you've got hidden data backing it up then it's just a guess, and suggesting it's any more than that is actively misleading.  (Sorry, I'm a little particular about these things.)

Also, I have data to the contrary.

Exhibit A is the origins mythics dupe data collected just before booster crafting arrived (see this thread and the ORI mythics tab of the drop rates spreadsheet).  I've compared the actual number of drops n it took to get the actual number of non-dupes, with the statistically expected number it should take to get that many non-dupes, based on simple random drops.  If dupes are noticeably more likely to drop, then the total actual number of drops should be significantly more than the total expected number.  But the actual total (282) was not significantly more than the expected total (307).

Exhibit B is Marvaddin's registered drops from free boosters.  He not only recorded each card he opened but how many he already owned at the time, so it's possible to calculate his expected gains.  He's opened 1438 free cards, and based on simple random drops and his collection size when each card was opened, he should expect 150.89 new cards from those.  If dupes are noticeably more likely to drop, then he should get significantly fewer new cards than that from his drops.  He got 149 new cards, which is not significantly fewer.  (Spreadsheet with working is available on request.)

Thuran said:

What we call "superstition", is often just guesses at how the system works, based on certain patterns, and the devs have done nothing to alleviate these theories.

Sure.  But why guess or be superstitious, when you can get actual answers from data collection and analysis?

Thuran

We did however see a lot of data indicating just how large the percentage of dupes, compared yo nondupes were prior to booster crafting. I don't remember the exact discussion, since it was last summer, but several charts made it clear that drops were not evenly distributed, but instead heavily skewed towards dupes.

Furthermore, other data from back then showed how, as soon as one copy dropped, the frequency of that card appearing would soar drastically, often resulting in multiple dupes almost immediately, on a card that otherwise had not dropped for months, behaviour which occurred consistently.

We also know that the internal mechanics of boosters can take dupes into account, as confirmed when Oktagon stated a few weeks back that Elite packs would be tuned to provide dupes less frequently.

Ofc, we don't know if that means "artificially higher chance than random at nondupe", or " lowering the already artificially skewed chance of receiving a dupe." 

What we do know, is that the mechanics can take dupes into account, that it is common practise in the loot box industry to skew things towards dupes in order to keep customers buying loot boxes, and that people who received a nondupe suddenly saw a drastic spike in how often they got that card, compared to before they owned the first copy.

As stated, it seems that the booster crafting patch changed this towards numbers closer to what you would expect, except we see numerous anecdotal reports of mythics becoming almost extinct lately, compared to before booster crafting. 

You know a MTG mythic happens roughly every 15 boosters, which can be easily verified to be consistent. What I wonder is how so many people can fail to get some form of consistent droprates data.

But easy to figure out in theory, just gather 10 players with similar collections, have them open 100 boosters each, and check if the droprates are similar. In 100 boosters they should each get 1-2 mythics ( )with similar intervals,  with a few getting maybe 3.  

I'd gladly volunteer to stock up to 100, once the first month of ixalan is over

Volrak

Thuran said:

We did however see a lot of data indicating just how large the percentage of dupes, compared yo nondupes were prior to booster crafting. I don't remember the exact discussion, since it was last summer, but several charts made it clear that drops were not evenly distributed, but instead heavily skewed towards dupes.

Furthermore, other data from back then showed how, as soon as one copy dropped, the frequency of that card appearing would soar drastically, often resulting in multiple dupes almost immediately, on a card that otherwise had not dropped for months, behaviour which occurred consistently.

I'd certainly be interested to see that data.

Thuran said:

What we do know, is that the mechanics can take dupes into account, that it is common practise in the loot box industry to skew things towards dupes in order to keep customers buying loot boxes, and that people who received a nondupe suddenly saw a drastic spike in how often they got that card, compared to before they owned the first copy.

I'd be completely unsurprised to see complaints of skewed dupe rates appearing across every game that relies heavily on something like loot boxes.  But I don't know of a game where that practice has actually been confirmed (not that I've actively gone looking).

Thuran said:

You know a MTG mythic happens roughly every 15 boosters, which can be easily verified to be consistent. What I wonder is how so many people can fail to get some form of consistent droprates data.

Actually, it's one in 20 5-card boosters.  From about 50,000 recorded card drops the observed average mythic drop rate is 0.986%, with a 99% confidence interval of +/- 0.12% (and we can speculate the true mythic drop rate may be 1%).  In other words, from these observations, we can say with 99% confidence that you need more than 15 boosters on average for each mythic drop (since 1 in 15 boosters would give a drop rate of 1/75, which is 1.33%, which is outside the 99% confidence interval of 0.87% - 1.11%).

We find this drop rate is consistent across the "normal" pack sizes (i.e. single, super, premium excluding bonus rare, as well as the old fat pack and big box also excluding bonus drop).  It's also generally consistent across pack sets (ORI, BFZ, SOI, KLD, AKH, and HOU).

Thuran said:

But easy to figure out in theory, just gather 10 players with similar collections, have them open 100 boosters each, and check if the droprates are similar. In 100 boosters they should each get 1-2 mythics ( )with similar intervals,  with a few getting maybe 3. 

From 100 boosters (500 cards), you'd get on average 5 mythics, but there's a spread like you say, with for example 1 in 60 people getting as many as 10 mythics, and 1 in 142 people getting none at all.

If you're interested in this kind of stuff you should really check out the community drop rates spreadsheet.

While I mention it, we always want more data so we can pinpoint the true drop rates more precisely and detect if anything changes.  Most relevant, with XLN arriving imminently:
* The "Data - HOU bulk" tab, in which drops from stockpiled HOU packs can be entered, and
* The "Data - XLN bulk" tab, in which new XLN packs can be entered.
There's also a general "Data" tab where any and all ad-hoc drops can be recorded.

I ask anyone who will be opening a large number of packs to not let your data go to waste, and please consider entering it in the spreadsheet.

Gilesclone

I’ll be opening 49 HoD packs and probably 20 or so Ixalan packs.  I’ll record my results and add them when I get a chance.