Mastery Points

Home Forums Angry Birds Epic Forum Mastery Points

Viewing 25 replies - 2,201 through 2,225 (of 2,264 total)
  • Replies
  • burbman
    @burbman

    Fair enough guys!  Good luck!

    Sutekh137
    @sutekh137

    @burbman

    And to you!

    FWIW, I just had another Dojo run and ended up looking at three 15s. So you are 100% wrong from the perspective of my two massive data points. *smile*

    Thanks,
    sutekh137

    burbman
    @burbman

    …And that is why we test…

    burbman
    @burbman

    Grind is essentially over after the anniversary event.  All of Red’s classes are now maxed, along with four of Matilda’s classes, and one of Chuck’s.  All in, I only need 24,780 mastery points to level all seven of the remaining classes of Matilda and Chuck up to 80, then I can buy all the Blues mastery I need in one marathon dojo trip.  I will post my final statistical evaluation later today.

    Anonymous

    Update on my dojo research:

    I’m up to 95 data points now with 23 available classes:

    – Average # of Snout offers per visit it up to 5.45

    – Initial offers are 86.3% Snoutlings

    – Replacement offers are 52.5% Snoutlings

    Other than the classes I maxed out quickly, the rest are all M73 or M74, so I’ll have plenty more data points before it’s all said and done. I actively avoided doing well in the Anniversary event (I got one star, which spun up 2000 mastery for all classes), which helped.

    Sutekh137
    @sutekh137

    @burbman

    Wow, you will be maxxed out in 50 days or so — that is IMPRESSIVE.

    The Event was OK for me — I scraped out 10th place by the skin of my teeth beat out 11th place 104,000 to 102,000 or so). I think the other guy must have used all of his FE and Stamina just like I did (no FE left, only 4 stamina left). I took a calculated risk and didn’t hit the Golden Pig cloud castle at all, so I left 40 LC on the table. But my two stars rolled for 6000 on Blues and 6000 on Bomb. That means I got an extra 36,000 mastery with that second star whereas with 40 LC I would have gotten around 24,000 at best via Dojo resets. And I didn’t spin Martha (my laggard), which would have sucked. So, all in all, can’t complain.

    @poptimus

    The Event reward did mean Bomb’s Seadog got maxxed out. So now I am grinding with Wizard (which I got the Elite for!), Spies, and Paladin to get more singletons maxxed out. I know my data is already skewed, because my first five Dojo trips with one class maxxed hit for 10, 9, 5, 9, and 6, for an average of 7.8. Not going to get nearly as many data points in this stage as the “pure” checks (where I got 253 points in and ended at 5.90 +/- 0.1752), so we shall see. Your average has come up a bit, and I would have expected it to be a bit lower with having 7+ single classes maxxed out.

    Great work, everyone!

    burbman
    @burbman

    @sutekh137, @poptimus

    I have combined all three of our now complete “pure” data sets and come up with 573 data points.  Our three averages came in within .04 of an offer of each other, so I am very confident that this is a representative sample.

    Combined Average – 5.90 offers

    Median – 5 offers

    Mode – 4 offers

    Max – 16 offers

    Min – 0 offers

    St. Dev – 2.7348

    St. Err. – .1142

    In comparison, I was able to get exactly 100 data points with only the six bomb classes maxed out.

    Average – 5.66 offers

    Median – 5 offers

    Mode – 5 offers

    Max – 12 offers

    Min – 1 offer

    St. Dev. – 2.3665

    St. Error – .2367

     

    Since receiving the award for the Anniversary event, I have not had more than one or two dojo trips where I have had the same number of potential offers, so I have stopped any further tracking.

    Sutekh137
    @sutekh137

    @burbman
    @poptimus

    So, at 5.66, nothing really to indicate taking 6 snoutling offers out of circulation and one LC offer out of circulation at the Dojo makes much of a difference? At least, everything is within error bars… But, also nothing really to indicate maxxing out whole birds is beneficial, so it would seem an even grind upward is the way to go?

    By the way, the first post on this thread is from 2 years and 8 months ago. We just had the 3rd birthday party event for Epic. Did mastery really get introduced when the game had only been out 4 months? Or am I just confused by the timeline and timestamps here?

    Thanks,
    sutekh137

    burbman
    @burbman

    @sutekh137

    I have come to the conclusion that the best approach is to keep all 30 bird classes in play for as long as possible, while focusing efforts on four of the five bird families.  The downside to maxing out a few classes early is not much, but it is real.

    Mastery after four months of the game being released sounds about right.  Early on, there was a different Gold/Silver/Bronze upgrade system in place that allowed players to pay Snouts for the first two upgrades per class, but cost LC to reach the gold class.  The earned mastery was a welcome change for those of us who were playing at that time.

    Sutekh137
    @sutekh137

    @burbman

    Ah, yeah, good memory! I had forgotten about that old way of doing mastery. I just remember being done with the standard game campaign, maybe for a long time, but if I was playing a lot, 4 months was probably a reasonable time for me to have been finished a while… Really puts into perspective how (if you leave off the Arena, which I do) little has changed in the past 2+ years. Quite honestly, nothing much new at all, just variations on the theme with more Epic gear added, more classes added, the Chronicle Caves, and the addition of Elite classes (which are barely even a variation. Most changes are 10% here and 10% there with no actual new content being added in over 2.5 years. That’s why this grind is that much closer to being my last.

    I agree with your overall mastery assessment, except I am not convinced yet that maxxing out classes is that big of a deal. As late as 137 data points in, I was at an average of 5.61 on “pure” visits — that’s lower than your 5.66 when you had slightly lowered your snoutling/LC offer ratio (though not by much). I hope I can get enough data points to find out more, though I only have two classes maxxed out so far and have had lucky data points to kick things off (nothing below 5 so far).

    Thanks,
    sutekh137

    bchild
    @bchild

    With 3 maxed classes, I averaged 5.5 in 84 visits.

    With 7 maxed classes (6 chuck classes and 1 bomb class), I’m averaging 6.52 in 46 visits thus far.  The distribution shows two humps, one around 4 and 5 (8 and 10 visits, respectively) and another at 9 (7 visits).  6, 7, and 8 had (4, 5, 4 visits respectively).  I did have one visit each with offers of 11, 13, 14 and 15.

    I expect this average will slowly start to drop.

    Anonymous

    Now I’ve got 134 data points with seven classes maxed:

    5.46 Snoutling offers per visit

    85.6% of initial offers are Snoutling offers

    53.0% of replacement offers are Snoutling offers

    Sutekh137
    @sutekh137

    @poptimus

    @burbman

    Sixty-four days in, and my data with 8 classes maxxed out (but no birds maxxed out):

    Number of data points: 72
    Average snoutling offers at Dojo: 5.53
    Max offers: 13
    Min offers: 1

    Average since I started maxxing classes (at least one class maxxed) is 5.82 on 101 data points.

    The 5.53 is close to poptimus’s numbers, so there definitely seems to be a decline. What is odd (to me) is that the decline seems so modest. My ratio of snoutling offers to LC offers is 3.67 instead of the optimal 5.00, so if this really is about fewer snoutling offers resulting in fewer snoutling refreshes as slots are used, why wouldn’t the average number of offers be a lot lower for me? At least 20-25% lower?

    All in all, I am ending up like burbman, saying that it is still best to level everything up at one time, not maxxing anything out until everything is close (and still leave one entire bird behind). I have no idea what actual algorithm is in play, but removing snoutling slots definitely seems to make the average take a hit (around 6% for me thus far).

    Thanks,
    sutekh137

    burbman
    @burbman

    @sutekh137 – Keep up the good work!  One thing that I found, after I had maxxed out one full bird, was that the 75 for all option showed up more often than I would have expected it to on a purely random basis.  Approximately 80% of my dojo visits finished up with that as one of the three offers showing.  Far more than the 60% I was anticipating, which leads me to believe that the game replaces the 15LC for bird class X, where X is the completed bird with an additional 75 for all, but cannot show more than one of the 75 cards at any one time.  That said, I also saw some of my longer runs after the 75 for all card was one of the initial offerings, which reinforces my thoughts.  Again, I only got 100 visits with the one complete family to test, so this could all be small sample size bias, but my theory is as follows:

    • Player starts with a randomly shuffled deck of cards consisting of one snoutling offer per bird class available, one 15 LC offer per bird family, and one 75LC offer
    • Selected offers are replaced randomly within the deck, and for every card replaced, an additional LC card is also added to the deck
    • Identical offers cannot be displayed simultaneously, so if a duplicate offer is drawn, it is replaced randomly into the deck
    • As 15LC offers are no longer available due to the lack of available classes, they are replaced by 75LC offers.

    I will need to test this theory in November when we get the next cap increase, but it supports (at least theoretically) the findings that @poptimus, yourself, and I have seen this go around, and supports the declining % of snout offers that @poptimus saw between the initial presentation, and replacement offers.  It also explains why I did not see dramatically longer runs as I got down to just two bird families remaining.  I would have expected runs to get much longer as there were only three LC offer types, and the likelihood of hitting all three LC offers quickly in a population where the snout offers were being replaced as selected seems to me to be quite low.  I should have had a 1 in 13 chance of the third space filling with an LC offer, yet my runs were never much longer than they were in the beginning.

    Anonymous

    @sutekh137 @burbman

    I agree with everything you two said today. Meanwhile, here’s the latest on my data collection effort:

    – 186 dojo visits with seven classes maxed

    – 5.47 snoutling offers per visit

    – 84.6% of initial offers are for snoutlings

    – 53.6% of replacement offers are for snoutlings

    For each bird, here are the %s of all snoutling offers:

    – Red (17.4% expected): 15.4% actual

    – Chuck (17.4%): 16.4%

    – Matilda (21.7%): 23.3%

    – Bomb (26.1%): 27.7%

    – Blues (17.4%): 17.2%

    Almost all my remaining classes are at M75, so still plenty of data to collect, but these numbers seem to have basically solidified at this point.

    Sutekh137
    @sutekh137

    @burbman
    @poptimus

    I agree as well, and I like the “deck” idea.

    Further analysis of expected reductions to snoutling offers, then:

    In a “pure” deck (nothing maxxed), that means 30 silver cards (83.3%) and 6 gold cards (16.7%).

    In my current deck (8 maxxed classes), that means 22 silver cards (78.6%) and 6 gold cards (21.4%).

    Even though 8 maxxed classes seems like a lot, it really isn’t that much difference when it comes to the percentage of snoutling “cards” in the “deck”. Eighty-three percent down to 79%. If I keep all the decimal places, the 30 vs 22 percentage reduction is 5.7% — not all that far off from the reduction off “pure” I am seeing in my 8-maxxed numbers.

    I guess the numbers do make sense when viewed from a “deck” viewpoint…

    Thanks,
    sutekh137

    Sutekh137
    @sutekh137

    @burbman
    @poptimus

    OK, after 76 days it is done — all bird classes are at M80!

    Final stats:

    • No birds or classes maxxed: 5.90 snoutling offers per visit (on average) with 253 data points (+/- 0.175)
    • 8 classes maxxed but no maxxed birds: 5.82 snoutling offers per visit (on average) with 130 data points (+/- 0.231)
    • At least one class maxxed out and no maxxed birds: 5.91 snoutling offers per visit (on average) with 170 data points (+/- 0.200)

    I had the Dojo showing all LC offers but then maxxed out my final non-Matilda class. Went back in and the “All Bomb” had turned into a Matilda class, and I was able to wind out all of her remaining classes (took around 33,500 snoutlings). I am still around 225,000 snoutlings at this time.

    Overall, I am not entirely sure about the best way to Dojo-up. It would seem doing all birds at once and not maxxing out anything would be best, but my overall average with at least one maxxed class is actually HIGHER than when I had nothing maxxed. But it stands to reason that keeping everything un-maxxed is the smartest thing to do, plus, of course, saving one whole bird back to level up en masse at the end.

    Here is the spreadsheet link one last time, where the top three lines contain the statistical breakdowns I mention above:

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

    Thanks!
    sutekh137

    Anonymous

    @sutekh137 @burbman

    252 data points with seven classes maxed:

    5.54 Snoutling offers per visit

    83.6% of initial offers are Snoutling offers

    54.8% of replacement offers are Snoutling offers

    Seems pretty clear to me at this point that all of our initial guesswork was right. The best thing to do is level all classes up at roughly the same pace, leaving one bird behind, etc., etc. I’ll keep collecting the data until the first of my remaining classes hits M80, but I’m not expecting any new revelations at this point.

    Jamairoqui
    @jamairoqui

    @poptimus That means it is more cost effective in terms of LCs to buy the 5 LC reset of the offers than the 1000-All 75 LC buy. There are five birds, 6 hats per bird, so a 75LC buy gives you 30,000 mastery points. But that 75 LCs would give you 15 resets, at 5.5 buys of 500 each per reset, giving 2700 per reset, or a little over 41,000 mastery for 75 LCs (and around 6000 snoutlings).

    And the economics gets worse the more birds are maxed out. By the time you get to 10 birds maxed, then you are getting only 20,000 mastery for that same 75 LCs, but would still get 41,000 for the 15 resets.

    Of course you need to have loads of snoutlings, but with the mouthpool snoutling farming technique, you can generate maybe 2000 snoutlings per hour, which is better than SRC used to be.

    Sutekh137
    @sutekh137

    @jamairoqui

    You are correct. In fact, @burbman pointed out that fact when I was talking about using saved up LCs at the Dojo when maximum mastery is revised upward. I had been inefficiently using the 75 LC option, and he set me straight right quick.  *smile*

    In summary for leveling up mastery:

    • Use Cave 9 for decent farming.
    • Always use LC on Dojo resets, never on LC-based offers.
    • Do not max out any birds. Level them all up until they are all close to maximize snoutling offers (not to mention event rewards and calendar mastery upgrades).
    • Leave one bird behind entirely in your grinding. This bird will show up in hundreds of snoutling offers once it is the only bird with available classes left.

    I think that’s about it.

     

    Thanks,

    sutekh137

    bchild
    @bchild

    @sutekh137

    I may have skewed the trend a bit…  With 7 classes maxed, 1 family (Chuck) and 1 Bomb class, in 74 dojo visits, I averaged 6.3 offers.  Need to confirm the exact numbers since I’m in a business trip and do not have my spreadsheet with me.  But it was 6+ average offers.  Much better than my average before maxing out one family.

    Jamairoqui
    @jamairoqui

    I am a master at restating the obvious, or already known.

    Just to be contrary, I think that you get better payout per minute farming Cave 4-5 for mastery than the Cave 9 levels. The Cave 9 levels take a while to get through on autoplay, and Cave 4-5 plays fast. Finally Cave 4-5 works for any two relatively high-level combination of birds.

    I’ve said that before too. Soon I’ll be reduced to telling the joke about the two peanuts who walked through the dark alley, and one was assaulted (the other blanched), over and over and over.

    Sutekh137
    @sutekh137

    @jamairoqui

    Please never change.  *smile*  I agree that which Cave to use is very subjective, especially when bird combinations come into play. Wizard and Spies rip through Cave 9 right quick. Other combinations — not so much.


    @bchild

    My average when having some classes maxxed (but no birds maxxed) was actually higher than my average before maxxing anything out. So I am with you. But @poptimus got more data points, and his numbers were somewhat lower. That being said, just a couple days of “lucky” trips to the Dojo and the average can really change. Still, it stands to reason leveling up all birds at the same rate will maximize snoutling offers (and has no downside unless you really need a maxxed class for some weird reason).

    Thanks,

    sutekh137

    DD
    @toolow

    These are some serious statistics. No doubt it’ll be useful for some, thanks to @burbman, @sutekh137, @poptimus and others who contributed.

    While I’m fascinated by it, I largely stuck to leveling up all the classes at the same time, purely for personal satisfaction. All my birds are now M80; would have got there earlier if not being mislabeled again as a cheat and missed out mastery payout from events.

    Pointless
    @pointless

    I myself grind 17-5 with borrowed, or 13-5 with two. Prefer the larger rewards, as I’m not always able to restart a level immediately…really does show the subjective preference, doesn’t it?

Home Forums Angry Birds Epic Forum Mastery Points

Viewing 25 replies - 2,201 through 2,225 (of 2,264 total)
  • You must be logged in to reply to this topic.