Resume of how Titan loot works – Tiers, Scores, theories

[Arien]

Well I am always A or A+ on 7-9* titans, so loot tier 9-12. I see a lot of 3* ascension items drop, one problem is shown in this picture. Not all 3* items are good loot. This was “lucky” and “expected” outcome. Best loot for A+. Two 3* items. But the items themselves are crap. Both can be farmed on the map and I have tens of them each. They still are on the same roll.

The problem for me is 4* items. They are the limiting factor. Chance for them to drop is much smaller and thus the variation much bigger.

[Mai]

Your reasoning is correct, but the proposal is not. Let me give you an example so you may find the flaw yourself:

You have two people with 2 different dices. One roll 100 numbers, the other 90. But you do not know which numbers were taken off from the 90 faces dice. They can only roll once a day. If you need to prove the dices are different using only the data, how many tries do you need?

1-2 month is about 60 tries. Even if they have the luck to get only different numbers, each getting 60 different number proves anything? Probability states that the 100 one could actually roll more repeat numbers just by chance. So in a limited pool you can think that clearly he is the one will lesser options.

So how many tries do I need to know for sure all the 10 numbers missing in the 90 face dice? There’s no right answer, this is probability, you can get it in 90 tries or 2 billion tries. You can literally spend your whole life throwing that dice and never get 90 different results; unlikely yes, impossible no.

Now imagine you have no idea how your dice looks like, how many faces it has, how many options and possibilities, you only see the results. How many tries do you need to understand what’s happening? A ■■■■ load.

Two months, 60 tries is nothing, absolutely nothing. In 60 tries C can get better loot. The more alike the two “dices” are the more try you need for their difference show in the data. A 1-100 and 1-99 dice has VERY similar data, that 1 number missing will barely show. You need a lot of tries to stress that little diference and make it visible.

I said it already, you guys don’t seem to get the magnitude of the situations. You need thousands of tries in the same EXACT tier loot for years. That’s not a exaggeration, and my example were a dice with 100 sides, so that things has minimal 1% chance, but the probability that some might be 0,x% chance is very high, the dice could have a thousand faces or more. We don’t know, we would have to keep rolling until the data would be stable.

[Mai]

Arien, you said before that 4* only drop after tier 7, what’s the source of that info? If it’s something the devs said, I would like to add in the main post.

[Arien]

That is the lowest I have heard. I don’t usual do that low tiers so haven’t got that personally.

[Elpis]

Again your thoughts are very solid, and my logic missing one point if your first post is true: the difference between loot 4 and loot 12 is 1 (!) ascension material, and that is too small of a difference to get an answer with my method.

If the difference between 3 steps (12-9 or 9-6 for example, that is the difference between A+ and C) have doubled the ascension material, then even if we don’t know the face of this hypothetically dice that we throw, the fact that 1 person throw only 1 dice and the other throw 2 to get the same number make the difference bigger enough even if we have to see the results in only 1 or 2 months, casuality or not casuality.

So if your tier description is kinda reliable, that is the source of the problem.
The difference is too small.

[Rook]

You misunderstood me; I’m perfectly happy to post pictures to add to the pile for science’ sake. But since I’m told that all such evidence doesn’t matter, since it’s too small a sample, really what’s the point?

These are recent. Posting goes backwards from present to July 1 (I didn’t take pics of every Titan)

You’ll notice I have included pics of:

• A, B, C
• Winning and Losing
• Minor and major ascension items

These are mine; I didn’t save the A+ stuff, but I can certainly find and post it for you in the next post.

Remember, my contention is not that B getting loot makes no sense…but that A/A+ should be getting it 1/3 the time, according to your math. No conspiracy here; please give that a rest.

[Rook]

From my alliance:

These are from present backwards, no cherry-picking. Cheers.

[ZapperCat]

Mai, you may rest now… “those with ears will hear…” and good work!

[Rook]

Today A+ won (yay!), though B got a double goodie and A got nothing: these are the pics posted/reported inside our alliance for today. More data for you.

[Mai]

@Rook

But since I’m told that all such evidence doesn’t matter, since it’s too small a sample, really what’s the point?

Again, I am NOT looking at odds in this post, but patterns of loot distribution. As you can see there’s a variation, but the pattern is pretty solid. It’s not about odds regarding quality, but quantity!!!

[ALL]

[ALL]

Example of a bad loot)))

[Rook]

@ALL

Example of a bad loot)))

Technically, you are getting ascension material (chainmail), but yeah, not what I consider [rare] ascension loot: shields, orbs, cloaks, etc.

[Rook]

@Mai

It’s not about odds regarding quality, but quantity!!!

We look at both things in our alliance:

1. Quality: What loot is being won? (Orbs, rings, cloaks, trap tools, etc. Yay!)

2. Quanity: Who is getting that rare ascension loot (A+, A, B, C), and how often?

My initial posts were because we noticed a pattern of A+ and A not getting good stuff while B did.

We accept the statement that it’s due to probability (how many B vs A+/A in the Alliance). Even so, we don’t see A+/A receive goodies as often as they “should”. Hopefully it’s just a blip in the pattern and will change (like it did this morning).

[Mai]

No, no, you are not understanding. You are messing 2 different things.

Quality of the items (the stars of those items) and odds (how often they are rewarded) that are both about ODDS of different options in a single roll inside the loot itself.

I’m looking at distribution of rolls and loot. How many items a person in rewarded every tier, regardless of quality and odds. I have NO interest in paying attention at what specific item was rolled, but the classes of it.

This is NOT a post about ascension items, there’s plenty of others about that. This is a general post about loot tier and loot distribution. Not odds.

Every time someone gives me a print screen I check how many Ascension Materials, Ingredients and Battle Items were rewarded, also how many special rolls of heros, tokens and flask. I’m documenting this.

I did pay attention to Iron and Food at first, to see if the amount would increase as the tiers go up, they do. But that’s it, I won’t include in the data the exactly items.

And the only useful printscreens are the ones that show a situation not yet documented in my main post:

@Mai

Tier I –
Tier II –
Tier III – ? Rolls (1 Food, 1 Iron, 2 Ingredients, ? ascension materials, ? battle items, ? gem)
Tier IV – 9 Rolls (1 Food, 1 Iron, 2 Ingredients, 2 ascension materials, 1 battle items, 1 gem, 1 hero/token/flask)
Tier V – 8 Rolls (1 Food, 1 Iron, 2 Ingredients, 2 ascension materials, 1 battle items, 1 gem)
Tier VI – 8/9 Rolls (1 Food, 1 Iron, 2 Ingredients, 2 ascension materials, 1 battle items, 1 gem, 0/1 hero/token/flask)
Tier VII – 8/9 Rolls (1 Food, 1 Iron, 2 Ingredients, 2/3 Ascension materials, 1/2 battle items, 1 gem, 0/1 token/hero/flask)
Tier VIII – 11 Rolls (1 Food, 1 Iron, 3 Ingredients, 3 Ascension materials, 1/2 battle items, 1 gem)
Tier IX – 12/13 Rolls (1 Food, 1 Iron, 3 ingredients, 3 ascension materials, 3/1 battle items, 1 gem, 0/1 token/hero/flask)
Tier X – 12/13/14 Rolls (1 Food, 1 Iron, 3 ingredients, 3 ascension materials, 3 battle items, 1 gem, 2/1/0 hero/token/flask)
Tier XI – 13 Rolls (1 Food, 1 Iron, 3 ingredients, 3 ascension materials, 3/2 battle items, 1 gem, 0/1 token/hero/flask)
Tier XII – 12 Rolls (1 Food, 1 Iron, 3 ingredients, 3 ascension materials, 2 battle items, 1 gem, 0/1 token/hero/flask)

I’m currently traveling and didn’t yet check all the ones you and others shared after. I’ll soon and update the list above. That’s all I’m doing. That’s my only objective in this space. But I have to be honest that’s a lot a data, maybe others might have a use for it.

[Fledoble]

I love it! Keep sharing data!
Someone can always use it!

[Revelate]

@Arien

Good post, one note is that rare ascension items are 3*. epic are 4*

4* ascension items start to drop at tier 7.

In my alt’s alliance on a 5* titan.

Damage #8 (should be rank B), therefore loot tier VI, received Darts (4*).

Don’t think there was a screenshot, just saw it in chat unfortunately where he was commenting it was weird to get the high level one rather than the low level one he actually needs right now for Joon.

Course my one and probably only ever A+ on anything (combination of fortuitous level and way better than average boards), got butkus heh.

Edit: he confirmed in chat loot tier 6 but no screenshot.

[ThunderStrikes]

Look what I got today

[jorma]

[Revelate]

@Mai

Rev, like said before a few dozen times, that would require a ■■■■ load of data. And it would have to be of the same exactly tier. It would take years to have anything that might resemble the real odds due the limitations and access to that data. It can’t be done.

Other games use addons that collect that data as you play of millions of players to get a decent idea of the probabilities of loots. That’s not possible here, just isn’t. You guys don’t seem to get the “thousands” part of it. Any conclusion taken from a small pool can NOT be trusted. That’s simple statistic.

And even if you don’t know the exactly odds, doubting what the devs said it’s the true is completely absurd. They made clear a dozen times, using their FAQ, answering mails and even posts that the loot for higher tier are better. That’s a fact then. If you can’t perceive that, either you have a ridiculous bad luck or you don’t understand what probability means. Not getting a single piece of loot for weeks and even months are not proof that it’s wrong, it’s actually a normal outcome from a system of probability that doesn’t have any kind of guarantee system.

So instead of letting your mind starting to get paranoiac, see things that doesn’t exist, start to doubt what is given us as a fact of the game and waste time trying to disprove it, you better focus your effort in discussing ways the devs could improve the system and make it less frustrating. And if you really want to know odds, start begging, that’s the only way you’ll ever know.

I didn’t state my opinion on the matter, but if you’d bothered to read my other posts I’m a staunch supporter of the developers on this one: RNG is RNG. Something like 20% of players seem to get a titan specific ascension item off a 8* or 9* titan just from anecdotal reports… I think there are too many potential drops and they need to be reduced to common ascension items rather than individual by color, but I don’t get a vote on that.

Further, I have taken some upper division math and economics classes and that includes both statistics and probability. You don’t need anything like a million datapoints to get a pretty good general idea: don’t throw out the good chasing the perfect, that’s a terrible way to live one’s life and the same applies here. Case in point: how many billions of USD in advertising annually were based on the Nielsen ratings which until something like 2014 used on the order of 4000 households across the US to base their metrics? Do you really think that all the organizations who paid for advertising based on those metrics were mathematically incompetent?

Basic math: roughly 10 different alliances in this thread, and on the order of 5 to 10 people in alliances share their results. There may be some skew from non-regular contributors (and those could be tossed out) but over a month that’s over 1500 datapoints if people collectively rolled up their sleeves and contributed. It simply can be done to a degree to prove or disprove the more outlandish claims in this thread.

That’s better than arguing about the subject when nobody has any real data to back their claims.

That said, I don’t think it happen but I am open to being pleasantly surprised.