Seriously Fizzer, there's enough people on the ladders, it doesn't need to be this loose.


Game Link Date Ratings
vicnus and rocky1 vs THE AWESOME OTTERYCAT and Glass
20858765 THE AWESOME OTTERYCAT and Glass: 1585

That is ridiculous, it just matched up teams rated 2086 and 1585 on the 2v2 ladder.

There are 147 teams with active ratings.

It just matched the team with the 8th highest rating vs the one with the 53rd highest.

Making it match within 20% (of "Ranking", not of rating), or even 10% or 15%, would be a vast improvement.

147 * .2 = 29.4

"Rank 8" could play as low as "Rank 37". 1683 would be the lowest they could go with it set to 20%.

15% would be "Rank 30", 1735 rating.

10% would be "Rank" 23, 1804 rating.

To compare wait times those would entail:

Their game ended:

vicnus and rocky1 defeated Ju and pignouf71
20743749 2/17/2020 14:55:13 Ju and pignouf71: 1735, vicnus and rocky1: 2086

The previous game with a team over 1683 ended:

Fc Bayern and Janine defeated [-GR-] ΜΑΚΕΔΩΝ and Iceman
20712428 2/17/2020 07:27:18 [-GR-] ΜΑΚΕΔΩΝ and Iceman: 1696, Fc Bayern and Janine: 1656

That's roughly 7.5 hours.

1735:
Summer and Nauzhror defeated Negan and pete
20700846 2/16/2020 16:10:04 Negan and pete: 1920, Summer and Nauzhror: 2034

22 hours

1804:

Summer and Nauzhror defeated Negan and pete
20700846 2/16/2020 16:10:04 Negan and pete: 1920, Summer and Nauzhror: 2034


22 hours

I guarantee top teams would rather wait close to a day for a good match than get a terrible match quickly.

Technically, Otterycat and Glass are a decent team, and their rating will increase, but that's not relevant. A typical team with their rating playing a team like rocky1 and vicnus is just a lose-lose situation for rocky and vicnus.

If they win the game, their rating goes from 2086 to 2074. Yes, that's right, WINNING the game REDUCES their rating by 12 points.

If they lose the game they drop to 2040.

These matchups shouldn't happen. They're demotivational to teams that care about their rating, and ruin the integrity off the rating system. The rank 1 team left the ladder not long ago because they were frustrated that winning a game dropped their rating from 2225 to 2217.

Hold on, you can lower your rating by winning!?!?!?
That is just unfair to any good player.

Also, the rating/ranking systems on WZ could definitely use some improvements, but I have no idea how.

There are many things that could be improved, but that's a good thing to start with.

"The rank 1 team left the ladder not long ago because they were frustrated that winning a game dropped their rating from 2225 to 2217. "

Can confirm. Surdiac and I were ranked number 1 and we left ladder after winning a game actually dropped our rating to 2217, and now it's at 2210. The only game we have lost is against a team which had 2284 rating.

Edit: I understand it may be put in place so good teams don't get 1st spot by selectively beating noob teams, but we should at least get SOME rating increase, losing rating for winning doesn't make sense to me.

I'm not saying it's completely stupid, I'm just saying I don't understand why it's set up the way it is.

Edited 2/17/2020 21:07:29

Tightening matchmaking makes it so you can't play noob teams, thus no potential to get rank 1 by beating them.

Try going from 30% to 5%

Hold on, you can lower your rating by winning!?!?!?
That is just unfair to any good player.

Technically, the end product you see which we call a rating can indeed decrease by winning. Bayesian ELO (which is used on warlight's ladders) uses some fairly complex maths, but I can give you a simplified version of what's happening.

Imagine you measure two things to determine how good a given player is (so you're calculating their rating).

Measure 'X' will be what you think their rating is.
Measure 'Y' will be how sure you are they are actually at that rating.

Now, when a new team is introduced on the ladder, our variable Y will be very low. If a team wins or loses some games, Y will get better. But when a team only loses games or only wins games, the rating system is not sure how good its estimate is. This is also why runs can happen. You can basically abuse these mechanics. (I got a ladder rating of 2340 once. Am I that good? Absolutely not.)

Measure X and Y are both used when calculating one's rating. To effectively calculate it, We take the estimated ladder rating (X) and adjust it for our certainty (Y). When you see a team dropping in rating after a win, the rating system is essentially saying they're now more sure about their rating than first. But since we only see the end product instead of the volatility (Y) with it, it seems like a team can lose points for winning a game.

EDIT: to clarify: you can indeed lose points after winning a game. This happens rarely but it happens. It means that your next games will influence your rating less. Hence the many complaints about ratings not being reflective of one's skill after sustained ladder play. This is also why Bayesian ELO is a rather poor choice for a continuous event like a ladder and a better choice for tournaments. Continuous events should benefit from a rating system like Glicko2 or regular ELO.

Edited 2/17/2020 21:44:04

This makes sense actually, thank you Farah. So in other words, lowering the rating is essentially the rating system making a correction?

Edited 2/17/2020 21:57:30

It's slightly more complicated that that; there's another reason for fluctuating ratings!

So, to lay it out in simple terms, let's explain what Bayesian mechanics are first:

Farah is a drunk and tends to lose his keys around the house.
The first time he's drunk, he finds his keys on his bed.
The second time he's drunk, he finds his keys on the couch.
The third time he's drunk, he finds his keys on the bed again.

Now Farah is drunk for a fourth time (what a slob). Where should he start looking for his keys? The answer is the bed. This is Bayesian statistics in a nutshell. We adjust our beliefs all the time and incorporate that in our thinking. I lost my keys three times, and they were on my bed twice; ergo, I start looking on my bed first because I have a slight bias towards the belief that they're there.

The ladder does the same thing when evaluating your rating. When you join, you're rated average. You then play a few games and your rating starts to change. That's just how a rating system works and should work. However, what if you're on one of the extremes? Either exceptionally good or exceptionally bad?

Let's say I'm extremely good (I'm not) and I just joined the ladder. The ladder rates me as average first, as it should. Then I win a game against you, so you lose a lot of points; you just lost against an average player after all.
This is where Bayesian ELO will adjust. It remembers all games on it that have been played in the past 5 months. If my rating rises a lot because I'm such an awesome player, Bayesian ELO will adjust YOUR rating because it now sees that I wasn't an average player, but just underrated.

These corrections happen every two hours on the ladder. Actual rating updates occur every four hours. There's a 50% chance that the rating update occurs at the same time as a correction update. So if you win a game and expect some points from it, you might actually get them. But the correction update which might have happened simultaneously might deduct some points at the same time.
Combine that with being able to lose points from a win, and see why it's more common than we might expect.

"These corrections happen every two hours on the ladder. Actual rating updates occur every four hours. There's a 50% chance that the rating update occurs at the same time as a correction update. So if you win a game and expect some points from it, you might actually get them. But the correction update which might have happened simultaneously might deduct some points at the same time.
Combine that with being able to lose points from a win, and see why it's more common than we might expect."

At least in the context of warzone, this isn't right. There's not two different kinds of updates. Every update is the result of new games completing. There is some weird timing involved though, where sometimes an update adds some of the games that were completed since the last update, but not all of them, which results in what you're referring to as a "correction update". But it's still just adding new game results to the log, it's just not adding any games that you personally partook in.

Edited 2/18/2020 00:14:47

I find this weird but it gives lower teams a chance to get their ratig up

Me and Ollie were rated ~1550 and beat negan and pete who were around 1900

Gained quite a lot of rating points on that one