I continue to get unrealistic numbers of 1-3 kingdom tributes - haven’t been tracking in detail like the others but statistically there’s just no way with all kingdoms 3* or above and level 9 or 10. This has only been happening for a couple of weeks for me at least - it was much more in line with expectations before that.
@Sirrian @Nimhain @Saltypatra Your input would be greatly appreciated here!
Here’s my tribute I track since I saw this thread:
8, 5, 5, 7, 3, 9, 6, 9, 2, 5, 5, 4, 4, 4, 4, 6, 9, 2
4, 7, 9, 8, 4, 7, 2, 5, 7, 9, 7, 5, 8, 5, 4, 6
8, 6, 5, 9, 5, 5, 5, 3, 4, 6, 4, 6, 4, 6, 7, 4
11, 6, 7, 2, 10, 7, 6, 8, 5, 7, 7, 7, 7
3, 3, 10, 4, 5, 7, 10, 3, 3, 3, 9, 6, 2, 4, 3
Can you calculate it for me @Rasper? I’m not so good with math and numbers XD. Thanks. 
you’re averaging at 21.2% over 78 collections
If you combine yours with mine and @Grundulum you get:
p = 19.6%
n = 329 collections, 8883 kingdom checks
99% confidence interval of 18.56% to 20.73% , so even your “good” run doesn’t bring our totals close to expectations.
Btw… What’s confidence interval?
I hate accusing someone of making the same mistake twice, but…
If something like this is the issue, it should be an easy enough check and fix. With wonky guild tribute amount bonuses in the mix, we’d have pretty much no way of knowing if a specific kingdom was being checked or not (barring someone with a screenshot of that one kingdom paying out, by itself) without a massive sample size and a bunch of estimates.
More data (with a 5% guild bonus, and all 27 kingdoms at 3+ stars):
With 3132 samples, I’m down to 20.72% rate of success.
95% confidence interval: 0.193 to 0.2214
Hey Guys,
We’ll try to take a look into this when possible. Unfortunately we’ve all been busy with Guild Wars.
As long as it is on the radar. In the grand scheme of things it is rather minor, at tops averaging roughly 0.5 fewer kingdom per tribute (5.2 vs 5.8). But if you collect a dozen a day, over a week it comes to 40 gems and 80 gold keys, and some gold/glory. But if there is an issue best to get it ironed out to prevent it compounding over time.
If you go by the full price currency value they assign to gems, gold, glory, souls in the store or in their specials, that’s a lot of value lost, like you said, especially multiplied by time and number of players.
(I collect 12/day.)
I agree, just getting it on the radar is fine. This is one of the longest support threads ever (you can sort by posts), it was getting a little frustrating that they wouldn’t acknowledge it, so I finally just sent a manual support request.
I’ve been continuing my tracking, more out of curiosity than anything else at this point since the devs said they would take a look.
177 collects
938 kingdoms
Average: 5.3 per collect
If it checked all 27 kingdoms, that would be 17727 = 4779 kingdom checks with 938 success, or a rate of 19.63%
If it checked 26 kingdoms, that would be 17726 = 4602 kingdom checks with 938 success, or a rate of 20.38%
If it checked 25 kingdoms, that would be 177*25 = 4425 kingdom checks with 938 success, or a rate of 21.19%
Since the expected result is 21.2%, that last number looks VERY likely! So, without having seen any of the code, my guess would be they put a counter to make sure we never got more than 25 kingdoms in one collect (which would make sense down the road, if we had 35+ kingdoms we’d need a cap to keep the rewards from scaling forever). But that counter is instead only checking 25 kingdoms instead of stopping when 25 report success.
(Just for giggles, adding mine to the other data gives a total of over 400 tribute collections, with success rates of 19.70, 20.46, and 21.28 for 27, 26, and 25 kingdoms.)
edit: Another very distinct possibility, the guild bonus is just not doing anything. At 205 collects my rate is 19.93%, if there is no guild bonus you’d expect 20%. And the aggregate data of 430 data points is 19.8%.
Sounds good. I’ve also been watching this for a long time and wondering why the statistics are off. Here’s hoping @Nimhain and @Sirrian will find time to look into this soon!
NowayJoe2Go
“Thanks for letting us know about this. We have added it to our bug list. We hope to have fix out as soon as possible.”
This was the support response I received today.
While better than nothing, it’s a little bit too copy-and-paste to be very convincing.
The support team (who you talk to when you submit a ticket not through the forums) doesn’t have access to the code base; they’re not the dev team (who you interact with here in the forums). That message you got might be the only thing they can say – that they’ve alerted the devs that there is a problem.
My final update. I’ve personally collected 250 tributes over the last few weeks, at 27 kingdoms each for a total of 6750 kingdom checks. Total kingdom returns 1341.
Average chance for a Kingdom to return a tribute: 19.87%
Average number of Kingdoms per collect: 5.36 (that works out to 4.36 gems and 8.72 gold keys)
From a statistical perspective, the chances of the actual rate being 21.2% is roughly 1 in 400. Not impossible, but certainly unlikely.
After all is said and done, my conclusion is the guild bonus from red statue is simply not working and we are getting the base 20% chance.
Adding in other folks data brought the total collections to 476 with an average of 19.8%, and the chances of it actually being 21.2 being closer to 1 in 3500.
I agree that’s always going to be the situation with customer support, but when they say “we have added it to our bug list” to whom does we and our refer?
And does “bug list” mean confirmed bugs to be fixed or possible/reported bugs to be verified?
I’d prefer they say “I submitted this to the [insert] team for review as a possible bug” if that’s what they meant.
Maybe I shouldn’t have said the copy and paste feel of the response wasn’t convincing, I should have said it was vague.
The CSR did a perfectly adequate job, I’m not trying to call them out in the slightest. GOW support has actually been the best I’ve ever had in any game that wasn’t small team/small community.
The Bug List is for reported bugs to be investigated. While investigating bugs, we will try to reproduce them on our end, and then provide the programmers with steps on how to cause the bug. This helps them to identify where the problem may be occurring.
The reason we sometimes ask for videos or screenshots of bugs, may be that we are unable to reproduce an issue internally. From there we can take the footage from players, to look at what they are seeing and present it to the programmers.
I’ll drop my final update here as well.
With 5% red statue bonus and all 27 kingdoms at 3+ stars:
1430 tributes from 7128 kingdom checks (264 collects).
Average chance: 20.06%
95% confidence interval: 0.1913 to 0.2099
At 26% (n of 27), the chance to receive 12+ Tributes is 10x the chance to receive 1 or 0 (about 0.3%).
At 20%, those chances are roughly reversed.
In my set of 250 I had a 1 Kingdom Tribute happen 9 times. Even had a 0 Kingdom Tribute happen once. But I did have a 13 and two 12s and 11s also.
If you combine my 250 with @SirCookie 's 264 you get
Average chance: 19.97%
95% confidence interval: 0.193 to .206
Nice when you get two big datasets that agree closely.
Just resurrecting this thread to give you guys an update.
Over the weekend we found and confirmed that the Red Guild Statue’s bonus was not applying as intended (it was giving a percentage bonus rather than an additive bonus - it was just in a difficult place to spot).
We were able to fixed this and deployed it out (thankfully it didn’t require a client update to resolve).