Thanks to rykoszet for the transcript 🙂
What changes did you make to Matchmaking and did it bring the results you expected?
– Andrey Biletskiy: Yes, the goal has been achieved. Impressions at many levels of the game have been significantly improved primarily at levels VII and VIII, where eventually the number of battles has significantly decreased, where these vehicles started at the bottom of the list. This is not the end for us, we have further iterations that we are gradually introducing at even a good pace. We are testing a system that allows us to control situations where the battle system is very strange, which we are never satisfied with. Think about it this way: Matchmaker is fed with a huge amount of random elements at once and what we are trying to teach him is to manipulate time, because it is the only variable that we can play and control. So Matchmaker uses these minimum delays in the selection of players to assemble more predictable teams. This will avoid situations that are theoretically possible, where we have 16 light tanks in the battle.
Or it will give us the ability to control how often specific levels will land in MM +1 or +2. This, of course, does not mean that every player will have mathematically such chances that he will play one such battle, the other such battle, sometimes of course there may be differences. We also want to make sure that we provide players with the most diverse experiences possible. We are also testing the mechanism controlling the effects of battles with three, two one and no artillery and it seems to me that after one more iteration of the test we will be able to say that we have achieved what we wanted. Interestingly, we are testing this MM on official servers, introducing new rules for a short time to get the most organic, real data, not their statistics. Sometimes players write to support that they had a strange battle, or that they waited too long in the battle queue, and that gives us a lot of information. When we’re done with what we’re working on now, we’ll extend work on MM to many situations that he hasn’t considered before.
For example – we have a vehicle tagging system, used even in Ranked Battles. When it happens that MM finds too many alpha tank destroyers in the battle, for example, the queue will be rebuilt by delaying matchmaking for some people in a few seconds and allowing new ones.
In general, work on Matchmaking is like skateboarding blindfolded.
– What Matchmaking improvements are you planning in the near future?
– Andrey Biletskiy: I have already talked about the most important improvements, the MM control mechanism. All the time we have these strange battles from time to time due to the way MM manages the queue. This is where our work is focused. We want teams to be predictable and comfortable for players.
– Are there any plans to redistribute specific types of vehicles, for example light tanks, for battles?
– Andrey Biletskiy: We’re already doing it with artillery! However, before we extend it to other types of vehicles, we need to make sure it works. Step by step, iteration after iteration we will try to introduce more control over what MM does.
– Alexey Ilyin: I would like to add a few words here: our main goal now is not to break up what we already have. At the moment, the queue – especially on the European and yew server – works stable and well due to their large population. We don’t want to spoil anything because the system works well. We have reduced the number of battles with MM +2, where, for example, Tier VIII met Tier X tanks in 40% of the battles, but we were able to reduce this value by half to 20%. We are happy with it. Above all, however, we are working on what Andrey mentioned and we are working on a series of small changes that will allow the new architecture to learn how to deal with specific situations. For example, we are talking about situations where there are a lot of the same vehicles in the queue. We want to solve these situations without dramatically extending battle waiting and MM errors.
– Will there be a +1 balance?
– Andrey Biletskiy: I don’t think so, I wouldn’t advise entering it either. However, our changes are intended to make sure players have a fair distribution of battles. We further believe that MM +2 introduces a more diverse experience, puts the player in a large number of different situations to which he must adapt, build new tactics and strategies.
– Why won’t there be skill-based matchmaking?
– Andrey Biletskiy: As you probably know, people sometimes break out on ranked battles. They blame ranked battles for being super thick, nervous, and so on. And it is Ranked Battles that are the closest implementation of skill-based matchmaking. To be honest: based on the popularity of Ranked Battles, I don’t see the need to introduce something like that.
Again – it’s all about a multitude of experiences, sometimes being the winning side, sometimes losing, so that players can try to turn the tide of battles. Lack of skill MM also gives interesting opportunities to measure the performance of players relative to others. What we observed in the rankings – the winrate of all players was very close to 50%. This way no player will ever feel that he is getting better, he will not feel the fun of developing. Switching to skill matchmaking would force us to copy League of Legends with their league and rank system and I’m not sure if it would be a solution for our game that would maintain this contrasting, varied gameplay.
So for now – even on the front line, we confront players based on their experience in this mode, but although maybe in the future we will introduce some skill MM elements into different modes, the randoms should remain random.
– Alexey Ilyin: The same question also often appears on the CIS server, it seems to me that players would like more intense battles, but not everyone could handle it. Ranked battles have shown us that it’s not everyday gameplay.
– We had a large balance of various vehicles, such as the FV4005 or Type 5 Heavy. How are they doing after the changes? Have you noticed a decrease in players’ interest in exploring these lines?
– Alexey Ilyin: Let’s start with Type 5: We tried not to remove this tank from the game, but to reduce the number of HE shells. Now the popularity of both sections is quite equal. We are happy with this and the tank remains popular, although its number in battles has decreased. The performance of both guns is quite similar. That’s good because we want to give people choice and use for both of them, although in different roles.
Speaking of FV: The level of popularity has not changed at all. The nerf wasn’t that big and what we wanted to achieve was a reduction in the survivability of the vehicle, not its firepower. This was achieved, his injuries are still very large, but survivability has decreased. We are also happy with the rebalance, the vehicles are still popular, but not so toxic and too effective.
– Will you change vehicle characteristics this year? If so, which ones?
– Andrey Biletskiy: We’re testing missile rebalance. This is a huge initiative, the biggest balance challenge in years. The whole balance focus is on this. We are pleased with these changes and tests in the sandbox. People also seem to like it more and more. At the moment, as long as we believe that we are in the right place with changes, we will not introduce small changes anywhere. Slava Makarov, my boss, once shared a weak joke like “finish cutting the tail of a cat piece by piece,” so we try to do it with one cut. If in one of the next sandboxes we see a deterioration in the perception of players on these changes and decide to stop them, then we have plans to return to the old model. But for now we want to introduce many small changes in a large package.
– Alexey Ilyin: I would like to talk about specific vehicles, because at the previous WG Fest we announced a change in the tops of five great lines and we have already done the Kranvagen, STB-1 and Leopard 1 rebalance, which we are happy with, and the reason we are waiting with IS-4 and The E-100 relies on changes in missiles that will affect them much more than the three already changed. Also first we want to change it and then see what the change in their performance will be and adjust them later. We remember them, we know that we should focus on them, but we will do it only after changes in ammunition.
– When is the next sandbox and what will we test?
– Alexey Ilyin: The next sandbox will be in October, we are trying to find the right date, because there will be a lot of events. We don’t want to disturb them. The next test will focus on HE shells. We have already changed the ammunition and pool of tanks of HP, the changes in them will be small, but we will add fragments and the mechanics of dealing damage.
– Andrey Biletskiy: I don’t think we are ready to provide details, because they are not yet agreed, but the main thought is that now frags are completely unpredictable. How many times it happens that even very large calibers work against unarmored vehicles, they do not deal enough damage. The goal here is to increase the predictability of these missiles and fulfill their role – they are to finish the target when there is no time to aim, defend the base or deal damage to tanks that we are unable to move otherwise. We are working on the HE missile damage module, we have new instruments for it for various tanks. It will also allow us to differentiate the system of normal tank missiles from artillery. We will adjust these components so that the missiles achieve the intended effects on impact.
– Alexey Ilyin: A few more words: Players, take part in these tests. This is important because we can analyze your opinions and get key data from it. The test will be in October, everyone will be able to take part in it, just download the client. Of course, the right article with details will appear in due time, so please help us make the game better.
– Andrey Biletskiy: Particularly worth noting is the fact that this time there will be no invitation system as in previous iterations. Simply put – if you want to take part, take part. The more opinions and data from your gameplay we get from you, the better we will understand where we are going and how to make it better.
– Why are British light tanks so weak? Do you plan to improve them?
– Alexey Ilyin: I’ll start from another side: This year we released several vehicles and we didn’t want to repeat from previous lines and clone previous vehicles. We want something new, fill the options in the game where there have been loopholes so far. For example, when releasing Swedes – a new suspension was a curiosity. The same with wheeled or Swedish medium tanks. In British light tanks everything is based on a passive spot, because this sphere may not have been empty, but …
– Andrey Biletskiy: Not full enough. What Alexey talks about are different styles of play. Swedish tank destroyers allow an interesting approach to the verticality of maps. Swedish heavy tanks are a different approach to magazines, in a different role. The taxi drivers are ultimately active scouts, as we saw in the 3vs3 tournament a moment ago. Yes, the wheelers were created so that they could not be easily stopped, encouraging active gameplay.
In turn, British light tanks are for this group of players – whom we know because we have statistics – who prefer passive meeting. We don’t want to create one line for everyone, because it would be a failure for us. We don’t want players to say “I will stop playing my style because everyone is playing it”. Interestingly, we have different wheeled vehicle pickups on different servers. Some say they are OP, while others are much less. The same with British light tanks. We have opinions from the world that they are OP. It’s very subjective.
– Alexey Ilyin: I will be very boring talking about numbers, but they are quite popular. Players who can and want to play them play them. There are quite a lot of them, and the efficiency is not bad at all, because they are at the level of the best light tanks. The number of Manticore spots is similar to the AMX 13 105, T-100 LT and wheeled vehicles. They are quite good, not for everyone, but they manage in a passive spot.
– Why did you cancel the 430U rebalance?
– Andrey Biletskiy: Probably thanks to your reaction to the news .
– Max Chuvalov: We had a big debate in which direction we should go and when we noticed the reaction of the community we realized that it would not work. We immediately undo these changes, leaving only STB and Leopard.
– Andrey Biletskiy: This does not mean that we will never go back to this topic, but in the upcoming major changes we will definitely see how it turns out in practice.
– Can you tell a little more about the new mode?
– Andrey Biletskiy: It’s a Battle Royale style mode, if you remember we had it for April Fool’s Day, it was a very, very early prototype that almost leaked. I mean, we made fun of this leak, but it was a really early prototype. We wanted to see the reaction of recipients to unusual mechanics. We want to introduce an environment where players can adapt their playing style dynamically during the battle. We will also have many asymmetrical duels with each other. Map control, maintenance or navigation will be completely different. The meeting system, which makes every move think twice, will be completely changed and more dangerous. And the rules of the battlefield – the map is getting smaller, you have to keep up with the changes in the battlefield. Everything is based on territory control and collecting equipment dropped by planes and found in different places on the map. Of course, you can also grab the wrecks of opponents.
Tanks will evolve. Everyone will start at a similar level, but different people like different styles of play, so we will allow experience and change tanks as you like in the progression of the battle. Maybe a smaller, harder to see vehicle, or maybe a larger, better armored vehicle? You can evolve vehicles your way. The mode will be available solo and in three-man platoons and will face the same number of groups. The mode has already been checked at the top of developers in Minsk and it seems to me that we have prepared something really fun.
– Something more about the mechanics of this mode?
– Andrey Biletskiy: First of all, we wanted to keep this feeling of discovering the unknown, but the mechanics have changed a lot since April Fool’s Day 2018. Every time we release new modes we want to show something new and look for conclusions on how to use it or what not to do anymore. . There will be various new consumables that may also go to the front line in the future. If we see a positive reception, we are serious about making this a seasonal event, like a front line. This will give us a completely new approach to competition in the game, but avoiding the problems that are associated with competition modes.
It can also be much more interesting to watch for viewers, even on Youtube, because it requires less specific knowledge and is more affordable. We have made many changes based on our experience, even with the respawn system. We’ve also added a radar that will let you find yourself more or less in a situation. We’ve changed many things in the consumables, we’ve introduced the evolution of the tank to reduce the randomness of what you can and cannot do in battle, so it’s a very big test for us. I hope that people will enjoy it not only because it is Battle Royale, but also a different approach to the game of World of Tanks that divides the basics and part of the mechanics. You can still play armor, you can still use the terrain to gain an advantage, all the time you have to predict your opponent’s moves to succeed. At the same time, however, you can adapt to what others are doing and to what is.
– Why did you turn off friendly fire?
– Andrey Biletskiy: It’s a very complicated question.
– Alexey Ilyin: Because we can!
– Andrey Biletskiy: You know, from the perspective of how good we give communication opportunities to players – basically it’s “bad”, but it will improve – it turned out that damage to the ally has become one of the methods of transmission “hey, man, I do not like What you’re doing”. However, this negatively affected the whole team, especially when such a signal was sent by a guy from a 150mm caliber gun … Secondly, we got a lot of complaints that the players behave strangely and ruined all the fun.
So we began to wonder what friendly fire gives us. Of course, without discussion, this is realistic, but is it useful in the game? Maybe we should give players other communication options? And we started to do it in different modes, turning off allied fire just to see how it works. Maybe people would be dissatisfied with that, or it would have a big impact on the game? It turned out not, and people were quite happy about it. Of course, there are still problems with abuse, such as throwing off cliffs or drowning, but it is much more difficult and requires more effort and maybe – just maybe – people in troll mode will think “not worth it” and let go.
– Alexey Ilyin: One of the main reasons was that players who shot at allies didn’t want to. It was accidents that upset both sides. Players who accidentally fired at allies had to suffer excessive consequences of unintended actions. If nobody wanted to, why not just turn it off?
– Last question: Can you share the effects of the last event – “Homefront”?
– Max Chuvalov: We tested several mechanics and systems. First, we checked artificial intelligence and how PVE mode could work in the game. To remind you – these were several waves of opponents that had to be beaten in five. Bots are nice. Very good. It would seem that sometimes they even rotate the hull when you shoot them. AI is of course the most important element of every PVE mode. We are pleased that we were able to create such an AI.
Secondly – the narrative. We’ve tested missions, maybe it can be developed more. We checked the superplutons – five people. And the biggest information for us – players like PVE and want to spend a lot of time there, so we had to limit their access to it with fuel. Many players used all the fuel they got. These are important tips and green light for the concept.
– Andrey Biletskiy: This is our first approach to PVE, our crew is not used to it. We are learning, but I hope that we will soon announce further steps in this direction and our success will continue. Who knows what we’ll get? Maybe with time you can tell a story!