Mastery Points

[Sutekh137]

@toolow

That is just…disgusting. Back in the day of SRC I could have gotten 600,000 in a year just petting things grind while I work — and that was with ads! If you bought a bit of LC to get rid of ads, even better!

Not to mention folks who used emulators to grind overnight, which still isn’t really cheating. Heck, my time zone manipulation of the old Golden Pig offers was way worse than that. If they look for snoutling totals, why wouldn’t they look for LC totals? A few thousand LCs without ever having purchased any? These guys really are dolts sometimes, even when trying to enforce against baddies!

I’m amazed you still play (if you do…). I probably wouldn’t have and would have done everything I could to make other people uninstall every Rovio game possible.

@burbman

Right on average then, so far. *smile* We need a lot more data points!

Thanks,
sutekh137

[Anonymous]

@sutekh137

i maxed all seven classes within a day of each other. I didn’t bother recording data in between. And it will be a long time before I max any more classes, so I should have an adequate amount of data when it’s all said and done.   :)

[DD]

@sutekh137
I’ve stopped playing with my second account. Until the data on my main account sync, I’m just carrying on to keep touch with the fine company here. I simply don’t have the time and means to publicize what Rovio did and wage a campaign against them. Besides, why spoil it for those who are still enjoying the game? Having said that, I do bring their action to people’s attention from time to time, as I just did here. People can make up their mind as to what actions they take, if any.

[bchild]

I should be able to start collecting data for 1 maxed out bird family dojo offers by the weekend. Chuck has 2 classes maxed with 2 at 79 (each with about 5 dojo buys left) and two others at 78 and 75.  My data may be slightly skewed since I will have one Bomb class maxed.

[burbman]

Time for the six week mastery grind check in.

Red – All classes at 78

Chuck – Four at 78, two at 77

Matilda – One at 78, five at 77

Bomb – all maxed at 80

Blues – Four at 75, two at 74

Currently in first place in the Birthday event by over 15,000 medals.  A prize wheel roll of 6,000 for all, Red, and Matilda will pretty much max out all of my red classes this coming weekend, and leave me with less than anther week of grinding to finish off the flock.  Certainly did not expect to finish up in 7-8 weeks.

[burbman]

@sutekh137 @poptimus – My statistical analysis of dojo offers with only four families has 50 data points so far.  Average right now is 5.48 offers, slightly lower than the figure for all classes available, but still a small sample size, and the average moves quite a bit with each new data point.

[Sutekh137]

@burbman
@poptimus

I was at 5.64 at 50 data points, and dipped as low as 5.62 at 74 data points. So yeah, long way to go…

My average currently is 5.84 on 219 data points, still “pure” Dojo runs.

This event is terrible. I’ll be lucky to get one star. The most stamina I have missed so far is about 40 and yet there is virtually no chance of me even getting a two-star rank. Not sure where all of these people are coming from who never miss a battle and have so much extra stamina and FE to burn, but it sure is annoying. I remember when I got Top Three in every event, big or small, for over a year straight. Not sure what gives. Right now I am at 38,300 and the lowest two-star score (tenth place) is at 45,050.

More time for “pure” Dojo data sampling, I guess…

Thanks,
sutekh137

[Anonymous]

@burbman @sutekh137

61 data points for me, since reducing to 23 available classes. I’m at 5.10 Snout offers per visit.

[burbman]

@sutekh137 – your 38,300 would be good for sixth on my board right now.  Do you use FE to respin for top award each play?  Can make a big difference.  I am at 60K, and have only used about 20 SD so far in ones and twos before bed or work to use all free Stamina available.

[Sutekh137]

@burbman

I don’t have any FE, so I can’t re-spin anything. That is crazy that I would be sixth on your board but am…sixteenth on mine (and am at zero stamina right now, so this isn’t a temporary lull). That has been my experience with all events since they went to the “more contestants on one board” format. I have to get lucky and do everything just right to even get two stars, much less three.

@poptimus
5.10, eh? So what is your overall ratio of potential snoutling offers to LC offers? That is, how many classes do you have maxxed out vs whole birds? Because if you are down to 5.10, @burbman should also have lower returns because he has so many classes maxxed (even if all Bomb is maxxed out).

I realize I should probably add columns to my spreadsheet for potential LC buys vs potential snoutling buys to compare that ratio against average snoutling offers per Dojo trip. Shouldn’t be too hard to add a few more columns…

Thanks,
sutekh137

[Anonymous]

@sutekh137

I don’t have any birds fully maxed out. I’ve only gotten the following classes to M80:

– Red: Paladin, Stone Guard

– Chuck: Lightning Bird, Thunderbird

– Matilda: Witch

– Bomb: None

– Blues: Rogues, Skulkers

In the (now 62) dojo visits since that’s been the case, 83.3% of my initial offers (155 out of 186 slots) have been for Snoutlings, and 51.1% of the replacement offers (162 out of 314 slots) have been for Snoutlings. Overall it’s a 62.7% ratio.

[Sutekh137]

@poptimus

So seven maxxed classes but no maxxed birds means your potential snoutling:LC ratio is 23:6, or around 3.83. That’s quite a bit lower than the “pure” average (nothing maxxed) of 5. Seems to be having an effect if you are still at 5.10 — that’s pretty low, even with only 61 data points.

Let’s compare to @burbman… According to his listing, he has maxxed out 6 classes (all of Bomb’s) and one bird (Bomb), so his snoutling:LC offer ratio is 24:5, or 4.8. Hm, maybe that will turn out to be about right, depending on how many data points he can get. He’s off the max ratio of 5, but not as low as your ratio of 3.83. As classes and whole birds are maxxed out (if that tack is chosen), the ratios go:

Nothing maxxed: 30:6 = 5
1 bird maxxed: 24:5 = 4.8
2 birds maxxed: 18:4 = 4.5
3 birds maxxed: 12:3 = 4
4 birds maxxed: 6:2 = 3 (moot, since you will be able to max out last bird)
5 birds maxxed: (doesn’t matter — you are done)

Looks like it might turn out that not maxxing any classes at all is the right way to do this thing. Though, it is still totally worth it to leave one whole bird behind and then slowly increase all classes of the other four birds. The data should tell the tale for sure…

My spreadsheet now has columns for potential offers and the snoutling:LC offer ratio:

https://docs.google.com/spreadsheets/d/1bbpM4fUkhjb_w6bZMnQQi6zT-MKBcCMkJMQmCKXl-P8/edit?usp=sharing

Thanks,
sutekh137

[burbman]

@sutekh137, @poptimus – Now for the next question.  Are Dojo offers determined as each space becomes available, much like a dice roll, or are they determined more like a deck of shuffled cards, with the previously selected card replaced within the deck randomly?  The differences that @poptimus is seeing between original & subsequent offers has me leaning toward the latter, as a dice roll scenario would not vary significantly between types of offers, where odds of an LC offer would tend to increase progressively as cards are turned.

[Sutekh137]

@burbman
@poptimus

I think it is more like the deck of cards, or a scenario where the number of snoutling offers is determined from the outset. There are times when the last slot still offers up quite a few more snoutling offers, but usually once the LC starts taking over it spreads fast. I am speaking anecdotally, though — my method of taking data in no way helps with your questions statistically…

I added a histogram to my data, and it is sort of weird — the number of 6-offer Dojo trips I have had is far less than 5 and even less than 7. Very odd. Especially since our average is relatively close to 6, I was surprised to see the histogram results looking that way.

Thanks,
sutekh137

[Anonymous]

@sutekh137

Remind me what your maxed classes situation is? I thought you were doing something similar to me…?

@burbman

Interesting theory. The way I’ve been tracking the data, it lends some possible credence. If you check out the spreadsheet I had e-mailed you recently again (the one that shows my data from before I had maxed out any classes), you’ll see that the percentages do indeed decrease slightly in the first, second, third, and fourth rounds of replacement offers (60.6%, 54.6%, 52.6%, and 50.0% respectively, based on sample sizes of 287, 174, 95, and 50). Beyond that, the samples are too small to be significant. (Really, even the fourth replacement offer sample may be too small, right? I don’t know as much as stats as I would like to, at the moment!)

My data so far from after having maxed out seven classes is currently returning replacement slot Snout offer rates of 54.2%, 46.4%, 48.7%, and 42.1% based on sample sizes of 155, 84, 39 and 19. It’s certainly possible that more data will solidify these numbers further, obviously.

Oh, and I should also add that my Snoutling offers have tracked very closely with my expectations in terms of which birds they’re being offered to. Since I haven’t maxed out any Bomb classes yet, my theory is that he should receive roughly 26.1% of Snout offers (6 of my remaining 23 classes), and he’s received 25.2% in reality. Here are the corresponding results for all five birds:

Red: 17.4% expected, 12.6% actual

Chuck: 17.4% expected, 22.1% actual

Matilda: 21.7% expected, 22.4% actual

Bomb: 26.1% expected, 25.2 actual

Blues: 17.4% expected, 17.7% actual

[Sutekh137]

@poptimus

I am still “pure” — nothing maxxed. I am doing that as long as I can to get the most data points possible of the pure state.

Once I have one class go max (Seadog is in the mid 79 level), then I will probably have 8 max as quick as I can, 2 for each bird that is not my laggard (Matilda).

I realize I should probably go ahead and max more quickly so I get more data points in that more interesting scenario, and because all of our “pure” data has ended up in the 5.8 range. But I’m sticking with my original plan just ’cause.

Thanks,
sutekh137

PS. Great data analysis! Any thoughts on why my histogram has such a low occurrence of 6s?

[Anonymous]

@sutekh137

Ah, I didn’t know what you meant when you said “pure” earlier, but I should’ve been able to figure it out from your sample size. Either way, thanks for clarifying!

“PS. Great data analysis!”

Thanks!

“Any thoughts on why my histogram has such a low occurrence of 6s?”

I have no earthly idea!   =-]

[burbman]

@poptimus @sutekh137 – I can shed a little light on the occurrence of different values.

I have all three of our “pure” data sets on a single spreadsheet with 539 data points.  Four is the most common result, which makes sense with an average of just below six, as we have an absolute minimum value of zero, and theoretically an unlimited top value.  (our maximum recorded is 16).  Six itself is slightly below what I would expect it to be, but is probably just noise in the data.  Here are our results:

0 Snout – 2 times

1 Snout – 12 times

2 Snout – 22 times

3 Snout – 56 times

4 Snout – 96 times

5 Snout – 87 times

6 snout – 67 times

7 Snout – 76 times

8 Snout – 39 times

9 Snout – 31 times

10 Snout – 14 times

11 Snout – 15 times

12 Snout – 7 times

13 Snout – 7 times

14 Snout – 4 times

15 Snout – 2 times

16 Snout – 2 times

[Sutekh137]

@poptimus

@burbman

Do you have your “pure” data around any more to get the distribution of 4s, 5s, 6s, and 7s? I wonder if you have a “six dip” as well?

I wonder if 6 isn’t supposed to be the average number of snoutling offers in “pure” mode, and if they actively work against 6 being the peak to make it harder to decipher. That doesn’t really make sense, but 6 should clearly be the top dog in the distribution if the average is 5.8+ and the distribution is truly random/continuous. It’s just really weird.

Thanks,
sutekh137

[Sutekh137]

@burbman
@poptimus

Ah, yeah, forgot about zero. But then why wouldn’t 5 be the peak? Or are you saying 4 is the peak because the “low” side of the peak is weighted more heavily since the values can range so high? Makes sense. The distribution is not a classic Gaussian sort of thing? (I suck at statistics).

But then why do we have more 7s than 6s? In my data I have 21 6s and 32 7s (a more dramatic difference than in the combined data you present). The distribution of everything else looks pretty good, but that six-dip is weird.

Thanks,
sutekh137

[burbman]

@sutekh137 – Six dip is odd in our pure data, but in my Four Family list, I have 11 – 6’s and only 2 – 7’s.  I think it is statistical noise. (you will probably get a big run of 6’s now to make all this mute)

[Anonymous]

My personal distribution, based on “pure” visits:

0: 0

1: 4

2: 4

3: 17

4: 19

5: 17

6: 18

7: 10

8: 3

9: 10

10: 2

11: 4

12: 2

13: 1

14: 3

15: 0

16: 1

@sutekh137

I think the main point @burbman is getting at is that this isn’t your typical “bell curve” data, since there’s an absolute minimum (zero) possible, but no absolute maximum possible (theoretically, a slot could keep filling in with Snoutling offers forever!). So, the distribution is naturally going to have a longer “tail” towards the high end, and a stouter “tail” towards the low end. None of that accounts for your dip in six-Snout-offers visits, but I think that simply must be random “noise,” like @burbman said.

[burbman]

@sutekh137, @poptimus – My recent observations, and anecdotally my recollections over time have shown a nearly 100% occurrence of the 75LC for all option coming up in the dojo by the time all snoutling offers have been exhausted.  Are the two of you seeing the same thing?

I would think that perhaps not because you each have six potential LC type offers, while I am down to five, and in exhausting snoutling offers, I would need to hit three of the LC options.

I am going to continue tracking this trend, as right now it is my suspicion that 75 for all is more highly weighted than the 15 for one family, which is in turn more highly weighted than the snoutling offers.

[Sutekh137]

@burbman
@poptimus

Not sure, but I just went to look at the Dojo, and it has three 15s staring at me — no 75. *smile* So, my first data point says the 75-for-all is not more likely…

I don’t trust myself for anything anecdotal, as the 75 is simple more noticeable because it is different. I am naturally going to think I see it more just because it stands out.

So, you’ll have to see what you come with statistically… I am getting tired of taking data so am not going to start doing this, too. And I am disgusted with this event, especially since out of the last 15 or so spins I have done (going back into yesterday afternoon), only one has included a 750-medal slot (I am trying to take advantage of the reward doubler, and it is causing me to lose almost entire days of time on the doubler refresh). Still languishing at 15th place on my board, virtually no chance of two stars. We’ll see if the pig dungeon opening up forces anyone to slow down above me as they go for LC instead of medals. But I really am getting to be entirely over this game. Taking data is fun and feels purposeful, but I am pretty sure that after my birds are all maxxed for mastery that I am done.

Thanks,
sutekh137

[Anonymous]

@burbman @sutek137

I just scrolled through some of the screenshots I’ve taken during the course of this research, and I’ve definitely got some screens with three 15 LC offers on them at once.

I hope you don’t mind, but I’m not going to track that aspect of dojo offers. I’m not going to ever buy an LC offer either way. (I would use the 5 LC reset option, if anything, but I’m fairly certain I’m never going to use that either. I prefer to use my LCs on the GPM and occasionally the Ancient Relict.)