We are constantly on a quest for improvements in the Matchmaking system. In this regard, as an experiment, today we will increase the speed of Matchmaking for one week (as an experiment) to implement a more rapid expansion in the range of acceptable levels. This will increase the search speed of players in battle, but also it will increase the number of players with equipment above or below your level. These changes will affect only the PvP game mode. We will monitor the results of this experiment and draw conclusions about further steps in this direction. We invite you to actively participate in testing, please also give your impressions on the game forum.

Discuss on Forum

We are constantly on a quest for improvements in the Matchmaking system. In this regard, as an experiment, today we will increase the speed of Matchmaking for one week (as an experiment) to implement a more rapid expansion in the range of acceptable levels. This will increase the search speed of players in battle, but also it will increase the number of players with equipment above or below your level. These changes will affect only the PvP game mode. We will monitor the results of this experiment and draw conclusions about further steps in this direction. We invite you to actively participate in testing, please also give your impressions on the game forum.

Discuss on Forum

Hello everyone!

Recently we have received a lot of requests to explain how the calculation of rating works in ranked matches. For ranked matches in Survarium we opted for the ELO rating system, which has been successfully used to determine the strength of players in chess, and it also has a number of successful implementations in other games. In this article we will explain the principle of ELO rating, and also tell you about some features of the ELO implementation in Survarium.

What is the ELO rating
ELO – numerically expressed feature of player power; based on the ratings of the two players it is possible to predict who will emerge victorious in the internal opposition and the score he wins.

Before combat system, based on the ELO rating of a player predicts its performance in battle, after the battle it compares the predicted result with the actual, after this, the assessment of the player is changed based on the difference between predicted and actual outcome.

The rules for computing ELO rating in the game.
As a criterion of assessment of the level of the player in each match, we will use the personal score of useful actions. Since the score of a particular player depends heavily on the nature and timing of the match, then we will use the relative score of the player – the percentage of total score of all players scored by this player. First, let’s consider a simple situation – duel of two players.

Evaluation of the percentage of the player's score against a particular player is carried out according to the following formula:
ESTa= 1 / (1 + 10^((Rb – Ra)/dif_factor))
Where:
ESTa – the expected number of points earned in the range from 0 to 1.
Ra – ELO rating of player A
Rb – ELO rating of player B
dif_factor is a constant that determines at what difference in ELO rating the probability of winning for one player is ten times higher than the other. At the moment it is 400.

For example, player A (rating 1100), and player B (rating 1350). The difference of 250 is in favor of player B, so in personal confrontation player A needs to score:
ESTa= 1/(1+10^(250/400)) = 1 / (1+4.22) = 0.19 or 19 % of the total number of points
Estb stock= 1/(1+10^(-250/400))= 1/ (1 + 0.24) = 0.81 or 81% of the total number of points

By the end of the match there is a comparison of prognosis of points scored with real-time indicator. After that there is a change in the ELO rating, depending on the difference between prognosis and indicators.

The ELO rating is determined by the formula:

Ra' = Ra + K_factor * (Sa - Ea)
Where:
Ra’ - new rating of a player
Ra – old rating of a player
Sa – real score of a player during a match
Ea – projections of a player during a match
K_factor – value, regulating the rate of rating change. It depends on the number ofmatches played by the player:
100 - for qualifiers
50 – for players who have played less than 10 games since the start of the season.
25 – for players who have played more than 10 games since the start of the season.

For example, in the above case, player A has played very well and scored 30% of the total number of points (though lost the fight). Player B respectively scored 70% of the points. For example, this was the second match for player A (K_factor = 50), and for player B It’s 11th match (K_factor = 25)
Ra’ =1 100 + 50 * (0.3-0.19) = 1105,5
Rb’ =1 350 + 25 * (0.7-0.8) = 1347,5

But, unlike chess, our matches are played 8x8 survivors. For this case we use a more sophisticated algorithm to predict the performance of players. We proceed from the assumption that each match is something like a tournament in which each player of one team meets every player of the other team equal number of times (for simplicity of calculations we consider this to be 1 time). Thus we can predict the result of the player as an average result of a duel with each player of the opposing team. After the match we calculate a real score for each player and then change his ELO rating.


More information is available here.

Hello everyone!

Recently we have received a lot of requests to explain how the calculation of rating works in ranked matches. For ranked matches in Survarium we opted for the ELO rating system, which has been successfully used to determine the strength of players in chess, and it also has a number of successful implementations in other games. In this article we will explain the principle of ELO rating, and also tell you about some features of the ELO implementation in Survarium.

What is the ELO rating
ELO – numerically expressed feature of player power; based on the ratings of the two players it is possible to predict who will emerge victorious in the internal opposition and the score he wins.

Before combat system, based on the ELO rating of a player predicts its performance in battle, after the battle it compares the predicted result with the actual, after this, the assessment of the player is changed based on the difference between predicted and actual outcome.

The rules for computing ELO rating in the game.
As a criterion of assessment of the level of the player in each match, we will use the personal score of useful actions. Since the score of a particular player depends heavily on the nature and timing of the match, then we will use the relative score of the player – the percentage of total score of all players scored by this player. First, let’s consider a simple situation – duel of two players.

Evaluation of the percentage of the player's score against a particular player is carried out according to the following formula:
ESTa= 1 / (1 + 10^((Rb – Ra)/dif_factor))
Where:
ESTa – the expected number of points earned in the range from 0 to 1.
Ra – ELO rating of player A
Rb – ELO rating of player B
dif_factor is a constant that determines at what difference in ELO rating the probability of winning for one player is ten times higher than the other. At the moment it is 400.

For example, player A (rating 1100), and player B (rating 1350). The difference of 250 is in favor of player B, so in personal confrontation player A needs to score:
ESTa= 1/(1+10^(250/400)) = 1 / (1+4.22) = 0.19 or 19 % of the total number of points
Estb stock= 1/(1+10^(-250/400))= 1/ (1 + 0.24) = 0.81 or 81% of the total number of points

By the end of the match there is a comparison of prognosis of points scored with real-time indicator. After that there is a change in the ELO rating, depending on the difference between prognosis and indicators.

The ELO rating is determined by the formula:

Ra' = Ra + K_factor * (Sa - Ea)
Where:
Ra’ - new rating of a player
Ra – old rating of a player
Sa – real score of a player during a match
Ea – projections of a player during a match
K_factor – value, regulating the rate of rating change. It depends on the number ofmatches played by the player:
100 - for qualifiers
50 – for players who have played less than 10 games since the start of the season.
25 – for players who have played more than 10 games since the start of the season.

For example, in the above case, player A has played very well and scored 30% of the total number of points (though lost the fight). Player B respectively scored 70% of the points. For example, this was the second match for player A (K_factor = 50), and for player B It’s 11th match (K_factor = 25)
Ra’ =1 100 + 50 * (0.3-0.19) = 1105,5
Rb’ =1 350 + 25 * (0.7-0.8) = 1347,5

But, unlike chess, our matches are played 8x8 survivors. For this case we use a more sophisticated algorithm to predict the performance of players. We proceed from the assumption that each match is something like a tournament in which each player of one team meets every player of the other team equal number of times (for simplicity of calculations we consider this to be 1 time). Thus we can predict the result of the player as an average result of a duel with each player of the opposing team. After the match we calculate a real score for each player and then change his ELO rating.


More information is available here.

Q: What is the difference between anomalous protection and gas tanks? Could you give some examples where each is best used?
A: As opposed to anomalous protection, the gas tanks give you 100% protection against anomalies affecting your breath/face, such as larkspur and sponge, but acting for a limited time only.

Q: Why did you decide to make a second Smallshot ammo type for 0.28 PTS? Was its purpose mainly to increase maintenance costs for the high-level shotguns (Saiga/Vepr-12)?
A: This was made for the balance purposes - we needed a more powerful ammo to balance out those shotguns against other weapons of their level.

Q: Are you still planning on implementing detectors for collecting artifacts? How do you see this working with the new random anomalies?
A: Please, remember Survarium is not S.T.A.L.K.E.R., it's a different game) From game designer's point of view such detectors are not needed for the random anomalies, so no plans for that at the moment.

Q: What percentage protection does the Sponge artifact give, and does it cover all types of anomalies?
A: It does cover all types of anomalies. As for the protection percentage, we are still adjusting it.

Q: Why does the pay-for-premium VSS have a 10 more round capacity than the free version? Why was this decision made? What other guns will have such treatment?
A: That was done for the balancing purpose. We needed to balance out premium VSS as a level 9 gun.

Q: Will the 24 hour delay for switching regions be dropped to an amount that lets us play on the populated regions as population changes during the day?
A: The regions system will be changed with the next game update; one of the changes - a possibility to freely switch between the game servers. More details to be shared later.

Q: Can you say how big the new London location will be in comparison to existing ones? Will it be small like "Tarakanovsky Fort" or large like "Mamayev Kurgan"?
A: The map is quite spacious. Soon you'll be able to judge for yourself 

Q: Is there any chance that we'll see the "London" map in 0.29?
A: yes

Q: What happened to the idea of scanners, radars and other electronics mentioned in the Dev Diary #5 video*? Are there plans to add such gadgets or other deployable items in the near future?
A: That's still on our plan list, yet to be implemented.

Q: In 'Dev Diary Part 5: Story' you mentioned that a bit of story will be implemented in PvP. When will we be able to see the first signs of something related to the main story in PvP?
A: We don't have a specific ETA for you at the moment. We'll make sure to announce it once the details are available.

Q: What is the difference between anomalous protection and gas tanks? Could you give some examples where each is best used?
A: As opposed to anomalous protection, the gas tanks give you 100% protection against anomalies affecting your breath/face, such as larkspur and sponge, but acting for a limited time only.

Q: Why did you decide to make a second Smallshot ammo type for 0.28 PTS? Was its purpose mainly to increase maintenance costs for the high-level shotguns (Saiga/Vepr-12)?
A: This was made for the balance purposes - we needed a more powerful ammo to balance out those shotguns against other weapons of their level.

Q: Are you still planning on implementing detectors for collecting artifacts? How do you see this working with the new random anomalies?
A: Please, remember Survarium is not S.T.A.L.K.E.R., it's a different game) From game designer's point of view such detectors are not needed for the random anomalies, so no plans for that at the moment.

Q: What percentage protection does the Sponge artifact give, and does it cover all types of anomalies?
A: It does cover all types of anomalies. As for the protection percentage, we are still adjusting it.

Q: Why does the pay-for-premium VSS have a 10 more round capacity than the free version? Why was this decision made? What other guns will have such treatment?
A: That was done for the balancing purpose. We needed to balance out premium VSS as a level 9 gun.

Q: Will the 24 hour delay for switching regions be dropped to an amount that lets us play on the populated regions as population changes during the day?
A: The regions system will be changed with the next game update; one of the changes - a possibility to freely switch between the game servers. More details to be shared later.

Q: Can you say how big the new London location will be in comparison to existing ones? Will it be small like "Tarakanovsky Fort" or large like "Mamayev Kurgan"?
A: The map is quite spacious. Soon you'll be able to judge for yourself 

Q: Is there any chance that we'll see the "London" map in 0.29?
A: yes

Q: What happened to the idea of scanners, radars and other electronics mentioned in the Dev Diary #5 video*? Are there plans to add such gadgets or other deployable items in the near future?
A: That's still on our plan list, yet to be implemented.

Q: In 'Dev Diary Part 5: Story' you mentioned that a bit of story will be implemented in PvP. When will we be able to see the first signs of something related to the main story in PvP?
A: We don't have a specific ETA for you at the moment. We'll make sure to announce it once the details are available.

Dear Survivors,

We've had many players asking us what our plans are to sort out the imbalance that is currently found in the game with regards to new players in sweaters running around in high-tier or high-league level matches. We would like to apologise for this temporary situation and are currently looking into it and plan to rebalance and amend the situation with the forthcoming game version V0.29.

Again, we apologise for this and thank you all for your invaluable feedback. Together we will make this game better and better!

See you on the battlefield!

Dear Survivors,

We've had many players asking us what our plans are to sort out the imbalance that is currently found in the game with regards to new players in sweaters running around in high-tier or high-league level matches. We would like to apologise for this temporary situation and are currently looking into it and plan to rebalance and amend the situation with the forthcoming game version V0.29.

Again, we apologise for this and thank you all for your invaluable feedback. Together we will make this game better and better!

See you on the battlefield!

Dear Survivors,

In recent weeks there has been an outcry by the community regarding the regional servers and our decision to merge the EU and NA servers to increase the speed of matchmaking. This had a two-fold effect; players were matched at a faster rate but at the sacrifice of NA servers and thus many players revolted and have chosen not to play until such time as we separate all regions once again.

So, here’s the current situation. For the remainder of Version 28 the servers will remain as they are. Come update V0.29 there will be an in-game server browser where you, the player, will be able to view various available servers in real-time and select the one you wish to play on. This will include NA, EU and CIS along with any other regions that may launch with 29. We are hoping to release the next version sometime in June. So, you won’t have to wait too much longer.

We are sincerely and deeply sorry for the trouble caused and hope you can be patience a little while longer.

See you on the battlefield.

Dear Survivors,

In recent weeks there has been an outcry by the community regarding the regional servers and our decision to merge the EU and NA servers to increase the speed of matchmaking. This had a two-fold effect; players were matched at a faster rate but at the sacrifice of NA servers and thus many players revolted and have chosen not to play until such time as we separate all regions once again.

So, here’s the current situation. For the remainder of Version 28 the servers will remain as they are. Come update V0.29 there will be an in-game server browser where you, the player, will be able to view various available servers in real-time and select the one you wish to play on. This will include NA, EU and CIS along with any other regions that may launch with 29. We are hoping to release the next version sometime in June. So, you won’t have to wait too much longer.

We are sincerely and deeply sorry for the trouble caused and hope you can be patience a little while longer.

See you on the battlefield.