Immortal Soul: Black Survival - AB Lars
Hello! This is GM Meiji, hiding in the research center and observing user behaviors.

Many fans have been questioning the gameplay speed difference between the mobile versions and the Steam version. In response, we’ve been analyzing this issue from multiple angles.

There have been multiple reports that the Steam version is noticeably fast compared to the mobile versions. We apologize for our delayed response.

In game development, we need to be able to replicate any gameplay issues that are reported. Even when many people feel that there is an issue, we cannot do anything unless that issue can be objectively replicated and tested. Without a clear cause we are forced to reconsider several hundred thousands of lines of code and that is simply impractical.

However, this does not mean we did nothing to fix it. Several GMs have teamed up as a task force and investigated the speed gap. It took us quite a bit of time as we had to work without a clear cause. But your feedback and input have helped us!


1. Steam User Demographics
The following table shows the percentage of Steam users per league. About half of all Steam users are based in Korea. Since the Steam version is only a few months old, Steam users tend toward the Herbivore League. However it is true that, higher you go in the Carnivore League, more often the game is played on Steam.



2. Problem at hand
After testing on smartphones and PCs of above-average specs, we discovered that the Steam version can attack 1 more time per 16 attack cycles on average. This is roughly around 6% in speed difference. However, this is a symptom, not the cause. Working from here took some time.

After a while, we have identified four possible causes.

1) Button activation moment
Upon encountering an enemy, buttons slide into the screen from the sides and become active (pressable) once the animation is over. Mashing hotkeys seems to cut the time between button activation and input recognition. The time difference is too short to be noticed with naked eyes, but replaying the video recording in half time showed significant difference. With the next update, the button activation moment will be cleaned up so that this no longer happens.

2) Button press vs. button release
On the Steam version, user input is registered with the button press. This is different on smartphones because of physics and engineering – on smartphones, user input is registered with the button release. Mashing the button practically solves the issue but this is not an insignificant problem. With the next update, we will fix this issue by implementing a tiny user input moment correction.

3) Touch/slide sensitivity
Sometimes we get feedback saying that button taps were ignored by the game. We suspect this is due to minute sliding when tapping the screen. On touchscreen devices, pressing the screen and releasing immediately registers as a touch event. Pressing the screen and then moving a bit before releasing registers as a slide event. Black Survival may be too sensitive in distinguishing the two. However, this is not something we can fix immediately as we use Unity3D’s NGUI. We will investigate this issue in the long term. In the meantime... please be quick with your touch input. :(

4) Hotkey > Touch
Tapping/pressing a button repeatedly eliminates most speed differences when 1) and 2) are resolved, but it still does not fix other issues, such as button/hotkey distance.
We believe it is not possible to balance platforms perfectly, but it is still something we should aspire to accomplish. If there is a one-tenth second difference in crafting an item, that is 5 seconds of delay after crafting 50 items. We will try to prevent these inevitable speed gaps from being unfair exploits. One exploit we are removing is the hit-and-run tactic that exploits minimap movement during counterattacks/resting. When Action UI pops up in the middle of the screen, activating the minimap will cancel the action.
If there are other issues, please let us know!

3. Device spec factors

Well, above issues seem crystal clear. What took us so long?

During analysis, we constantly came across the issue of device specs. You are not just playing Black Survival on a smartphone – the device is going through a number of technical hoops and hurdles to get it running. This makes precise testing difficult. As examples…

First, there is the issue of signal carriers. Telecommunication companies use a range of techniques and technology to ensure that all customers receive quality service. One major issue is load distribution. They must avoid overloading a cell tower; compensate for topography and concentration; and take all sorts of measures to make service consistent. As such, even when your device shows you are connected to an LTE network, your signal may be constantly jumping networks you are actually connected to. Some signal carriers have their networks configured so that there is data loss when the jump happens.

WiFi networks are also problematic. Your device showing that your WiFi network is at full strength is actually showing the network strength between your router and your device and says nothing about your actual Internet connection. Some cell carriers provide dynamic IP, which is comparable to what we explained above. If data is lost while your connection jumps, you may suffer latency issues.

These are nearly imperceptible when browsing the Internet or watching streams, but with Black Survival these can cause errors with the event log. From the server side, this resembles speedhacks and such abuses, so it will be handled as a special case as a precautionary measure. “Everyone else is fine but I lagged to death!” is often attributable to this issue.

Further, with Android devices…

Android and iOS are fundamentally different from the architecture level. iOS uses CPU and such resources directly, but Android is akin to an emulator operating in layers. With the exact same device, Black Survival will perform better on an iOS one. The OS side of software engineering is very difficult – even the people who made the OS themselves deal with bugs and lags. Android firmwares are more colorful in their problems. That’s why sometimes even expensive Android phones have dealing with simple games.

Black Survival does not use much 3D graphics but it does a lot of calculations. It places more strain on the CPU than the GPU. That’s why Black Survival might feel slower compared to its required specs. Speaking of which, devices with poor GPU might be the better device for Black Survival, given right conditions.

Beyond that, sometimes bad weather, old cell tower, mismanaged firmware updates, etc. can be sources of problems. Speaking from experience, something truly random might affect game performance.

We emphasize that minimizing device differences is our responsibility. Compared to when Black Survival first launched, we’ve made significant progress in that regard. However, there are always variables we did not account for and it is extremely difficult to solve for all of them. We will still try regardless.

So far, bugfixes have been highly technical issues and we only talked about their end results in patch notes. In the future, however, we will put an effort to talk about frequent 1:1 feedback in notices and development journals.

Thank you.
Immortal Soul: Black Survival - AB Lars
Hello! This is GM Meiji, hiding in the research center and observing user behaviors.

Many fans have been questioning the gameplay speed difference between the mobile versions and the Steam version. In response, we’ve been analyzing this issue from multiple angles.

There have been multiple reports that the Steam version is noticeably fast compared to the mobile versions. We apologize for our delayed response.

In game development, we need to be able to replicate any gameplay issues that are reported. Even when many people feel that there is an issue, we cannot do anything unless that issue can be objectively replicated and tested. Without a clear cause we are forced to reconsider several hundred thousands of lines of code and that is simply impractical.

However, this does not mean we did nothing to fix it. Several GMs have teamed up as a task force and investigated the speed gap. It took us quite a bit of time as we had to work without a clear cause. But your feedback and input have helped us!


1. Steam User Demographics
The following table shows the percentage of Steam users per league. About half of all Steam users are based in Korea. Since the Steam version is only a few months old, Steam users tend toward the Herbivore League. However it is true that, higher you go in the Carnivore League, more often the game is played on Steam.



2. Problem at hand
After testing on smartphones and PCs of above-average specs, we discovered that the Steam version can attack 1 more time per 16 attack cycles on average. This is roughly around 6% in speed difference. However, this is a symptom, not the cause. Working from here took some time.

After a while, we have identified four possible causes.

1) Button activation moment
Upon encountering an enemy, buttons slide into the screen from the sides and become active (pressable) once the animation is over. Mashing hotkeys seems to cut the time between button activation and input recognition. The time difference is too short to be noticed with naked eyes, but replaying the video recording in half time showed significant difference. With the next update, the button activation moment will be cleaned up so that this no longer happens.

2) Button press vs. button release
On the Steam version, user input is registered with the button press. This is different on smartphones because of physics and engineering – on smartphones, user input is registered with the button release. Mashing the button practically solves the issue but this is not an insignificant problem. With the next update, we will fix this issue by implementing a tiny user input moment correction.

3) Touch/slide sensitivity
Sometimes we get feedback saying that button taps were ignored by the game. We suspect this is due to minute sliding when tapping the screen. On touchscreen devices, pressing the screen and releasing immediately registers as a touch event. Pressing the screen and then moving a bit before releasing registers as a slide event. Black Survival may be too sensitive in distinguishing the two. However, this is not something we can fix immediately as we use Unity3D’s NGUI. We will investigate this issue in the long term. In the meantime... please be quick with your touch input. :(

4) Hotkey > Touch
Tapping/pressing a button repeatedly eliminates most speed differences when 1) and 2) are resolved, but it still does not fix other issues, such as button/hotkey distance.
We believe it is not possible to balance platforms perfectly, but it is still something we should aspire to accomplish. If there is a one-tenth second difference in crafting an item, that is 5 seconds of delay after crafting 50 items. We will try to prevent these inevitable speed gaps from being unfair exploits. One exploit we are removing is the hit-and-run tactic that exploits minimap movement during counterattacks/resting. When Action UI pops up in the middle of the screen, activating the minimap will cancel the action.
If there are other issues, please let us know!

3. Device spec factors

Well, above issues seem crystal clear. What took us so long?

During analysis, we constantly came across the issue of device specs. You are not just playing Black Survival on a smartphone – the device is going through a number of technical hoops and hurdles to get it running. This makes precise testing difficult. As examples…

First, there is the issue of signal carriers. Telecommunication companies use a range of techniques and technology to ensure that all customers receive quality service. One major issue is load distribution. They must avoid overloading a cell tower; compensate for topography and concentration; and take all sorts of measures to make service consistent. As such, even when your device shows you are connected to an LTE network, your signal may be constantly jumping networks you are actually connected to. Some signal carriers have their networks configured so that there is data loss when the jump happens.

WiFi networks are also problematic. Your device showing that your WiFi network is at full strength is actually showing the network strength between your router and your device and says nothing about your actual Internet connection. Some cell carriers provide dynamic IP, which is comparable to what we explained above. If data is lost while your connection jumps, you may suffer latency issues.

These are nearly imperceptible when browsing the Internet or watching streams, but with Black Survival these can cause errors with the event log. From the server side, this resembles speedhacks and such abuses, so it will be handled as a special case as a precautionary measure. “Everyone else is fine but I lagged to death!” is often attributable to this issue.

Further, with Android devices…

Android and iOS are fundamentally different from the architecture level. iOS uses CPU and such resources directly, but Android is akin to an emulator operating in layers. With the exact same device, Black Survival will perform better on an iOS one. The OS side of software engineering is very difficult – even the people who made the OS themselves deal with bugs and lags. Android firmwares are more colorful in their problems. That’s why sometimes even expensive Android phones have dealing with simple games.

Black Survival does not use much 3D graphics but it does a lot of calculations. It places more strain on the CPU than the GPU. That’s why Black Survival might feel slower compared to its required specs. Speaking of which, devices with poor GPU might be the better device for Black Survival, given right conditions.

Beyond that, sometimes bad weather, old cell tower, mismanaged firmware updates, etc. can be sources of problems. Speaking from experience, something truly random might affect game performance.

We emphasize that minimizing device differences is our responsibility. Compared to when Black Survival first launched, we’ve made significant progress in that regard. However, there are always variables we did not account for and it is extremely difficult to solve for all of them. We will still try regardless.

So far, bugfixes have been highly technical issues and we only talked about their end results in patch notes. In the future, however, we will put an effort to talk about frequent 1:1 feedback in notices and development journals.

Thank you.
31 Ιαν 2018
8-in-1 IQ Scale Bundle - alexgreendev
At the request of users, it was added:
1) Full screen mode for each game.
2) Scaling the game zone in some games.
3) Select the level of the game in some games.
PS: Do not hesitate to create discussions in the community and express your suggestions for improving the game. Your feedback is very important to us.
31 Ιαν 2018
8-in-1 IQ Scale Bundle - alexgreendev
At the request of users, it was added:
1) Full screen mode for each game.
2) Scaling the game zone in some games.
3) Select the level of the game in some games.
PS: Do not hesitate to create discussions in the community and express your suggestions for improving the game. Your feedback is very important to us.
31 Ιαν 2018
Community Announcements - alexgreendev
At the request of users, it was added:
1) Full screen mode for each game.
2) Scaling the game zone in some games.
3) Select the level of the game in some games.
4) DLC.
PS: Do not hesitate to create discussions in the community and express your suggestions for improving the game. Your feedback is very important to us.
31 Ιαν 2018
Alice Mystery Garden - AMG_GAMES
[Update Contents]
  • Corrected the display mirroring.
  • Overall performance has been improved.

31 Ιαν 2018
Alice Mystery Garden - AMG_GAMES
[Update Contents]
  • Corrected the display mirroring.
  • Overall performance has been improved.

31 Ιαν 2018
Community Announcements - alexgreendev
At the request of users, it was added:
1) Full screen mode for each game.
2) Scaling the game zone in some games.
3) Select the level of the game in some games.
4) DLC.
PS: Do not hesitate to create discussions in the community and express your suggestions for improving the game. Your feedback is very important to us.
31 Ιαν 2018
Community Announcements - alexgreendev
At the request of users, it was added:
1) Full screen mode for each game.
2) Scaling the game zone in some games.
3) Select the level of the game in some games.
4) DLC.
PS: Do not hesitate to create discussions in the community and express your suggestions for improving the game. Your feedback is very important to us.
31 Ιαν 2018
Community Announcements - alexgreendev
At the request of users, it was added:
1) Full screen mode for each game.
2) Scaling the game zone in some games.
3) Select the level of the game in some games.
4) DLC.
PS: Do not hesitate to create discussions in the community and express your suggestions for improving the game. Your feedback is very important to us.
...

Αναζήτηση νέων
Αρχείο
ΝΕΑ ΑΝΑ ΕΤΟΣ
2024   2023   2022   2021   2020  
2019   2018   2017   2016   2015  
2014   2013   2012   2011   2010  
2009   2008   2007   2006   2005  
2004   2003   2002