Golden Pig Machine

[KerrAvon]

@mvnla2 Yes, it might just be look-up tables based on values calculated from some simple equations. The rounding can account for many differences. But with the Red/Matilda weapons, the factor is a whole number (i.e. 6), so rounding is not going to affect calculations. But as you say, maybe they have slightly adjusted values in the look-up tables.

[MVNLA2]

@KerrAvon @Dr_Ishmael — None of the numbers for the best-fit least-squares line through the data is an integer. For example, the multiplier for Red/Matilda is 6.047, not 6.

[dr_ishmael]

@mvnla2 If you plot a trendline in Excel, sure, that’s what it calculates (or something very close to that) for the trendline’s slope. But like I told @eric2203, you have to test that formula by seeing how accurate it is at re-calculating the known values – and it’s not, it’s worse than using 6.

[MVNLA2]

@Dr_Ishmael — I did compare the values calculated from Excel’s trend-line formula to your data + mine for several sets, and it was all +/- 1. I need to post my data, but I think it might only be 3 additional points.
What I said on the previous page, which you might not have read, was:
As far as I’ve seen so far, the values are no more than + or – 1 from perfectly linear after rounding to the nearest integer. What if we assume the Rovio programmers take undue delight in sending people like us on wild goose chases? They could have a look-up table for the set item values, and either intentionally or unintentionally increased/decreased some by 1.

[dr_ishmael]

@mvnla2 Yes, I did see that, I just didn’t have anything significant to say to it at the time.

I’ll agree that it is possible that they could have written out lookup tables for all of this, but as a programmer myself (albeit not a game programmer), that seems inefficient. For instance, whenever they increase the level cap, they would have to append the data for the new levels to all of the lookup tables; on the other hand, they don’t have to update anything at all if everything is driven by equations.

[KerrAvon]

@dr_ishmael I played D&D in my younger days. It was all look-up tables in those days.

I agree simple equations would be the most obvious and the easiest option. But they may have their own reasons. You just never can tell what programmers will do.

At this stage I think collection of data is the best we can do.

@mvnla2 A fraction above 6 does not really make a difference to the rest of the calculations. And you have to remember the 6.047 obtained from the calculation is biased by the data that we have in the data set. If the lowest number is one of the odd values (divisible by 6), then the value will be closer to 6.00.

[KerrAvon]

@partshade There was one value you provided that I had meant to ask you about. The Blues offhand item you listed as level 41 with a value of 864. It would make more sense if this was a level 40 item with a value of 864, or else the value should be 884. Sorry to ask again, but can you check. @dr_ishmael this may provide one more data point.

[dr_ishmael]

@kerravon D&D and other pen-and-paper games have lookup tables primarily because you can’t print an equation in your manual and expect players to calculate all the values themselves. Also, a lot of them simply didn’t follow any formulas at all – AD&D 2nd Ed. experience levels were all over the place (thankfully this was simplified to a quadratic formula based on level in 3rd Ed.).

As for @partshade’s data point, I must have ignored it because it said “level 41”, for which we already knew all the values. 864 definitely fits for Blues offhand level 40, though.

[MVNLA2]

It occurred to me that a look-up table might be faster than equations, but I am also not a game programmer although I’ve had way more than average programming experience.
Will post my complete data set later today. Off to play golf shortly; going to be warm/hot. I say this as I’m looking at the news stories about lots of snow in the northeast. : D. Weather person just said wind chill of -30 F!

[MVNLA2]

Here is my data. Other than the level 41 set items, I did not keep track of the level where I got them. The levels for the rest are either by comparison with @Dr_Ishmael’s table or from determining where they fall on the least-squares fit line.

Bird	Damage		Health	
	Level	Value	Level	Value
Red			17	668
			33	1152
	33	228		
			41	1394
				
Chuck	33	332	33	580
	41	400		
				
Matilda			20	609
	27	192		
	29	204		
			32	889
	33	228		
			41	1077
				
Bomb	29	326	29	826
			32	899
	33	364	33	924
			41	1118
				
Blues	29	326		
	33	364	33	730
	41	442	41	884

[dr_ishmael]

@mvnla2 Thanks for your data! It only added 4 data points to the collective table, though, since most of it is from the incremental level caps: 29, 33, 37, and 41. We had most of those levels filled in already; the only point we’re still missing from them is Bomb’s offhand for L37. [edited – I’d originally said Blues’ offhand was the one missing]

BTW, you’re off a bit on a Matilda’s offhand numbers: 609 fits at level 21, and 889 fits at 33.

[MVNLA2]

Dr_Ishmael — I knew there were only a few you didn’t have.You could be correct about mistakes on Matilda’s values. I guess I never actually calculated the values on the best fit curve for those. Sorry.

[KerrAvon]

@dr_ishmael These are the values for my set items at levels below 30. The levels for some of the items for my level 12/13 team are known.

weapons
red 144
matilda 197
chuck 208
bomb 182, 96
blues 288, 278

offhand
red 515 (12)
matilda 609
chuck 276 (13), 261 (12)
bomb none
blues 326 (12), 308

[Anonymous]

@dr_ishmael @kerravon You know now that you said that you just reminded me that I did in fact get the Offhand for Blues at level 40. Sorry for the mix up but at least it got us a new number.

[Razorblade]

There seems to be a bug with the golden pig machine
and i admit i exploited it two times already as i believed the first time was only a glitch :)

i don’t have friends that spin my wheel, but the green pig spins my wheel from time to time (every two-three days?) i noticed a few days ago and yesterday again, that when you click on the golden pig after the green pig spinned, your machine adds a little to it. when i close and immediately open the golden machine again, it adds a little again, and again, and again until the golden machine is full and you get a guaranteed set item by spending three golden coins. yesterday i obtained seven set items in less than an hour (for the rediculous amount of 21 golden coins). The bug stops if you don’t click the pig machine for some time or get out. I’m curious if it will happen again in a day or two …..

try it, want to know if others get the same benefit, so i don’t feel like a cheater :P

[Eric2203]

@razorblade

Interesting. There’s also the bug mentioned by @partshade where logging out of your Rovio account gets you the free roll by watching an ad every day instead of every 2 days.

Well, on my game (using iOS here), I get the free roll by watching an ad every time I log out and back in, as long as I forcibly close the app in-between, WITHOUT WAITING A FULL DAY. I’ve gotten several rolls in the past few minutes by logging in and out and terminating the app in-between.

[Eric2203]

Well it just stopped working… I got half a dozen rolls before it stopped.

[dr_ishmael]

@kerravon Thanks again!

GOOD NEWS! I just solved two of the series: Red/Matilda weapons and Bomb/Blues weapons.

Red/Matilda Weapon Attack = Bround(2.5 * (Level+5)) + Bround(3.5 * (Level+5) - 1) + 1

Bomb/Blues Weapon Attack = 2 * Bround(4.8 * (Level+5))

I started by figuring out the differential pattern for the Bomb/Blues series, which is +10/+10/+10/+10/+8 (@kerravon’s new data provided the clues I needed to get this). I quickly realized that if I divided this pattern by 2, I got 5/5/5/5/4, which indicated a factor of 4.8, which quickly led me to the B/B formula.

The Red/Matilda formula took longer, but I started by realizing that the series is actually +7/+7/+5/+5, not just +7/+5 as I stated before. Since these numbers are odd, instead of even like the B/B pattern, dividing by 2 wouldn’t be sufficient. Instead, I thought that maybe it indicated the sum of two different factors, which just happened to be equivalent for B/B. Even then, using factors of 2.5 and 3.5 still didn’t come out quite right until I introduced a couple unit offsets. I’m not exactly satisfied with having to do that, but at least it works.

As for the other series, I might be able to work out Chuck’s weapon and Matilda’s offhand when I have time, but everything else has insufficient data.

[dr_ishmael]

Oh, I also have another story to tell. Today was the Rainbow Riot on my main account, and since I didn’t have time to spin the GPM during last week’s RR, I had two weeks’ worth of LCs from the Golden Pig to throw at it today.

I made 23 3-coin rolls without getting a single set item.

That’s right. Twenty-three rolls. Not a single bloomin’ rainbow-starred item. This drastically affects my calculated drop rates – previously I showed a 17.6% chance for a set item, but after today it’s only 13.2%.

Dr Ishmael's Golden Pig Machine drop rates

      Rainbow Riot     Normal           Normal w/@mvnla2
      ---------------  ---------------  ---------------
Stars Count Frequency  Count Frequency  Count Frequency
----- ----- ---------  ----- ---------  ----- ---------
0        12     9.30%     13    26.53%     29    25.00%
1        17    13.18%     12    24.49%     23    19.83%
2        33    25.58%     14    28.57%     37    31.90%
3        50    38.76%      8    16.33%     20    17.24%
Set      17    13.18%      2     4.08%      7     6.03%
----- ----- ---------  ----- ---------  ----- ---------
Total   129               49              116

Full Rainbow Bar rolls are not counted here.

@mvnla2 I added our data together in order to get more accurate non-RR rates. In your last drop rate update, you noted that one of the set items was from a full bar, so I accounted for that here. Is that the only one you’re counting in your data, or were any others of those 6 set items obtained from full-bar rolls?

[MVNLA2]

@Dr_Ishmael — I only had one set item from a free roll in my dataset.

@eric2203 — @partshade said the log-out trick stopped working for him a few days ago, and that he did not get any free rolls from ads for several days. The log-out trick also stopped working for me yesterday. Will let you know when I get another roll. I was wondering if maybe it works only as long as there is an ad you haven’t seen before. I wasn’t paying close attention, but it seemed that a lot of the ads for free rolls were ones I hadn’t seen before.

[dr_ishmael]

@mvnla2 That’s what I thought, but thanks for confirming. :)

[Anonymous]

@eric2203 @mvnla2 So my trick was that logging in on Rovio Account and logging out at night would get you a free roll everyday. What happened is that I went 2 days without a roll from my average routine. Now after going 2 days around 10:00 PM I did get a roll on my normal machine. After that I logged out. When I logged back in I didn’t have a roll on the normal machine but I did on the Arena GPM. This is strange because this isn’t really a 24 hour thing. I suspect that if you DO follow my trick that it kinda works like a 2:00 AM reset thing. Somehow its been messed up but I have no idea how it happened. When I log in today I can say what I get.

[KerrAvon]

@razorblade Someone mentioned that bug in the main bug thread a few days ago. I only got one repeat when I last got a message from Piggy McCool, and am waiting for another go at this.

[Eric2203]

@mvnla2 @partshade Not sure why it worked for me a few times in a row today. Perhaps it’s a combination of things: logging out gives the ad every day and logging out a few times in a row works as long as the ads are different. Just a guess. I’ll try again tomorrow.

[MVNLA2]

@partshade @eric2203 — I got my normal (once every 48 hrs) free roll from an ad today, and it was a new ad.