Jump to content
Sign in to follow this  
You need to play a total of 5 battles to post in this section.
DarkKrickett

Matchmaking

57 comments in this topic

Recommended Posts

Members
13 posts
88 battles

How is this calculated? Because it seems most games over 75% are lop-sided where one team obliterates the other. Only roughly 25% are what I would say are balanced fights that come down to the last few players. 

 

One thing in all video games I have a gripe with is the matching system (not a single company has come up with a decent matching system) I want to know why? Are the unable or unwilling to make a balanced approach to a matching system?

 

 

Share this post


Link to post
Share on other sites
1,617
[SYN]
[SYN]
Members
6,708 posts
11,323 battles

Aren't you still fighting primarily bots?  Pretty sure with only 77 matches you are still in the protected noob pool MM. 

 

Share this post


Link to post
Share on other sites
Members
13 posts
88 battles
21 minutes ago, renegadestatuz said:

It is balanced by ship type.

Well there in itself is the problem...  Ship types are all good but you also need to take into consideration upgrades/captain skills/ and that persons win/loss ratio in my book. If the matching system puts 4 players all who have pretty maxed out skills against 4 that do not that will be a lop-sided battle regardless of who has skill. So yet again another game company who fails on the matching system.

  • Cool 1

Share this post


Link to post
Share on other sites
Members
13 posts
88 battles
15 minutes ago, Kapitan_Wuff said:

Aren't you still fighting primarily bots?  Pretty sure with only 77 matches you are still in the protected noob pool MM. 

 

Not really but keep telling yourself whatever you need to to sleep at night.,

Share this post


Link to post
Share on other sites
57
[-DTM-]
Members
147 posts
4,958 battles
16 minutes ago, Kapitan_Wuff said:

Aren't you still fighting primarily bots?  Pretty sure with only 77 matches you are still in the protected noob pool MM. 

 

Funny enough he has a who 14 battles at T4. Yeah still kicking around in the protect MM. I smell a troll with this one though. 

  • Cool 1

Share this post


Link to post
Share on other sites
6,799
[SYN]
Alpha Tester, Beta Testers
30,523 posts
4,420 battles
33 minutes ago, DarkKrickett said:

How is this calculated? Because it seems most games over 75% are lop-sided where one team obliterates the other. Only roughly 25% are what I would say are balanced fights that come down to the last few players. 

 

One thing in all video games I have a gripe with is the matching system (not a single company has come up with a decent matching system) I want to know why? Are the unable or unwilling to make a balanced approach to a matching system?

 

 

All they do is match by tier and ship type aka battle weight.

 

You seriously need to go to the wiki and do some reading

 Cause obviously you don't really understand the game mechanics. Here's the link.

 

http://wiki.wargaming.net/en/World_of_Warships

 

Also on YouTube look up Notser and iChase. They have tons of videos on game mechanics, tutorials etc that can help you. Here are more links

https://m.youtube.com/channel/UC-c1d4QVKztu8_3f8VqMtKw

Notser

https://m.youtube.com/user/ichasegaming

iChase

Share this post


Link to post
Share on other sites
1,418
[HINON]
Supertester
7,520 posts
7,578 battles
4 minutes ago, DarkKrickett said:

Well there in itself is the problem...  Ship types are all good but you also need to take into consideration upgrades/captain skills/ and that persons win/loss ratio in my book. If the matching system puts 4 players all who have pretty maxed out skills against 4 that do not that will be a lop-sided battle regardless of who has skill. So yet again another game company who fails on the matching system.

The mm is fine. And it'll never have a skill based mm.

Share this post


Link to post
Share on other sites
1,418
[HINON]
Supertester
7,520 posts
7,578 battles
8 minutes ago, DarkKrickett said:

Not really but keep telling yourself whatever you need to to sleep at night.,

If you're not above tier 4 then you're still in protected mm.

Share this post


Link to post
Share on other sites
Beta Testers, In AlfaTesters
17 posts
624 battles
7 minutes ago, RipNuN2 said:

You have witnessed the snowball effect.

Snowball effect or snowflake effect?  lol

Share this post


Link to post
Share on other sites
175
[TF16]
Members
825 posts
4,747 battles
44 minutes ago, Taggeth said:

Snowball effect or snowflake effect?  lol

 

Snowball effect.

 

When one team starts to take losses and it isn't balanced out, it snowballs to the effect that the stronger team can expand its advantage and win the game handily.  The game is tilted towards this effect, hence the number of blowouts compared to close games.

Share this post


Link to post
Share on other sites
2,564
Members
17,757 posts
5,084 battles
2 hours ago, DarkKrickett said:

How is this calculated? Because it seems most games over 75% are lop-sided where one team obliterates the other. Only roughly 25% are what I would say are balanced fights that come down to the last few players. 

 

 

https://en.m.wikipedia.org/wiki/Lanchester's_laws

 

Lanchester's square law.

 

Basically means that when you have differing numbers of combatants, the difference in combat power is not equal to the difference in numbers.

 

So basically, at the start of the match, your team has a 50% chance of losing the first ship. Once you do, now you have a 52% chance of losing the next one. If you do, now you have a 56% chance of losing the next one, then a 64% chance of losing the next one, then an 80% chance, and so on. (numbers are made up, for illustrative purposes only) of course, if the next ship sunk at any time is on the other team, you back up a step, reducing your chance of losing the next ship.

 

Note that this takes nothing into account but numbers, so would be the case if player skill and captains skills/modules were perfectly matched, so the percentages will vary more or less, depending on the actual team compositions. But the important thing to take away here is that, even with everything being exactly equal, snowballs, collapses, whatever you want to call them, in a game with no respawns, are the norm, and that when you are down a couple ships, it's easier for the enemy to put you down a couple more, than for you to even things up.

Edited by Skpstr

Share this post


Link to post
Share on other sites
4,582
[SALVO]
Members
16,618 posts
17,276 battles
1 hour ago, renegadestatuz said:

The mm is fine. And it'll never have a skill based mm.

renegade, honestly there will always be people who believe that if the MM doesn't include some degree of skill balancing in the MM, that the MM is inherently flawed.

 

Share this post


Link to post
Share on other sites
1,418
[HINON]
Supertester
7,520 posts
7,578 battles
17 minutes ago, Crucis said:

renegade, honestly there will always be people who believe that if the MM doesn't include some degree of skill balancing in the MM, that the MM is inherently flawed.

 

:Smile_sceptic: I know

Share this post


Link to post
Share on other sites
4,582
[SALVO]
Members
16,618 posts
17,276 battles
3 minutes ago, renegadestatuz said:

:Smile_sceptic: I know

I find it hard to blame them.  When people play a game, they want to think that they have at least a 50/50 chance of winning when the game starts.  And if there's no skill-based component to a matchmaker, there's a chance that you can end up facing a team with far superior players.  Also, some will say that it balances out over the long term, and it probably does.  But IMO that doesn't change the fact (or at least opinion) that people want a 50/50 chance of winning EVERY battle when it starts.  They don't want a long term, statistically averaged out 50/50 chance.

Furthermore, from my experience, the most fun battles are the close ones.  The ones that go down to the wire.  Not the massacres, even when you are on the winning team.  But I'd say that in the great majority of instances, those close battles are probably the result of teams that have balanced levels of skill, though of course right now that's a matter of random chance, not the intent of MM.

  • Cool 1

Share this post


Link to post
Share on other sites
335
[DPG]
Members
816 posts
4,211 battles

If we had skill based matchmaking in the game we might lose some unicums.  If the game had a more robust player base it could work.  I've played games that have a slight skill based MM and if it isn't a  runaway success or you didn't get in on day one, you're not likely to ever find matches at lower skill levels.

Share this post


Link to post
Share on other sites
1,823
[ARRGG]
[ARRGG]
Members
5,770 posts

The MM is designed to put together 24 players as fast as possible. It takes numbers of ships as per the design by the developers. The developers assign a number value to each ship.It has a RNG { random number generator } for the variables of capt skills and ship upgrades. { each possible choice of each player is different } therefore the need for the RNG. This being said there is zero way to determine "Skill" of the Individual. Some players say the Stats of a player is the baseline for which to base this on, but as everyone knows this can be manipulated. For this very reason "Skill" will never be used to make a MM.

Share this post


Link to post
Share on other sites
4,582
[SALVO]
Members
16,618 posts
17,276 battles
15 minutes ago, CLUCH_CARGO said:

The MM is designed to put together 24 players as fast as possible. It takes numbers of ships as per the design by the developers. The developers assign a number value to each ship.It has a RNG { random number generator } for the variables of capt skills and ship upgrades. { each possible choice of each player is different } therefore the need for the RNG. This being said there is zero way to determine "Skill" of the Individual. Some players say the Stats of a player is the baseline for which to base this on, but as everyone knows this can be manipulated. For this very reason "Skill" will never be used to make a MM.

it's not that difficult to come up with a simple filter for skill.  Would it be perfect?  Of course not.  Would it be better than nothing?  I think so.

Share this post


Link to post
Share on other sites
1,823
[ARRGG]
[ARRGG]
Members
5,770 posts
1 minute ago, Crucis said:

it's not that difficult to come up with a simple filter for skill.  Would it be perfect?  Of course not.  Would it be better than nothing?  I think so.

Skill can be manipulated by the individual player..... a very skillful player can degrade his stats to inject himself into a non skilled game...A filter can not detect this... Cheaters never prosper but it doesn't stop them from trying.

Share this post


Link to post
Share on other sites
2,408
[PNG]
Supertester
5,655 posts
6,455 battles

Assuming player win rate (which is a horrible determination of skill since it is dependent on class) you'd still get garbage matches. A randomly generated team would match a skill balanced team 40% of the time.

 

 

Creating perfect parity with just whoever is in queue is almost impossible.

Share this post


Link to post
Share on other sites
2,564
Members
17,757 posts
5,084 battles
1 hour ago, DarkKrickett said:

Well there in itself is the problem...  Ship types are all good but you also need to take into consideration upgrades/captain skills/ and that persons win/loss ratio in my book. If the matching system puts 4 players all who have pretty maxed out skills against 4 that do not that will be a lop-sided battle regardless of who has skill. So yet again another game company who fails on the matching system.

 

So do we use overall WR, or by ship?

 

Are two players balanced if they have the same WR, but one has a 19-pt captain, in a new stock ship with no upgrades, and the other has a fully upgraded ship with a brand new captain? If not, what WR difference would make them equal? What about two players, everything being equal except one has captain skills that fit the current meta better? What happens when the meta shifts? When you can't match players exactly because of queue limitations, what do you do? What happens when you have a unicum division of 3 Tier 10s, and 3 solo unicums all in Tier 8s?

 

The list of "what ifs" is practically endless. You'd end up with some bloated system that had a habit of glitching and spitting out ridiculous matchups, and collapses would STILL be the most common occurrence.

 

The system we have is fair and works reasonably well for what it's required to do. If you want to balance out skill by having better players on your team, then become one.

 

Share this post


Link to post
Share on other sites
4,582
[SALVO]
Members
16,618 posts
17,276 battles
10 minutes ago, Compassghost said:

Assuming player win rate (which is a horrible determination of skill since it is dependent on class) you'd still get garbage matches. A randomly generated team would match a skill balanced team 40% of the time.

 

 

 

Creating perfect parity with just whoever is in queue is almost impossible.

No, you wouldn't want to use WR because if the teams were perfectly matched, you'd end up with WR's moving closer to 50%.  Better to use a non-WR stat.  maybe use each player's average damage in their ship, compared to the average damage in that ship?  Can you game that?  Maybe.  But if you do poorly in an Emerald, but great in a Moskva, a poor DPG in the Emerald wouldn't help you get better MM in the Moskva.  Sure, you might get good MM for your emerald, but to what end?  What do you really get out of having a bad DPG in one ship if it only affects the MM when you're in that ship?    Also, while one might say, well you can try to pad your overall WR. Maybe.  But how do you really help pad your WR in a low DPG ship, since with a low DPG would probably indicate that you're not contributing much to the team.  (Also remember that when I say low DPG, I mean compared to that specific ship's universal average DPG, not your personal overall DPG.)

Anyways, just a thought.

 

Share this post


Link to post
Share on other sites
4,582
[SALVO]
Members
16,618 posts
17,276 battles
8 minutes ago, Skpstr said:

 

So do we use overall WR, or by ship?

 

Are two players balanced if they have the same WR, but one has a 19-pt captain, in a new stock ship with no upgrades, and the other has a fully upgraded ship with a brand new captain? If not, what WR difference would make them equal? What about two players, everything being equal except one has captain skills that fit the current meta better? What happens when the meta shifts? When you can't match players exactly because of queue limitations, what do you do? What happens when you have a unicum division of 3 Tier 10s, and 3 solo unicums all in Tier 8s?

 

The list of "what ifs" is practically endless. You'd end up with some bloated system that had a habit of glitching and spitting out ridiculous matchups, and collapses would STILL be the most common occurrence.

 

The system we have is fair and works reasonably well for what it's required to do. If you want to balance out skill by having better players on your team, then become one.

 

Skp, you were doing well until you essentially said "git gud".  Not everyone has the time or desire to get good, and may just want to enjoy a few games now and again.  (This isn't me, just saying...)  Is it really fair to tell that person to get good?  I think not.  I think that all the average player wants is a 50/50 chance each and every time they enter a battle.  Why is that such a terrible expectation?  And why must good players always resort to the often offensive retort of "git gud"?

Share this post


Link to post
Share on other sites
2,564
Members
17,757 posts
5,084 battles
46 minutes ago, Crucis said:

 

Furthermore, from my experience, the most fun battles are the close ones.  The ones that go down to the wire.  Not the massacres, even when you are on the winning team.  But I'd say that in the great majority of instances, those close battles are probably the result of teams that have balanced levels of skill, though of course right now that's a matter of random chance, not the intent of MM.

 

I agree, but unless we add respawning, or some kind of dynamic handicapping, the best you can do is maybe go from 25% close matches to 30%. (made up numbers) Not worth the effort and possible side effects IMO.

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.

×