20 replies to this topic

[Vanize]

We recently(ish) got this dev blog explaining the match maker algorithm: http://blog.worldofw...asses-and-bots/

 

here are the "Basics"

 

Block Quote

Basics of filling battle rooms

1. Battle rooms are created starting from higher tiers and moving down.
2. The matchmaker forecasts how many rooms it can create at the specific battle tier, based on the number of suitable aircraft in the queue.
3. The team to be formed first is picked randomly.
4. All flights waiting in the queue are spread evenly among the forecasted battle rooms. At this point, the matchmaker has multiple underpopulated rooms that it starts filling with more players.
5. These rooms are filled with single aircraft from the queue. At the same time, teams are balanced by classes and tiers to eliminate disparity in team potentials. If necessary, the matchmaker picks additional aircraft from the queue and discards ones it already added to the room.

 

This algorithm runs every server tick, which means that every time the queue updates, the matchmaker checks it and, if possible, starts creating and filling new rooms. At the same time, it continues to balance teams in the rooms it created previously.

 

The real issue MM has is contained in that last sentence. It creates rooms, populates them with players, then allows those rooms to persist with the players waiting inside and not useable for other (potentially better) battle line-up solutions. It spends its time trying to populate rooms that already exist and then creating new rooms for things that don't fit the solution for a pre-existing room. This means it is SEVERELY limiting it's balancing options and distributing players among more rooms than the minimum ultimately required, as it would have far more variables to work with if players were not already committed to rooms as new players drop in.

 

If I were writing this algorithm, no actual rooms would be created until just before the battle drop, leaving me with maximum player base to create a battle line-up from, instead of already having them distributed to various rooms and waiting for new appropriate players to drop in.

 

IF MM simply held all players outside of rooms and recalculated theoretical battle line-up solutions regularly without actually making rooms and committing players to them, it could much more easily find a good solution and drop that entire set of players all at once (along with any bots required) into a room created at that moment for that final solution to the line-up and send them into battle straight away, presumably with some leftovers left in MM who would become the priority for the next solution.

 

The way MM is designed now basically assures a fragmented player base with lots of bots in each battle, even when there are sufficient players online for full battles. The inefficient manner the current MM uses wouldn't matter with a sufficiently large player base, but we do not have that luxury.

 

Edited by Vanize, 15 February 2017 - 11:41 PM.

[losttwo]

there are 10 different tiers or another term could be 10 different rooms.

Rather then the 7 rooms of pre 1.5.

In other words 10 rooms with a possible tier spread +/- 1 as apposed to

7 rooms with a +/- 2 spread.

 

How many rooms is MM creating and at what time interval is an unknown factor.

Therefore i am using the most basic of what is known,.

 

Let's say the population is running at 500.

Split that population by rooms which will give you about 50 players per room.

Provided of course that you have equal number of people choosing to play each tier and pressing the "BATTLE"

button at the same time.

Battles are 15 minutes which leaves 20 people waiting 15 minutes for a battle.

Now we need an extra room for those people so they dont have to wait.

Now we have 10 extra rooms for a total of 20 rooms.

 

Problem with the entire idea is that people do not press battle at the same time and they don't all fly similar tiers.

 

Now, Wargaming has hired many programmers and designers and have a team established to work on these idea's.

I am sure they have already ran through any IDEA or scenario we could come up with.

More so after, what has it been, 5 years. So to think that Match Maker is flawed is nothing more than an opinion

of something not working the way you think it should work.

 

My suggestion to you would be to write a program for match maker and submit it to Wargaming as a SOLUTION

to your perceived problem.

 

Otherwise my verbose drivel just matches your verbose drivel.

 

[MelBrooks]

So, things you don't understand are verbose drivel....agreed to what Vanize said, they took the backdoor approach on the algorithm by creating a new room when current players do not fit the parameters. Creates larger battles for sure, also fills in with a lot more bots. WG responded to the larger battle platform needed for sure but definitely at a price.

[GouIdy]

What people seem to be forgetting is that we have the same match maker as the other two servers & as far as WG is concerned the MM is working as intended. EU & RU server populations are 2 1/2 & 5 1/2 times that of NA  & obviously the MM is working better for them. At this point I really can't see WG giving NA any special treatment & only when MM becomes an issue on RU they might consider tweaking it. If & when that time comes I have no faith that it will actually be a change for the better as I'm of the opinion that the MM right now is probably as good as they are capable of getting it taking into consideration WG track record.

Edited by GouIdy, 16 February 2017 - 05:10 AM.

[cddlbunny0]

GouIdy, on 15 February 2017 - 11:13 PM, said:

What people seem to be forgetting is that we have the same match maker as the other two servers & as far as WG is concerned the MM is working as intended. EU & RU server populations are 2 1/2 & 5 1/2 times that of NA  & obviously the MM is working better for them. At this point I really can't see WG giving NA any special treatment & only when MM becomes an issue on RU they might consider tweaking it. If & when that time comes I have no faith that it will actually be a change for the better as I'm of the opinion that the MM right now is probably as good as they are capable of getting it taking into consideration WG track record.

 

Agreed. It seems that the NA population is more interested in World of Tanks and Warships than our little population. If one recalls correctly, with the amount of hype that WG put into their "showcase" game (Warships), the staff were transferred from this project so they could meet their deadline and our world became stagnant from that point on. Like both Victor and Serb said "this game has been a disappointment and has not lived up to expectations". In other words, there was not enough interest in North America. Remember the historical battles in 2015 in WoT. Everyone in the NA server wanted only to play King Tigers and JSII. It just seems that in the Americas people are not interested in their history the way the Europeans are, or should one say, the extent. That opinion is formatted from living on both sides of the pond.

Edited by cddlbunny0, 16 February 2017 - 11:11 AM.

[losttwo]

MelBrooks, on 15 February 2017 - 09:29 PM, said:

So, things you don't understand are verbose drivel....agreed to what Vanize said, they took the backdoor approach on the algorithm by creating a new room when current players do not fit the parameters. Creates larger battles for sure, also fills in with a lot more bots. WG responded to the larger battle platform needed for sure but definitely at a price.

Drivel...DRIVEL...of course it is all drivel...Someone doesn't even work for the company and wants to spout off whats wrong.

As if the company doesn't know what it is doing.

You know like the new kid that gets hired into a company, complains about his job and keeps spouting off about how he can run it better.

 

GouIdy, on 15 February 2017 - 11:13 PM, said:

What people seem to be forgetting is that we have the same match maker as the other two servers & as far as WG is concerned the MM is working as intended. EU & RU server populations are 2 1/2 & 5 1/2 times that of NA  & obviously the MM is working better for them. At this point I really can't see WG giving NA any special treatment & only when MM becomes an issue on RU they might consider tweaking it. If & when that time comes I have no faith that it will actually be a change for the better as I'm of the opinion that the MM right now is probably as good as they are capable of getting it taking into consideration WG track record.

 

I agree with you about " this point in time " I do not think match maker can get any better than it already is, given the circumstances.

As much as i would like to see a return to a +/- 2 spread. I do not think aircraft balance and match maker can handle it.

Although it will kick one out on rare ozzassion.

[cddlbunny0]

I think that you meant occassions, Lost. Yes if the server population were in line with the numbers on the other servers then a +/-2 would be feasible, but for now let us be happy with what we have.

[Topsight]

Dev Quote from the Blog and you have to be logged in to read this response to what I had written

 

"The matchmaker indeed fills the teams with bots of different skill levels. For example, on low tiers there are only newbie bots, on mediums there might be 1-2 newbies, most medium and 1-2 highly skilled bots. At high tier bots are medium and highly skilled. The matchmaker does not pair bots based on skillsets though, so one team might have an ace heavy fighter bot paired to a similar fighter or attack aircraft."

 

I am of the personal opinion that the more skilled you become the tougher the battles can become. On the surface that seems palatable to a point. Bottom line is population numbers I believe keep the game alive. I did give what I felt is a solution in their blog. They do not factor a flight of 2 to have an advantage and only factor them as single players. if they are pushing people to flight-up, I see a future problem with that if they balance bot skill sets and am of the opinion that those that play, at least on the NA server, fly mostly solo. It is as if the Developers are trying to push people to flight-up and from what I have seen playing on and off every day from T-6 and up 99% of the time I see a flight of 2 rarely. Not saying the Developers should see this as a problem at this time.

Edited by Topsight, 16 February 2017 - 03:28 PM.

[Psicko23]

losttwo, on 16 February 2017 - 02:18 AM, said:

Drivel...DRIVEL...of course it is all drivel...Someone doesn't even work for the company and wants to spout off whats wrong.

As if the company doesn't know what it is doing.

You know like the new kid that gets hired into a company, complains about his job and keeps spouting off about how he can run it better.

 

 

I agree with you about " this point in time " I do not think match maker can get any better than it already is, given the circumstances.

As much as i would like to see a return to a +/- 2 spread. I do not think aircraft balance and match maker can handle it.

Although it will kick one out on rare ozzassion.

Yes, Vanize doesn't work with wargaming, but his solution would be much better. Remember in 1.6 and 1.7 when there would be about 30 plus people in the combined tiers below you to above you, only to drop into a match that was 1v1 or 3v3, when it could have easily been at least a 10vs10. The company doesn't really seem to know what it's doing when it comes to retention for the NA server. Although I would like to see more there tier battles as there would be more humans and less bots.

 

cddlbunny0, on 16 February 2017 - 03:14 AM, said:

I think that you meant occassions, Lost. Yes if the server population were in line with the numbers on the other servers then a +/-2 would be feasible, but for now let us be happy with what we have.

Three tiered battles would only help us. We would have bigger games with more humans and less bots per match. 

[BrushWolf]

GouIdy, on 15 February 2017 - 10:13 PM, said:

What people seem to be forgetting is that we have the same match maker as the other two servers & as far as WG is concerned the MM is working as intended. EU & RU server populations are 2 1/2 & 5 1/2 times that of NA  & obviously the MM is working better for them. At this point I really can't see WG giving NA any special treatment & only when MM becomes an issue on RU they might consider tweaking it. If & when that time comes I have no faith that it will actually be a change for the better as I'm of the opinion that the MM right now is probably as good as they are capable of getting it taking into consideration WG track record.

 

While it started life from the WoT MM there have been so many changes, many of them bad, that this is not the same MM and in fact it is hard to even tell it comes from the same code.

[GouIdy]

BrushWolf, on 16 February 2017 - 12:36 PM, said:

 

While it started life from the WoT MM there have been so many changes, many of them bad, that this is not the same MM and in fact it is hard to even tell it comes from the same code.

 

Can you please elaborate on the point you are trying to make relating to my post as there was no mention of WOT or any comparison of the two games.

[BrushWolf]

GouIdy, on 16 February 2017 - 11:50 AM, said:

 

Can you please elaborate on the point you are trying to make relating to my post as there was no mention of WOT or any comparison of the two games.

 

WoT is the original game that the base MM was used with and still mostly does although the did start protecting the seals in the bottom tiers from the seal clubbers. That unfortunately sucks when someone has taken a long break and wants something easy to get back into shape. What I was trying to say is the changes that Persha has done to it has made it a completely different MM.

[losttwo]

Here, let me explain it in different way.

The Match Maker is not flawed the reason why is simple.

WG is thinking on a global scale rather than regional population scale.

It is an International company and will think like an International company.

" Build it, they will come " has been a major idealism for many an international company.

 

What makes you think it is flawed is your line of thinking and not liking the way it currently operates.

While you sit and perceive a problem with the current system because it doesn't operate the way you think

it should operate. You point out your perception but do not write out a program and offer a solution.

 

Sort of like, sitting on the side of the road with a flat tire.

There is a solution. rather than complaining and pointing out the flat tire you change it.

 

The problem with most people today is that when they have a flat tire they call AAA to change it for them.

At the same time they complain about how long it took AAA to get there and  try to tell the AAA mechanic how to do their job.

 

[BrushWolf]

losttwo, on 17 February 2017 - 04:27 AM, said:

Here, let me explain it in different way.

The Match Maker is not flawed the reason why is simple.

WG is thinking on a global scale rather than regional population scale.

It is an International company and will think like an International company.

" Build it, they will come " has been a major idealism for many an international company.

 

What makes you think it is flawed is your line of thinking and not liking the way it currently operates.

While you sit and perceive a problem with the current system because it doesn't operate the way you think

it should operate. You point out your perception but do not write out a program and offer a solution.

 

Sort of like, sitting on the side of the road with a flat tire.

There is a solution. rather than complaining and pointing out the flat tire you change it.

 

The problem with most people today is that when they have a flat tire they call AAA to change it for them.

At the same time they complain about how long it took AAA to get there and  try to tell the AAA mechanic how to do their job.

 

 

True but the MM could do a better job getting humans into the matches. A 20 - 30 seconds delay would probably make a huge difference.

[losttwo]

BrushWolf, on 17 February 2017 - 10:30 AM, said:

 

True but the MM could do a better job getting humans into the matches. A 20 - 30 seconds delay would probably make a huge difference.

 

My opinion is that it would not make a difference.

The problem is WHEN a person presses the battle button.

My self and I can only speak for myself. I do not exit a battle, switch planes and click battle.

Nor do I always click battle after my plane returns to hangar.

 

There are many factors that contribute to when the battle button is clicked.

Some of those factors are : Hangar load times ( mine is about 1 minute ), Internet, tier and plane.

 

I tired to play my XF-90 yesterday around 1PM EST and waited 5 minutes to get a battle.

Switched to tier 9 and then I had to click exit battle after 3 minutes of waiting to force a match.

Dropped down to tier 8 and the longest wait was 1:40  then logged out.

 

Was there a population in those tiers ( 9 and 10 ) ?  I do not know, the battles I was placed in showed no signs of life.

Yet tier 8 was populated with life but for me the queue time was a bit too long.

 

My normal operation is to exit the queue after 2 minutes and switch tiers or planes but I wanted the 5 days premium for the XF-90.

The question now remains " when is the next time  will be able to play the XF-90 in under 2 minutes ? "

 

Even if the queue time is extended another 30 seconds you could still end up with a longer wait than 2 minutes.

It would possibly increase the queue times in the populated tiers ( 3-7 ) and you still end up with a bot heavy match.

 

Like I said it is just my opinion.

 

This is a great topic and It would be nice if WARGAMING EMPLOYEES could chime in on the topic.

 

 

Edited by losttwo, 17 February 2017 - 04:11 PM.

[BrushWolf]

Lost, the top tiers have always been flaky outside of prime time. Personally I think most of the problems with the MM are from all the crap they have added to it. one example is its being psychotically focused on one & two tier matches to the point that people actually post on the forum when they do get a three tier match which with out population should be the norm.

[raulv04]

After glancing and scanning this thread, I'm reminded of the infamous 1.5 MM, where we had to wait up to 45-50 minutes for a battle or more, because the MM kept "balancing" tiers, aircraft and players. In 1.6, Gunlion came up with a solution for the Americas server that worked pretty well given the population of the server. So this means they updated the MM to pre-1.6?

[losttwo]

raulv04, on 17 February 2017 - 12:46 PM, said:

After glancing and scanning this thread, I'm reminded of the infamous 1.5 MM, where we had to wait up to 45-50 minutes for a battle or more, because the MM kept "balancing" tiers, aircraft and players. In 1.6, Gunlion came up with a solution for the Americas server that worked pretty well given the population of the server. So this means they updated the MM to pre-1.6?

 

The fix for 1.6 was a player limit.

The MM would not start a match unless there were 3 players per side.

But the queue times increased, unless you played the most populated tier.

Even then when a clan would gather and count in some players ended up on 3 vs 3 match.

The rest of the clan was fighting it out in the air.

 

Should the current match maker be given a minimum player drop count ?

I am a little tired of the 1 ( 7 ) count matches.

 

[BrushWolf]

raulv04, on 17 February 2017 - 11:46 AM, said:

After glancing and scanning this thread, I'm reminded of the infamous 1.5 MM, where we had to wait up to 45-50 minutes for a battle or more, because the MM kept "balancing" tiers, aircraft and players. In 1.6, Gunlion came up with a solution for the Americas server that worked pretty well given the population of the server. So this means they updated the MM to pre-1.6?

 

That was the undocumented skill factor which was rather uneven in its effects.

 

losttwo, on 17 February 2017 - 12:33 PM, said:

 

The fix for 1.6 was a player limit.

The MM would not start a match unless there were 3 players per side.

But the queue times increased, unless you played the most populated tier.

Even then when a clan would gather and count in some players ended up on 3 vs 3 match.

The rest of the clan was fighting it out in the air.

 

Should the current match maker be given a minimum player drop count ?

I am a little tired of the 1 ( 7 ) count matches.

 

 

Yeah, 1 vs 1 even with bots filling the teams out are still 1 vs 1 but with the ability to sic the bots on your enemy human. A three player minimum with a time limit to force a small match would make the game much more enjoyable and less of a dice throw with the bots deciding the matches with their performance or the lack of.

[Psicko23]

I remember that 1.5 still had decent sized matches, but not as good as pre 1.5. 1.6 was the terrible 1v1 after a long wait even though there were 10 people in the same tier as you waiting in queue. I once changed my oil and took a shower while waiting in queue before the battle started. 1.7 was the 3v3 minimum after a 30 second wait, even though there was 10 people in the same tier waiting in the queue that was the Gunlion NA MM fix. MM is definitely broken for the NA server.