Bogus percentages in events

[DeletedUser12171]

From observation it seems to me this game is giving away more "consumable" items like instants and buildings that have limited life spans. I think it's fairly obvious items of value like sectors and excellent buildings are being rationed to prolong the money earning lifespan of Elvenar.

Big accusation there. If true, it means Inno is making false advertising

 

[SoggyShorts]

The payoff is whatever the programmer wants it to be, just like the electronic slots in Las Vegas.

This is true- slots are simply set to pay out a percentage 75% by law, but due to competition they are set between 80 and 98%.
There is no live programmer in the basement tweaking the formula as you spin though.

From observation it seems to me this game is giving away more "consumable" items like instants and buildings that have limited life spans.

Yes, consumables are more common than buildings-- it says it right there in the mouseover that a building is only a 25% chance, so you should expect to get 3x as many "not buildings"
As for giving us time-limit buildings, that's really up to the player since other choices are available.
Really, with all of the complaints about limited space, having more buildings with limited time is brilliant.

I think it's fairly obvious items of value like sectors and excellent buildings are being rationed

Of course there are limits to how many expansions and buildings you get in the game... that isn't hidden at all.

 

[DeletedUser4576]

Judging by how long programming things take I would say that such a system would have the benefit of employing at least 3 full-time workers

Electronic slot machines are programmed to pay off a percentage of the cash they take in. When they pay a high percentage they are called "loose". Lower payoffs are called "tight". The way this is adjusted is by programming how many different random numbers from the random number generator result in winning results. To change the programming you swap a single chip in the machine. No memory of previous play is required and the result any single play is not predetermined, just the average over time. Like changing the number of black slots on a roulette wheel. It wouldn't be difficult for an online game to do the same except they would probably boot a different module in the software load instead of changing a chip. Seems like the elephant in this room is the game's requirement to make a profit by selling items. The less you give away, up to the point players quit, the more you can sell to those willing to spend money.

 

[SoggyShorts]

@CEEDUBYA
I still can't tell if you are suggesting that inno is using a secret set of numbers in the background or not.

Everything in your latest post lines up with what we see. As for how inno could change the odds, it's a simple alteration in the database- like changing a 7 to a 9. Still a 3 person job for inno

 

[DeletedUser19592]

There is an easy way to test the percentages as reported by Inno...simply post trials and results. I'll go first:

On 138 pulls of the 50 point small chest I have received:
5 kp -- 22 times (~16% observed, 11% posted)
Gold Instant -- 20 times (~14.6% observed, 15% posted)
Supply Instant -- 18 times (~13% observed, 15% posted)
50 splinters -- 16 times (~11.6% observed, 15% posted)
90 splinters -- 20 times (~14.6% observed, 19% posted)
5 hr boost -- 32 times (~ 23.2% observed, 19% posted)
daily prize -- 10 times (~7.25% observed, 7% posted)

Given enough data we can check to see if the posted percentages are legit.

If we collect information on the rewards based on whether the daily prize is "good" or "bad" we can test if the percentages are independent of the quality of the daily prizes. Instead of complaining about your particular situation, post data and together as a community we can analyze the information.

 

[SoggyShorts]

Given enough data we can check to see if the posted percentages are legit.

If we got everyone who ever posted on these forums to flip a coin there is a very good chance that we'd end up with something like 60% tails reported.
Getting the ~20 players who regularly post here to show results has a decent chance of something like 75% tails.

You need at least 100,000 data points to come close to a meaningful analysis, and I wouldn't consider anything under 1 million.

 

[DeletedUser19592]

If we got everyone who ever posted on these forums to flip a coin there is a very good chance that we'd end up with something like 60% tails reported.
Getting the ~20 players who regularly post here to show results has a decent chance of something like 75% tails.

You need at least 100,000 data points to come close to a meaningful analysis, and I wouldn't consider anything under 1 million.

100,000! For a goodness of fit test 1,000 or so should be more than plenty.

 

[DeletedUser4576]

@CEEDUBYA
I still can't tell if you are suggesting that inno is using a secret set of numbers in the background or not.

Everything in your latest post lines up with what we see. As for how inno could change the odds, it's a simple alteration in the database- like changing a 7 to a 9. Still a 3 person job for inno

 

[DeletedUser19592]

To be fair, though, the post above was more a call for people to track their next event results since this one is already most of the way through.

 

[DeletedUser4576]

I'm suggesting Inno is in the business of making money from us, the game players. Having watched Volkswagen get caught putting rigged firmware in their cars I see no reason not to be cynical about what Inno tells me. In the first chapters of the game we got 4-5 sectors per chapter; now we get 1. Buying sectors used to be pretty cheap; then the price quadrupled. We used to be able to fight as far ahead as we were able; then the "ring" system put a stop to that. AW we already won and deployed suddenly dropped in value one day. I see a pattern. My theory is that this was conceived as a five chapter game and was surprisingly successful. Now it's been stretched to thirteen chapter but they have to be more and more stingy with what they give players. I still play but only when the reward seems worth the candle. Some seem content to play the game for the rest of their lives, even if it's another 50 chapters just like the last 8. I'm constantly trying to judge when some other game will provide more entertainment for the time spent. But it's almost impossible to make a critical comment here without having it ground into dust by the same half dozen who seem to live on this board. Just another facet of the game that isn't as much fun as it once was.

 

[SoggyShorts]

100,000! For a goodness of fit test 1,000 or so should be more than plenty.

http://www.shodor.org/interactivate/activities/Coin/

First try with 1,000 tosses I got 520 : 480
That's enough deviation to make someone wonder about the odds.

 

[SoggyShorts]

@CEEDUBYA
The difference lies in your own examples.
Inno gave 5 expansions in early chapters and later chapters get 1, did they lie about it? No, they didn't have to.
Inno raised the price of expansions. Did they lie about it? No, they didn't have to.
The fighting system was totally broken (killing armies 3x your size with zero losses?) Did they lie about it? No, they didn't have to.

Unlike VW there is simply no motive for dishonesty.
Players will get the buildings that they get, whatever the odds.
If they don't get as many buildings as they want, they will either suck it up or buy some.
Posting the odds at 25% when it really is 30% or 20% will not change that, so there is no mechanism where lying gets them more money.

Another way to look at it:
If the odds are actually lower than posted, then players will be upset because they feel "unlucky".
It's pretty safe to say that a sad customer isn't a happy customer.
It also is pretty well understood that happy customers tend to buy more.
It follows then that inno doesn't have a very good motive for making customers sad.

 

[DeletedUser9601]

First of all, this whole premise of "odds" assumes that the daily building is the goal. That is not always true for some players. So even to "lie" about the odds makes no sense for Inno, because ultimately every roll is giving a prize. The slot-machine analysis is interesting, but ultimately its not an analogue.

Buying sectors used to be pretty cheap; then the price quadrupled. We used to be able to fight as far ahead as we were able; then the "ring" system put a stop to that. AW we already won and deployed suddenly dropped in value one day. I see a pattern. My theory is that this was conceived as a five chapter game and was surprisingly successful. Now it's been stretched to thirteen chapter but they have to be more and more stingy with what they give players. I still play but only when the reward seems worth the candle. Some seem content to play the game for the rest of their lives, even if it's another 50 chapters just like the last 8. I'm constantly trying to judge when some other game will provide more entertainment for the time spent. But it's almost impossible to make a critical comment here without having it ground into dust by the same half dozen who seem to live on this board. Just another facet of the game that isn't as much fun as it once was.

Cheap sectors is obviously a problem. Look at the top 300 or so on every server. Those people realized the value, bought out everything they could, and now they're sitting pretty. So much less incentive to spend diamonds anywhere. The change to the pricing model in no way is a "trick" or Inno somehow acting poorly. They mis-priced their initial offering, and they corrected it. It happens all the time in the real world.

Same thing for the fighting, though its hardly as stark. I'm at the end of Amuni. I stopped taking optional squad sizes, and I'm at 420 provinces cleared. I'll easily get to 500 while every fight is still "easy" or "medium." Should it be "very easy"? No, probably not.

The one AW change (Crystal Lighthouse) was again an obviously needed change. Again, I'm in Amuni. If I had an old level 6 Crystal Lighthouse and visited every neighbor each day (which takes 5 minutes on my mobile device), I wouldn't need tier factories. That's obviously a problem, and its a mechanic that doesn't match the game design.

I agree 100% that Inno is relatively quick at fixing game problems that make the game too easy, and they drag their feet on pro-player changes (see Royal Restoration). But thats, like, every Free-to-Play game developer. I don't want you to think that the boards live to serve Inno's interests. I don't think that's the case, personally. But there's a difference between relatively absurd complaints about "bogus percentages" (which should on their face be dismissed, because 1 person rolling a dice 20 times, etc etc), versus actual complaints/honest discussion about the game (which I like to think are encouraged). There's a few forum commenters (myself, Soggy, others) that don't really encourage discussions about pie-in-the-sky suggestions ('gee, wouldn't it be great it we got 10 diamonds a day for logging in?I?" "well no, because Inno would just increase diamond costs of stuff, and they're trying to make money"). But I think each of the more recent chapters is at least "trying" to change up the guest race experience, and while the tourneys are stagnant and the FAs stink, there is some tweaking over time.

 

[DeletedUser5521]

AW we already won and deployed suddenly dropped in value one day.

I hear ya, CW. But one day a whole bunch of them got soooo much better..
'Member? Oh, I 'member..

 

[DeletedUser5806]

Normal distribution & gaming theory....At its core, the population is normally distributed in a true Bell Curve. It is the key criticism when applying that "label" to the population being sampled. Like a teacher grading on a "curve". Is the population even large enough? What are the assumptions?

Those percentages when applied to the population of all users of the game - all servers, ect. I do believe those % are accurate.

Simply analogy, a teacher doesn't have enough time to hate you. They Devs are incented to provide accurate information. The algorithms would be infinitely complicated in order to single a player. It can be done, but its too much work. They are already working on the next Chapter, upgrades, re-balancing, etc.

In two years, I have never gotten all 3 Grand prizes. This event, I will in both. But I didn't get all of the other pieces - because it took all my splinters playing on double days.

 

[DeletedUser19592]

"If we got everyone who ever posted on these forums to flip a coin there is a very good chance that we'd end up with something like 60% tails reported.
Getting the ~20 players who regularly post here to show results has a decent chance of something like 75% tails." -- SoggyShorts

http://www.shodor.org/interactivate/activities/Coin/

First try with 1,000 tosses I got 520 : 480
That's enough deviation to make someone wonder about the odds.

Your first example gives a result that would happen less than 2% of the time. Your second example gives a result that would happen more than 80% of the time. Only one of these would be considered unusual and cause people to wonder about the odds.

I'm not sure if your comments are meant to imply that:
1) there is no point to the exercise,
2) there is no way we could collect enough data,
3) there is no way we could collect data where we could be sure of its veracity.

If the problem is 1) or 3) there's is not much argument I can make to convince you. If evidence can't convince you, or if any evidence gathered here is suspect then there's not much to say. If the problem is 2) then you're simply wrong. If sigma_daily = sqrt(0.07*0.93) = 0.255 and we want to be able to detect if the actual percentage is more than 2% less than posted percentages then we'd need a sample of size n >= 2[(1.645 + 0.84)*0.255/0.02]^2 = 2013. So I was off by a factor of 2, but it's still nowhere close to 100,000. I've already posted 138 of the data points, so we only need 1875 more, which means about 13 or 14 more posters who've kept track of their results. Fewer if they've managed to collect more of the 50 cost chests (which is very likely since I'm well below average for the 50 and 90 moon splinter prizes).

Note: I do not believe that Inno has posted false percentages on their tooltips. I am merely pointing out how this debate can be definitively put to rest.

 

[SoggyShorts]

Simply analogy, a teacher doesn't have enough time to hate you.

Ha! I love it.

Your first example gives a result that would happen less than 2% of the time. Your second example gives a result that would happen more than 80% of the time. Only one of these would be considered unusual and cause people to wonder about the odds.

Your logic is sound, but your knowledge of people is lacking. Just read through any threads during events, there are numerous examples of the innumerate complaining that they are getting screwed by a rigged system when their results are well within 1 standard deviation.

I'm not sure if your comments are meant to imply that:
1) there is no point to the exercise,
2) there is no way we could collect enough data,
3) there is no way we could collect data where we could be sure of its veracity.

1. Because for over 2 years across 9 major events no matter how many people have posted their exact results, and no matter how we compiled the data players still believe in voodoo.
2. Even getting 14 players to all open the same size boxes is a massive chore. Just read through the threads- very early someone posts which box gives the best odds, and for the next 3 weeks players post about their results using the other box. That said, we have managed to get about a dozen players to post their findings in the past with very predictable results. The % were as expected, and so were the responses "Yeah, well... you all must have been lucky, cause clearly inno is messing with my rolls"
3. The data we collect is bound to be tainted. Most importantly players are vastly more likely to report poor results. At the other end, we've even caught players bragging about how lucky they were, but either they hit a 1 in a billion chance or they used diamonds.*Note, I don't recall the exact numbers(speaking of tainted data), but the worst was something like 27 buildings and all 3 GP in the first week of the event before finishing half of the questline.

Note: I do not believe that Inno has posted false percentages on their tooltips. I am merely pointing out how this debate can be definitively put to rest.

Check the threads for the past 2 years, I've read every single post, and I assure you that your facts cannot penetrate a tinfoil hat.
Seriously, there is no limit to how strongly people will hold on to their beliefs in the face of overwhelming evidence to the contrary.
I had more than 1 games manager and several pit bosses who would pull an "unlucky" dealer and replace them with one who was more "lucky" to cool down a table where the players were winning too much.
The owner of the casino approved of this practice!

 

[Deleted User - 1178646]

There is an easy way to test the percentages as reported by Inno...simply post trials and results. I'll go first:

On 138 pulls of the 50 point small chest I have received:
5 kp -- 22 times (~16% observed, 11% posted)
Gold Instant -- 20 times (~14.6% observed, 15% posted)
Supply Instant -- 18 times (~13% observed, 15% posted)
50 splinters -- 16 times (~11.6% observed, 15% posted)
90 splinters -- 20 times (~14.6% observed, 19% posted)
5 hr boost -- 32 times (~ 23.2% observed, 19% posted)
daily prize -- 10 times (~7.25% observed, 7% posted)

Given enough data we can check to see if the posted percentages are legit.

If we collect information on the rewards based on whether the daily prize is "good" or "bad" we can test if the percentages are independent of the quality of the daily prizes. Instead of complaining about your particular situation, post data and together as a community we can analyze the information.

for such a limited set of data you are already darn close to the actuel numbers displayed on the chest.

I acuired 13 buildings out of 167 chests (7.8%) seems like the odds aren't bogus at all.
I could see this by adding the spoints of the main rewards together and dividing it by 5.

In this streak if 167 chests I did get a daily prize free streak of 42 chests!!!!, but later I also got multiple buildings in a few tries. this combined evened out the odds to near 7%.

 

[DeletedUser12171]

Long story short: people losing the roll tend to think it's rigged, regardless of the truth.

Think opening at certain times of the day determines reward rates is absurd? Look in the other thread. There's even insinuations of race (human vs elf) being involved. Ha! And in previous events there were accusations of Inno comping spenders.

The only alternate scenario I'll even give any consideration to is the "dice with memory" scenario, i.e. The system knows how many of each prize has already been awarded and is set to award prizes in a way such as to maintain the published rates. I.e. the server population of payouts will have proportions exactly as published but certain individuals may get specific prizes disproportionately. Why this would be the case though, I can't for the life of me imagine, since an unbiased die roll with the same published rates will give the exact same overall server population result. There is no overall incentive for Inno to do it this way. Any system "rigged" to produce rewards at rates other than published amounts to false advertising, which is illegal.

It's an old story. I guarantee this same topic will appear every event from now until kingdom come

 

[Deleted User - 1178646]

Long story short: people losing the roll tend to think it's rigged, regardless of the truth.

It's an old story. I guarantee this same topic will appear every event from now until kingdom come

It's the 42 chest long streak most will remember, not the 3 buildings in 5 tries later on in the event at a building we don't like as much as at the 42 miss streak.