Exposing the attrition lie!

[TotalTrash]

Since Inno introduced the new GBG format, I’ve tested the real attrition gain versus Inno’s claims. Every day for the last 12 seasons, I went to 100 attrition and recorded the corresponding number of fights. I never negotiated, and all fights were done STRICTLY on sectors with sufficient support for a theoretical 80% reduction.
Mathematically, an 80% reduction (or 20% attrition gain) should yield 500 fights for 100 points of attrition. Counting the initial zero attrition on Thursday morning, every GBG season has 12 resets. Therefore, 12 x 100 = 1,200 attrition should result in 12 x 500 =6,000 fights, and 12 seasons 12 x 1,200 = 14,400 should yield 12 x 6,000 = 72,000 fights if the 80% claim is accurate.



This is just one example GBG, I have 11 more like it.



Results: The data clearly shows that Inno’s 80% reduction claim is false. Over the course of 12 seasons, I only averaged a 76% reduction. In fact, I only met or exceeded the 80% mark 31 times out of 144 days. Hence, I lost about 300 fights and spent an extra 60 attrition every season to reach the hypothetical 80% mark. It is not due to bad luck or a funky random generator but a systemic and statistically valid result.

The remaining question is whether Inno’s deception is deliberate or due to programming ineptitude. The answer is the latter, but the reason might be surprising. This message is getting too long, so I will post the explanation another time - just one hint: fight slowly, and avoid races where many of your guildmates help!

 

[captain987]

Interesting data. Have you done any tests for statistical significance (e.g. T-test)?

 

[Dursland]

avoid races where many of your guildmates help!

That's a good way to lose by 5 hits lol.

 

[WillyTwoShoes]

It's curious to me that when someone posts a statistical summary that lends credence to what many players have suspected all along they get a rash of posts concerning their methods/math/motives. Inno never has that problem, they just put something out and say what it is and everyone is expected to believe it like it was gospel without them ever having to prove anything.

Thank you for the effort TT!

 

[Pericles the Lion]

It's curious to me that when someone posts a statistical summary that lends credence to what many players have suspected all along they get a rash of posts concerning their methods/math/motives. Inno never has that problem, they just put something out and say what it is and everyone is expected to believe it like it was gospel without them ever having to prove anything.

Thank you for the effort TT!

@captain987 asked a fair question. @TotalTrash compiled a lot of data but only shared a small fraction of it. Running a one-sample T-Test on his data would only take as much time as it takes to input +/- 130 data points and click a button. The result of the T-Test would be a good indicator that actual attrition deviates (i.e. statistically deviates) from what INNO states that it should be. Or not. TT has already put in a lot of time....what's wrong with taking the final step?

 

[Ebeondi Asi]

Inno as always fudged on percentage. If players id not know that? time to wake up. It is pervasive and not just in GbG attrition. They fiddle with percentages for everything! Apparently the excuse is what they call game balance.
If they were a casino. they would all be in jail for cheating customers. The worst and most annoying to my mind is on desirable Daily specials in Events. That is the most blatant and annoying IMO.
If they were not fiddling, the long runs of positive for the player would happen. nope... Far more of the long runs against the player happen.. Gee wiz
They would enhance the game if they stopped fiddling the percentages, and just allowed chance a fair run.

 

[WarMachineX]

The fact that it is 76% and so close to 80 is reassuring. Because some days it feels like 0%. I get attrition for every fight that I do (for like 10 fights).

 

[Flynn013]

12 seasons is a good start. I would like to see data from a lot more seasons to see if the trend continues or swings towards where it should be.
I don't keep that accurate track of my numbers but that seems very close to my results.

 

[Angry.Blanket]

Inno as always fudged on percentage. If players id not know that? time to wake up. It is pervasive and not just in GbG attrition. They fiddle with percentages for everything! Apparently the excuse is what they call game balance.
If they were a casino. they would all be in jail for cheating customers. The worst and most annoying to my mind is on desirable Daily specials in Events. That is the most blatant and annoying IMO.
If they were not fiddling, the long runs of positive for the player would happen. nope... Far more of the long runs against the player happen.. Gee wiz
They would enhance the game if they stopped fiddling the percentages, and just allowed chance a fair run.

That is all minor stuff, easy to live with.

 

[captain987]

It's curious to me that when someone posts a statistical summary that lends credence to what many players have suspected all along they get a rash of posts concerning their methods/math/motives. Inno never has that problem, they just put something out and say what it is and everyone is expected to believe it like it was gospel without them ever having to prove anything.

Thank you for the effort TT!

Depending on your background and education, it is very possible that you have never learned about statistical tests, or forgotten about them (I don't mean that in any way other than face value). The fact is that tests for statistical significance are very important; for example pretty much no journal will accept anything with a p-value >0.05 (forget about a study without statistical significance tests). And it only takes a second to run, as both Google sheets and Microsoft excel have built-in T-Test functions. I can't see any reason not to run one.

EDIT: A test for effect size should also be run, I forgot about that. Pearson's r would be the easiest, but Cohen's d would also work.

 

[Dominator - X]

It's curious to me that when someone posts a statistical summary that lends credence to what many players have suspected all along they get a rash of posts concerning their methods/math/motives. Inno never has that problem, they just put something out and say what it is and everyone is expected to believe it like it was gospel without them ever having to prove anything.

Thank you for the effort TT!

I know, right? How dare anyone employ critical thinking and ask for expanded data and or proof.

 

[TotalTrash]

Of course, I have done a complete statistical analysis including beautiful graphs of 12 GBGs, 68,000 fights and 144 data points. In fact, I have gone to 130 attrition every single day, and have an additional 144 data points for that range as well. If I post all that on this forums, 99% of the user base would be unnecessarily confused and would miss the whole point. Suffice to say, the p-values prove that Inno's claim of 500 fights per 100 attrition is false.

But Inno did not fudge the percentages because the problem lies in the code! I looked closely at the worst and best days of every GBG season. The worst, 10 out of 12 times, was the Thursday after GBG opened. The best, 9 out of 12 times, was after the final reset on Sunday night.

Why is that?
On Thursday morning, the whole guild races the entire map. Each sector is taken in seconds with 30+ guildmates helping. Each individual member gets very few fights, but a disproportional attrition increase. I fight as fast the game mechanics allow without violating the Terms of Service (aka bots), and that's about 40% faster than any of my very skilled guildmates. Hence, I suffer the greatest detriment (one example, I incurred 24 attrition points for 35 fights during one race on an 80% supported sector). The faster you fight, and the more guildmates are helping, the worse your attrition gain will become!

Conversely, after the last reset of the season, the round is decided and nobody races anything anymore. Hence, I can fight at a leisurely pace, and often without other members helping on that sector, and my results actually exceeded 8 out of 12 times Inno's 80% claim.

None of this is "minor stuff"!
It is a huge deal because if fighting fast and fighting with your guildmates disadvantages you with respect to attrition gain, nobody wants to race a sector anymore. Inno wanted to make GBG more competitive, but this glitch disadvantages the most skilled players AND the most active, most competitive guilds!

 

[Pericles the Lion]

Of course, I have done a complete statistical analysis including beautiful graphs of 12 GBGs, 68,000 fights and 144 data points. In fact, I have gone to 130 attrition every single day, and have an additional 144 data points for that range as well. If I post all that on this forums, 99% of the user base would be unnecessarily confused and would miss the whole point. Suffice to say, the p-values prove that Inno's claim of 500 fights per 100 attrition is false.

But Inno did not fudge the percentages because the problem lies in the code! I looked closely at the worst and best days of every GBG season. The worst, 10 out of 12 times, was the Thursday after GBG opened. The best, 9 out of 12 times, was after the final reset on Sunday night.

Why is that?
On Thursday morning, the whole guild races the entire map. Each sector is taken in seconds with 30+ guildmates helping. Each individual member gets very few fights, but a disproportional attrition increase. I fight as fast the game mechanics allow without violating the Terms of Service (aka bots), and that's about 40% faster than any of my very skilled guildmates. Hence, I suffer the greatest detriment (one example, I incurred 24 attrition points for 35 fights during one race on an 80% supported sector). The faster you fight, and the more guildmates are helping, the worse your attrition gain will become!

Conversely, after the last reset of the season, the round is decided and nobody races anything anymore. Hence, I can fight at a leisurely pace, and often without other members helping on that sector, and my results actually exceeded 8 out of 12 times Inno's 80% claim.

None of this is "minor stuff"!
It is a huge deal because if fighting fast and fighting with your guildmates disadvantages you with respect to attrition gain, nobody wants to race a sector anymore. Inno wanted to make GBG more competitive, but this glitch disadvantages the most skilled players AND the most active, most competitive guilds!

So what do you suggest? Fight slower and only when fewer guildmates are on?

The main reason that I am skeptical whenever I read claims that INNO is "lying" is I do not see a motive. In the context of this thread, if they wanted to set the maximum attrition reduction to 75% (which is what your data seems to suggest) then all they needed to do was do it. What do they have to gain from setting it to 75% then claiming that it was set to 80%? Look at it from their perspective. Eliminating zero attrition upset a lot of players. Setting the max reduction to 75%, versus 80%, probably would not have caused that many more players to be upset. However, setting it to 75% while falsely claiming - over and over - that it was 80% sets the stage for threads such as this one. TBH, if it were my decision and 75% was the goal, I would have set it to 75% and then claimed that it was set to 70%. Then, players would be getting less attrition than expected. Like I said, I fail to see any motive for lying when they can easily achieve their goals by being honest.

 

[TotalTrash]

he main reason that I am skeptical whenever I read claims that INNO is "lying" is I do not see a motive.

Obviously, you didn't read my initial post thoroughly. I said the deception was NOT deliberate, but due to faulty programming.

So what do you suggest? Fight slower and only when fewer guildmates are on?

I suggest that Inno fixes the problem. Is that so hard to understand?

 

[jaymoney23456]

Of course, I have done a complete statistical analysis including beautiful graphs of 12 GBGs, 68,000 fights and 144 data points. In fact, I have gone to 130 attrition every single day, and have an additional 144 data points for that range as well. If I post all that on this forums, 99% of the user base would be unnecessarily confused and would miss the whole point. Suffice to say, the p-values prove that Inno's claim of 500 fights per 100 attrition is false.

But Inno did not fudge the percentages because the problem lies in the code! I looked closely at the worst and best days of every GBG season. The worst, 10 out of 12 times, was the Thursday after GBG opened. The best, 9 out of 12 times, was after the final reset on Sunday night.

Why is that?
On Thursday morning, the whole guild races the entire map. Each sector is taken in seconds with 30+ guildmates helping. Each individual member gets very few fights, but a disproportional attrition increase. I fight as fast the game mechanics allow without violating the Terms of Service (aka bots), and that's about 40% faster than any of my very skilled guildmates. Hence, I suffer the greatest detriment (one example, I incurred 24 attrition points for 35 fights during one race on an 80% supported sector). The faster you fight, and the more guildmates are helping, the worse your attrition gain will become!

Conversely, after the last reset of the season, the round is decided and nobody races anything anymore. Hence, I can fight at a leisurely pace, and often without other members helping on that sector, and my results actually exceeded 8 out of 12 times Inno's 80% claim.

None of this is "minor stuff"!
It is a huge deal because if fighting fast and fighting with your guildmates disadvantages you with respect to attrition gain, nobody wants to race a sector anymore. Inno wanted to make GBG more competitive, but this glitch disadvantages the most skilled players AND the most active, most competitive guilds!

There is no chance that you have worse attrition increase odds if you fight faster than others. Also, it seems like an exaggerated boast to claim that you are 40% faster than anyone else in your guild that you claim are 'skilled players'. If you fight at GbG start then odds are that you end up hitting one or more sectors that have less than 80% reduction-this happens many times, as a lot of the time the sector in front of your base will have just one camp slot or the perhaps the sector after that will have the 1 slot or both might. One player tracking attrition over 12 seasons that has marginally worse than 80% attrition reduction on those sectors is not statistically significant given how many thousands of players and fights are going on each day in GbG. If you tracked 1000 players over 12 seasons then you could have something you could draw more solid conclusions on. One player's stats is less than 1/10th of one percent of the total player base that competes in GbG.

 

[jaymoney23456]

Obviously, you didn't read my initial post thoroughly. I said the deception was NOT deliberate, but due to faulty programming.



I suggest that Inno fixes the problem. Is that so hard to understand?

Tracking one player cannot prove that attrition is skewed or that its not. Too small of a sample size compared to the total amt of players doing GbG.

 

[jaymoney23456]

12 seasons is a good start. I would like to see data from a lot more seasons to see if the trend continues or swings towards where it should be.
I don't keep that accurate track of my numbers but that seems very close to my results.

Need to track many hundreds of players at least to come up with statistically significant numbers.

 

[TotalTrash]

There is no chance that you have worse attrition increase odds if you fight faster than others. Also, it seems like an exaggerated boast to claim that you are 40% faster than anyone else in your guild that you claim are 'skilled players'. If you fight at GbG start then odds are that you end up hitting one or more sectors that have less than 80% reduction-this happens many times, as a lot of the time the sector in front of your base will have just one camp slot or the perhaps the sector after that will have the 1 slot or both might. One player tracking attrition over 12 seasons that has marginally worse than 80% attrition reduction on those sectors is not statistically significant given how many thousands of players and fights are going on each day in GbG. If you tracked 1000 players over 12 seasons then you could have something you could draw more solid conclusions on. One player's stats is less than 1/10th of one percent of the total player base that competes in GbG.

You are not a professional gamer, nor do you understand statistics. But I won't argue with you because your posts are meant to discredit the uncomfortable truth that Inno might be incompetent - that's what sycophants do.

 

[jaymoney23456]

You are not a professional gamer, nor do you understand statistics. But I won't argue with you because your posts are meant to discredit the uncomfortable truth that Inno might be incompetent - that's what sycophants do.

If you actually read a lot of my posts you would know I am not a fan of how inno has changed the game lately. That doesn't mean I think there is some sort of conspiracy about attrition going on lol You are just making up nonsense in your posts about attrition and don't like it when someone contradicts your crap.

 

[Sharmon the Impaler]

They even stated very plainly that the attrition will be dynamic from now on about 4 months ago. The reason why you lose more attrition from fighting in a gang is because those are the triggers for the "surge" attrition. A simple algo = If fights per minute > 1000 then attrition mod = 1.1 else Attrition mod = 1. Attrition = Attrition X attrition mod