Castle change

[BantyRooster]

I noticed since the castle system came into the game the damage I receive in GBG has increased. What did inno do? It seems to be raise your castle or else go backwards...

 

[wolfhoundtoo]

There are various reasons you could be taking more damage in GBG versus last season. You don't give a lot of details so assuming that you haven't changed age or redesigned your city resulting in less attack/defense the results could simply be a combination of the terrain, the types of units that are facing each other and some random chance not going your way (AO for example or if you are in one of the ages where the troops have a random chance of doing extra damage you could just be having a run of bad luck).

Of course you could simply be misremembering the results from last season as well.

 

[BantyRooster]

I understand your point but, I did ask if anyone else in my guild was experiencing this and I was amazed so many thought it was just them. No, something changed. But thanks for your thoughts.

 

[The Lady Redneck]

I do not think it is anything to do with the Castle system as such. INNO has made changes (nerfed) battling in GBG I am in FE and despite increases in my Attack army A/D I noticed a while ago that I was taking more damage, and facing more difficult troop combinations. Last GBG and this one it became really noticeable ( for instance 4+ battles in a row where I was faced with a line up of 8 Hovers. Or a run where every 2nd battle is 8 hovers) I do not think it is to do with the Castle system, but is INNOs latest attempt to slow fighting down. As it means you have to stop to change your army combination and replace damaged units more often. Am not sure if they think that by slowing down those able to speed fight on autobattle it will give slower or weaker guilds more of a chance. Or if it is to try to prevent the use of bots some players use to fight. But all it is doing is slowing down the weaker guilds also and bots are just computer programs that can easily be reprogrammed to compensate. Or maybe they are trying to get people to spend more diamonds to heal troops with the heal all button so they can keep fighting fast without changing out troops. Whatever the reason it is making no real difference to my game as it is making me take a bit more care to make my actions more precise and the more often I have to change out troops the faster I become in doing so.

 

[RazorbackPirate]

I do not think it is anything to do with the Castle system as such. INNO has made changes (nerfed) battling in GBG I am in FE and despite increases in my Attack army A/D I noticed a while ago that I was taking more damage, and facing more difficult troop combinations. Last GBG and this one it became really noticeable ( for instance 4+ battles in a row where I was faced with a line up of 8 Hovers. Or a run where every 2nd battle is 8 hovers) I do not think it is to do with the Castle system, but is INNOs latest attempt to slow fighting down. As it means you have to stop to change your army combination and replace damaged units more often. Am not sure if they think that by slowing down those able to speed fight on autobattle it will give slower or weaker guilds more of a chance. Or if it is to try to prevent the use of bots some players use to fight. But all it is doing is slowing down the weaker guilds also and bots are just computer programs that can easily be reprogrammed to compensate. Or maybe they are trying to get people to spend more diamonds to heal troops with the heal all button so they can keep fighting fast without changing out troops.

Lots of conjecture based on nothing but anecdotal evidence. I see nothing here that would indicate anything other than a bad run of the RNG. There are too many factors at play that can effect results, terrain especially, which you don't see when all you do is autobattle.

While I agree that hard core PC fighters are negatively affected by the addition of the Heal All change, App players didn't miss a beat, making all your conjecture about Inno's motivations behind the change also nothing but conjecture.

Time after time, folks come to the forum with claims Inno changed this or that, with the one thing missing that will prove their point, documented evidence. Without it, it's just another player complaining about a bad streak.

Come back with documented evidence that all is no longer working according to the published RNG numbers and there will be something to talk about. Until then, sorry you're having a bad streak.

Whatever the reason it is making no real difference to my game as it is making me take a bit more care to make my actions more precise and the more often I have to change out troops the faster I become in doing so.

There's that muscle memory at play.

 

[Sharmon the Impaler]

Lots of conjecture based on nothing but anecdotal evidence. I see nothing here that would indicate anything other than a bad run of the RNG. There are too many factors at play that can effect results, terrain especially, which you don't see when all you do is autobattle.

While I agree that hard core PC fighters are negatively affected by the addition of the Heal All change, App players didn't miss a beat, making all your conjecture about Inno's motivations behind the change also nothing but conjecture.

Time after time, folks come to the forum with claims Inno changed this or that, with the one thing missing that will prove their point, documented evidence. Without it, it's just another player complaining about a bad streak.

Come back with documented evidence that all is no longer working according to the published RNG numbers and there will be something to talk about. Until then, sorry you're having a bad streak.

There's that muscle memory at play.

I carefully monitored my maximum attrition in GbG for the past 13 months and have adjusted my attack stats upward to maintain the number of battles that I fight at zero attrition before the units need swapping out at a certain and known level.

I have the stats for the past 8 months on this and it dropped by 8 fights in the past 3 seasons.

My max attrition before losing a fight on autobattle was 104 last month , this month it is 74. There was zero change except I raised my attack by 27 points. I average about 4500 fights per season and this is over 9 seasons @104 and the past 3 at 78 or as in this week 74. I consistently fight with the identical units and no rogues. With 50K+ fights used for that average I can exclude the RNG , Inno has changed the odds server side without a doubt.

 

[BantyRooster]

I appreciate everyone's honest input. I guess if it's happening to us all it's fair enough just disappointed about feeling that I went backwards. Happiness is a positive cash flow!

 

[Pericles the Lion]

I carefully monitored my maximum attrition in GbG for the past 13 months and have adjusted my attack stats upward to maintain the number of battles that I fight at zero attrition before the units need swapping out at a certain and known level.

I have the stats for the past 8 months on this and it dropped by 8 fights in the past 3 seasons.

My max attrition before losing a fight on autobattle was 104 last month , this month it is 74. There was zero change except I raised my attack by 27 points. I average about 4500 fights per season and this is over 9 seasons @104 and the past 3 at 78 or as in this week 74. I consistently fight with the identical units and no rogues. With 50K+ fights used for that average I can exclude the RNG , Inno has changed the odds server side without a doubt.

I concur. Before the current GBG season I was able to autobattle to 80 attrition before taking significant damage to my (1) hover tank (7) rogue army. I'm in PME. Depending on the number of guildmates on the field during a swap I get in between 225 and 275 battles each swap. This current season I'm not able to get past 60 attrition on auto. Nothing has changed except I added 96 attack and 32 defense when the Fall Event ended. More attack, more defense, less ability to fight without taking damage. I get it that terrain plays a role but I don't think it's likely to get "bad" terrain more often now, than before, unless something changed. I don't pay attention to the troops that I'm fighting because I'm using hovers.

 

[wolfhoundtoo]

But the troops you draw and the map you draw will impact your particular scenario quite a bit. You are using an advanced unit beyond the defense but the minimum damage you take is one when fighting (if you get hit). Maybe the auto battle changed a bit, maybe the troops you have been facing also have stealth and your one unit has been hitting them more often, or there are less plains on the maps you've drawn and your hover can be seen and thus shot.

 

[wolfhoundtoo]

I carefully monitored my maximum attrition in GbG for the past 13 months and have adjusted my attack stats upward to maintain the number of battles that I fight at zero attrition before the units need swapping out at a certain and known level.

I have the stats for the past 8 months on this and it dropped by 8 fights in the past 3 seasons.

My max attrition before losing a fight on autobattle was 104 last month , this month it is 74. There was zero change except I raised my attack by 27 points. I average about 4500 fights per season and this is over 9 seasons @104 and the past 3 at 78 or as in this week 74. I consistently fight with the identical units and no rogues. With 50K+ fights used for that average I can exclude the RNG , Inno has changed the odds server side without a doubt.

What age are you in?

 

[Sharmon the Impaler]

What age are you in?

I just aged up to SAAB today but was in SAM for 3 months. Units used are Sentinels X 8 and attack stats are 1699/1200.

 

[wolfhoundtoo]

I just aged up to SAAB today but was in SAM for 3 months

Well 1st your time frame you tracked was for different ages. That is a change. Your data for SAM is roughly 3 months worth (still a lot) but not 8 months. SAM troops have a 20% chance of doubling damage so it's not that hard to get a run of luck either way to boost or lower your outcomes. That random chance though means that you can lose earlier on any given day.

Let's explore your numbers for SAM and how many times you've lost at 74 this week and what the highest was for SAM alone.

 

[Sharmon the Impaler]

Well 1st your time frame you tracked was for different ages. That is a change. Your data for SAM is roughly 3 months worth (still a lot) but not 8 months. SAM troops have a 20% chance of doubling damage so it's not that hard to get a run of luck either way to boost or lower your outcomes. That random chance though means that you can lose earlier on any given day.

Let's explore your numbers for SAM and how many times you've lost at 74 this week and what the highest was for SAM alone.

That was why I only quoted the 9 seasons. I will look over the spreadsheet from prior to SAM but I don't recall any big change in "luck" for VF or OC. I was in Oceanic for over a year and my max att was fairly flat. It dropped a bit in VF and SAM but I always boost my attack to get back to the same level before moving on. I am a PKI and virtualisation specialist and have been for over 15 years , I know RNGs and statistics very well.

 

[RazorbackPirate]

I concur. Before the current GBG season I was able to autobattle to 80 attrition before taking significant damage to my (1) hover tank (7) rogue army. I'm in PME. Depending on the number of guildmates on the field during a swap I get in between 225 and 275 battles each swap. This current season I'm not able to get past 60 attrition on auto. Nothing has changed except I added 96 attack and 32 defense when the Fall Event ended. More attack, more defense, less ability to fight without taking damage. I get it that terrain plays a role but I don't think it's likely to get "bad" terrain more often now, than before, unless something changed. I don't pay attention to the troops that I'm fighting because I'm using hovers.

Is attrition rising faster than advertised? All things being equal, if not terrain, which you never see, what's changed? The system is either working as advertised, or it's not. Even the data presented is incomplete. If it were a change in the AI, it would show in every area of the game that uses that same AI. GBG, GE, and hood PvP, yet no one is reporting any changes in the other areas.

What terrain do you get? RNG. What units are you up against? RNG. Autobattlers never see the first, most don't pay attention to the second. What is the combination of facing tough units on tough terrain? Less attrition? I think so.

I get you're seeing different results, but it still sounds like a bad run of the RNGs. Of course no one ever comes here complaining when the RNGs over perform in their favor. It's never tracked, we never hear a peep.

About a year ago, shortly after getting my Himeji Castle to level 58, I had a run where I got 200 FPs 6-7 times in a two week period. I also remember thinking at the time, "I'll pay for this later." Sure enough, I got 200 FPs once more the following week, then not again for about three months. I had a great 3 weeks, then a lean 3 months, which averaged out to the RNG performing as advertised.

I see nothing different here. I could be wrong, but there's far too little hard data to make any other conclusions. Every time anyone has ever measured the RNGs with a large enough sample size to draw meaningful conclusions, they've always shown the RNGs working as advertised. I know it's not the popular opinion, but until there's proof to the contrary, I've got to give Inno the benefit of the doubt. Inno has have a track record of playing fair with their RNGs.

 

[Pericles the Lion]

But the troops you draw and the map you draw will impact your particular scenario quite a bit. You are using an advanced unit beyond the defense but the minimum damage you take is one when fighting (if you get hit). Maybe the auto battle changed a bit, maybe the troops you have been facing also have stealth and your one unit has been hitting them more often, or there are less plains on the maps you've drawn and your hover can be seen and thus shot.

I understand everything in your reply and do not disagree with any of it. My question to you is this....do you really believe that random chance is responsible for a 20 point reduction in attrition max when taking into consideration hundreds of battles? That same random chance would have applied last month wouldn't it? Ceteris parabus, I should be able to fight to a higher attrition since I've boosted my A/D higher not a lower attrition.

 

[Pericles the Lion]

Is attrition rising faster than advertised? All things being equal, if not terrain, which you never see, what's changed? The system is either working as advertised, or it's not. Even the data presented is incomplete. If it were a change in the AI, it would show in every area of the game that uses that same AI. GBG, GE, and hood PvP, yet no one is reporting any changes in the other areas.

What terrain do you get? RNG. What units are you up against? RNG. Autobattlers never see the first, most don't pay attention to the second. What is the combination of facing tough units on tough terrain? Less attrition? I think so.

I get you're seeing different results, but it still sounds like a bad run of the RNGs. Of course no one ever comes here complaining when the RNGs over perform in their favor. It's never tracked, we never hear a peep.

About a year ago, shortly after getting my Himeji Castle to level 58, I had a run where I got 200 FPs 6-7 times in a two week period. I also remember thinking at the time, "I'll pay for this later." Sure enough, I got 200 FPs once more the following week, then not again for about three months. I had a great 3 weeks, then a lean 3 months, which averaged out to the RNG performing as advertised.

I see nothing different here. I could be wrong, but there's far too little hard data to make any other conclusions. Every time anyone has ever measured the RNGs with a large enough sample size to draw meaningful conclusions, they've always shown the RNGs working as advertised. I know it's not the popular opinion, but until there's proof to the contrary, I've got to give Inno the benefit of the doubt. Inno has have a track record of playing fair with their RNGs.

Your reply makes sense. Instead of approaching this as a matter of RNG what if the variable change involves the damage inflicted/sustained figures for the defending troops? When manually battling, a player can see the expected amount of damage his unit will inflict when attacking a defender's unit. The amount is expressed as a range (e.g. 2-5). What if the code was "tweaked" so that more "2s" occurred and fewer "5s"? The RNG terrain and defender selection can remain unchanged but the battle outcome would be altered in favor of the defending army. Like you, I've had streaks with my HC, both hot and cold. However, a 58 HC is good for 9 charges per day, 810 in 3 months. 810 battles is easy in a single GBG day. I can see a cold streak in a single 250 battle swap but a dozen?

 

[RazorbackPirate]

what if...

What if...

See the problem?

While a reasonable hypothesis, it now needs proof. You are correct that, "When manually battling, a player can see the expected amount of damage his unit will inflict when attacking a defender's unit. The amount is expressed as a range (e.g. 2-5)." Now is the time to measure. Does it inflict damage at 25% 2-5 as expected, or is it now weighted to inflict more damage?

If so, why is it not showing up in the other places this same AI is used? Having just finished GE IV manually battling, and having manually battled in GBG at higher attrition today, I can tell you anecdotally, same as it ever was.

I also surrendered 5 times in a row in GBG because of crap terrain facing too many long range units. When I did finally got a favorable terrain, I took little damage, as opposed to being assured of losing multiple units.

 

[wolfhoundtoo]

I understand everything in your reply and do not disagree with any of it. My question to you is this....do you really believe that random chance is responsible for a 20 point reduction in attrition max when taking into consideration hundreds of battles? That same random chance would have applied last month wouldn't it? Ceteris parabus, I should be able to fight to a higher attrition since I've boosted my A/D higher not a lower attrition.

Random is random. So yes it could be quite different from last month. Especially since I doubt you've actually tracked the data in individual battles. Unlikely chances occur and when they do you tend to remember them. Also you seemed to imply that you only had one hover craft (how many did you get from the map it can't be that many) so at what point do you stop autobattling because you don't want to risk losing it? Assuming you get only 1 damage per hit how low will you allow it go and risk it being killed? So yes a bit of chance falling the other way would cause a swing. Your comparison of highs to lows really across hundreds of battles doesn't really enhance your argument.

 

[wolfhoundtoo]

That was why I only quoted the 9 seasons. I will look over the spreadsheet from prior to SAM but I don't recall any big change in "luck" for VF or OC. I was in Oceanic for over a year and my max att was fairly flat. It dropped a bit in VF and SAM but I always boost my attack to get back to the same level before moving on. I am a PKI and virtualisation specialist and have been for over 15 years , I know RNGs and statistics very well.

Ok but you are still comparing the highest to lowest events. How many battles do you lose like that when auto? Do you stop after losing one auto battle? As for knowing RNS and statistics very well good for you. Bring the actual statistics to this conversation then not vague statements that imply lots of data but don't actually present any data for anyone else to analyze. Numbers in other words. I'm willing to bet when you work you use actual data points to make your points.

P.S. Your job has nothing to do with validating your data.

 

[Sharmon the Impaler]

Ok but you are still comparing the highest to lowest events. How many battles do you lose like that when auto? Do you stop after losing one auto battle? As for knowing RNS and statistics very well good for you. Bring the actual statistics to this conversation then not vague statements that imply lots of data but don't actually present any data for anyone else to analyze. Numbers in other words. I'm willing to bet when you work you use actual data points to make your points.

P.S. Your job has nothing to do with validating your data.

I will parse the data from SAM start to SAM end and have it in a form that won't get deleted by weeks end. I count the 10th time an autobattle fails in a row as my Max attentuation.

I design and analyze hardened VPN tunnels for my employer and multiple seed sources are used with each envelope using 2 seeds for each to ensure randomness. I also have to constantly check the deployed installations daily to ensure that they stay within the required parameters , sending RFCs in whenever it deviates from what is expected.

A 50K dataset is considered as returning a valid sample to our engineering team when the RFC is sent in. I am sure that FoE is less stringent with what is an acceptable min/max value.