Mastery Points

[MVNLA2]

@Dr_Ishmael — I didn’t give more details, but at the beginning the blues were significantly in the lead, Red and Matilda were tied for last, and Chuck and Bomb were in-between. Recently (I would guess at more than 100 visits) the distribution started changing significantly to that described 2 posts above. Will be interesting to see if you distribution shifts.
This is my earlier distribution:

Mastery Points

[dr_ishmael]

@mvnla2 Hmm, and my earlier distribution was directly below yours, too. Guess I was wrong about my historical trends – at that point everyone was pretty close except for Bomb being in the lead. Overall, I feel safe assuming that our sample sizes are still somewhat small, and the actual probabilities for specific birds are all equal.

[MVNLA2]

@Dr_Ishmael — I combined our data, with this result:

All	30.4%
Bomb	15.8%
Blues	14.3%
Chuck	14.1%
Red	12.2%
Matilda	13.3%

Order agrees with yours, except Red and Matilda are reversed.
Interesting to see if all single birds end up at the same % or not.

[Sutekh137]

@KerrAvon

@KillerKea

Now that Epic is working again on my tablet (YAY!), I timed Bomb, Chuck, and Blues (Beserker, Wizard, and Spies) at right around 3 minutes to complete SRC.

Using Red (Paladin) instead of Bomb takes about 3:15. Bomb works better, as you said.

On my iPhone 4, using Bomb, Chuck, and Blues it took about 3:30…

Thanks,
sutekh137

[greatleo]

It appears that if you want to stop receiving e-mails about follow-ups to a topic, the only way to do it is to post something and uncheck the notify box.

So please accept my apology in advance for this off-topic post.

[Immortal… Larry]

@Dr_Ishmael, @mvnla2, with your data combined with mine, the average snoutlings offering is as follows:
My runs+mvnla2+dr_ishmael: (104*5.38+118*5,75+88*6,33)/(104+118+88)=5.79 in 310 visits.
If I use my data combined with the LC re-rolled offerings:
(404*5.88+118*5,75+88*6,33)/(404+118+88)=5.92 in 610 visits.

It clearly does not really matter a lot whether you re-roll or not. The statistical difference with the amount of measurements is not very significant (0.13).

Additionally, I have not seen any serious change in offerings with the addition of rank 10 birds. It seems clear that the dojo offerings are based on 3 pools regardless of the amount of classes available. Single class snoutling offering, all classes for one bird LC and all birds all classes LC. This has been a steady line all the way upto the last available class, after which the snoutling offering will always be present until that class also reaches rank 10.

Hope it adds somthing to the research.

[Immortal… Larry]

Actually, I think it might just be very simple. three offers out of 3 possible pools. Each pool has a 1/3 chance to show up. Two pools you don’t want, one pool you do. If you only accept the snoutling pool, there is again a 1/3 chance that this pool gets offered again. For me the math for calculating probabilities is a bit complicated, but I believe it should come out around our measured numbers for average snoutling offers.

I think it really is that simple…

[MVNLA2]

@larry — Can you give a detailed example of how you think the Dojo works? I couldn’t make sense out of what you said above.
Really need to update the main post. : (
I get irritated every time I see the tip that events are the best way to gain mastery. Maybe if you are just starting out, but it’s pretty pathetic when you are trying to upgrade from rank 9 to 10.

[Immortal… Larry]

@mvnla2, not sure what I have in my head, but it seems to be logical to me. I think the mechanics of this is really straight forward.

I’ll start with the basics. There’s three things that come up in the dojo. A 75LC 1000 for all offer, a 15LC 500 one bird offer and a 100Snout one class offer. Let’s visualise these options as closed chests. A golden chest (75LC offering), a silver chest (15 LC offering) and a bronze chest (snoutling offering). The dojo doesn’t care about what’s inside the chests, it just randomly shows you a combination of chests with a minimum of one bronze chest.

Now we come to the second part. Whatever is inside the chests (which bird or class) is determined by a secondary random pick. At least that is what the statistics show so far.

Let’s take an example. On average you will get 6 snoutling offerings per visit.
Your first classes reach level 10, so if the dojo takes all classes into account, this average must go down, since there are fewer classes to offer.

Our data does not support this at all. Therefore the dojo doesn’t look at how many classes are left to offer. It just looks at the 3 chests it has to offer (75LC, 500LC and 100 Snoutlings). The contents of the chests is irrelevant for the amount of offerings.

Now, onto the next bit. I think that the odds of getting a snoutling offering again after taking one is pretty straight forward, but calculating odds has always been my weak spot. This is where things get fuzzy for me. It does feel pretty straight forward in my head though but I simply cannot explain it. I just tried and typed theories and explanations here over and over again, but I get stuck each time.

I’d like someone else to do the math on this instead.

[Sutekh137]

@larry

This is purely anecdotal, but it did seem to me that I would always get the same “average” number of snoutling (bronze chest) offers on my trips to the Dojo, even when I was down to only a few remaining classes to add mastery to. So, yes, this meant the same class would come up many times, especially as I had fewer and fewer classes to advance.

This culminated in what I think was a bug when at the Dojo I completed mastery on one of two remaining classes and then the last class kept showing up 30-40 times in a row in the same slot (obviously) until I was able to master that one too.

Not sure if that has anything to do with what you are analyzing, but what you state above sounds reasonable to me and my experience.

@MVNLA

Yeah, Events are not that great of a means of getting mastery compared to simply farming CPC. Heck, in the time between Events (the one in December and the recent “Year of the Goat” one) I got all of my classes to level 10. Took about 5-6 weeks of auto-farming CPC. I highly recommend it, and it works well for gaining experience as well (if they ever raise max level again). You’ll end up with level 10 birds and around 40000-50000 in snoutlings.

Since then, I’ve switched the farming to SRC and now have…let me check…about 140,000 snoutlings. That’s because SRC is quite a bit better for cash, and I am obviously not spending any snoutlings at the Dojo…

Thanks,
sutekh137

[MVNLA2]

@Larry — Picky point, but I have gotten 0 snoutling offers on occasion. I will think about the rest of your theory. I thought you had it all figured out. : D

[Sutekh137]

@larry

One more thing… While your analysis makes sense, it seems that there isn’t a 1/3 chance per slot. Not quite, anyway. Because once a gold chest is used, then it is a 1/2 chance as to what fills the next slot, isn’t there? At least, I have never seen 2 gold chests in one listing, or maybe I am forgetting? One CAN get multiple silver and bronze options. Hm, now that I think about it, it WASN’T a bug that my one remaining headgear kept appearing. Because there was nothing else the Dojo could have displayed to me. It already used the 75LC option, the 15LC (for the last remaining bird) and then had one more slot to fill. And the only thing it could put there was the last remaining specific headgear, over and over. Other folks who have maxed out everything should be able to confirm that behavior. Once you are down to one bird and one class, the slots should ALWAYS be:

75LC for all birds, all classes.
15LC for last remaining bird, all classes.
N snoutlings for the last remaining class,

But yes, I would say that once the chances for gold/silver/bronze are determined, then it does another random choice as to what advanceable bird within that family to choose (for silver) or what advanceable headgear (for bronze). Obviously the gold chest is the gold chest, as there is only one of those.

Sorry if I am just repeating what others have said — I didn’t read the whole thread, as I don’t have that much free time. *smile*

Thanks,
sutekh137

[Sutekh137]

@MVNLA

Then it would seem like the first choice could be a 1/3 chance for gold, silver or bronze.

If gold was picked, the next choice is fifty/fifty between a silver or bronze, and then fifty/fifty again for the third slot? But I would think that would yield more zero-bronze scenarios. Perhaps it works that way except the odds of the gold chest are much smaller than one-third.

Thanks,
sutekh137

[Immortal… Larry]

@mvnla2, I think I just got it…
When you visit the dojo at first:
There’s 100% chance that there’s 1 bronze chest in one slot.
There’s 50% chance for each other slot to contain a bronze chest.
There’s 50% chance for each other slot to be a silver or gold chest.

This adds up to 66% of all initial offerings in the dojo to be snoutling offerings. If we take this figure into account for all following offerings in the dojo, you end up with the following.

The first offering in the dojo should average at 2 snoutling offerings.
After taking both snoutling offerings, you get 1,33 new snoutling offerings. This adds up to 3,33 (66% of the 2 new offerings).
Taking the 1,33 snoutling offerings, you get 0,89 offerings. This adds up to 4,22.
Taking the 0,89 offering, you get 0,59 offerings. This adds up to 4,81.
Taking the 0,59 offering, you get 0,39 offerings. This adds up to 5,21
Taking the 0,39 offering, you get 0,26 offerings. This adds up to 5,47
Taking the 0,26 offering, you get 0,18 offerings. This adds up to 5,65
Taking the 0,18 offering, you get 0,12 offerings. This adds up to 5,77

And so on. If you keep adding this up to infinite buys, you end up at an average of 6 total offerings, which is fully in line with our statistics.

The only flaw in this reasoning is the forced first offering which we start with…

[Immortal… Larry]

Wow, you guys are fast. I didn’t realise someone already replied to this.

You are all right, I am not taking into account the fact that the odds increase when there already is a 75LC offering. However, I am starting to think that all LC offerings are in the same chest.

Basically there could only be 2 chests. One for LC offerings and one for snoutlings. That would make my theory more elegant.

[KerrAvon]

@sutekh137 What you suggested for getting the final bird to mastery 10 is correct. You get 2 slots filled with LC offerings and the last continually gets replaced by a snoutling offering. This has been posted in this thread earlier.

You can also use this to get any new bird you buy up to mastery 10 quickly. i.e. if you get Bard, you should be able to buy snoutling offerings to get her to rank 10 straight away.

[KerrAvon]

@larry I think your initial suggestion is probably close to what actually occurs. Except, you can get the situation where there are only LC offers. I am surprised you did not come across this.

The selections may be something like this (you would have to do a lot of work to determine actual chances)
20% for 75LC offer (Gold)
30% for 5LC offer (Silver)
50% for snoutling offer (Bronze)

The system then runs through and fills each of the three dojo positions.
If the 75LC offer is already present, then a snoutling offer is given. When a silver or Bronze chest is selected, then there is a secondary selection to determine which bird or class offer.

[Immortal… Larry]

@kerravon, we’d need a proper mathematician for that…

The only thing to go on is the fact that all the stats combined end up just below 6 (5,92). Since the amount of data we gathered is quite significant with 610 depleted dojo visits, the statistics should end up really close to this number.

If you take into account 20% for 75LC and 30% for 15LC, and the assumption that we will never buy one of these options, could you consider both to be in the same chest and ending up at 50% anyway? Basically what I’m going at is that regardless of gold or bronze chests, they would not be touched in this method, so their percentages could be taken as one unit in the math.

[Immortal… Larry]

@kerravon, what I meant to add is the fact that the stats were continuously below 6. Which, considering the large amount of data points, could suggest that it is more complicated than what I described. This speaks for your statement that there’s a bit more to this.

[KerrAvon]

@larry I think the only way to work it out would be to write down what new selection comes up in each slot at the dojo after you collect an offering.

You would need to answer the questions
If there are 3 snoutling offerings and you take one, what does it get replaced with?
If there are 2 snoutling offerings and you take one, what does it get replaced with?
If there is 1 snoutling offering and you take it, what does it get replaced with?

This would get complicated, and you would need to know what is in each slot and collect a lot more detailed collection data.

Probably the simplest way to do this is ask. Is there a difference in what comes up in the third slot when slots one and two contain 2 x 15 LC offerings or 1 x 15 LC and 1 x 75 LC offerings?

[Immortal… Larry]

@kerravon, Would that not be a bit too much data to collect? Having an additional variable in there would increase the necessary data a lot…

I think the only important thing is that the chests principle applies in some way. Which means that having rank 10 birds should not matter at all for the dojo offerings. This in turn means that the measured average should be okay for all situations regardless current ranks (except the last class or bird to reach rank 10 of course).

[Immortal… Larry]

@mvnla2, I have really never seen a dojo without an initial snoutling offering. Have you included that in your stats? It would significantly lower the averages…

It might also be a bug. Either on your or my side…

@sutekh137, what do you think of the gold/silver in one chest option? It could equalise the statistics somewhat… @kerravon is making some good points in here as well.

I’ll sleep on it for now.

[KerrAvon]

@larry You are absolutely correct. If the system is based on the chests principle, then it shouldn’t matter if it is 2 or 3 chests.

Collecting the data may help determine whether it is 2 or 3 chests. But it won’t get you closer to answering the original questions – if the offerings changed as you got higher ranks and when you got rank 10 classes. And currently it looks like they won’t change, which does support the chests principle.

[MVNLA2]

@Larry — I must be dreaming. I could swear that I’ve gotten 0 snoutling offerings at the Dojo, but there are none in my spreadsheet.
It’s going to take me a while to digest all the recent posts.

[MVNLA2]

@Larry @KerrAvon @sutekh137 — What happens if the next item chosen is already shown on the board? We “know” that there are never two identical offerings on the board.