Ratings changes for next season

[Rab]

I also think twdt was better without .5 but now we have it I've noticed that it's hard to use the full 41 star cap so I think it should change to 41.5

[renzi]

I think in general most 7, 7.5, 8, 8.5 and 9 stars are broken. In the old days 8 stars were twl impact players, now its more like a "consistent twl starter" which was a 7 star in the old days. This all basicly says that the playerbase has gotten closer together when it comes to skill. While I feel like isnt true.

We used this for years:
10*Top 3-5 player in your league
9*Top 10 player in your league
8*Impact TWL starter
7*Consistent TWL starter
6*Impact TWD player/ Started 1-2 TWL matches
5*Average to above Average TWD player/ TWL Sub
4*Low Average to Poor TWD player
3*First time TWD'er
2*Pubber
1*If you are given a 1* rating you are being alias checked or are potentially a cheater and are being verified by staff

This is good. Use this next twdt

[project dragon]

Why do we do a 10* system if everyone is effectively rated from 6-10*? Why not just use a 5* system if noone is ever given 1-5*s?

[RedBull]

Why do we do a 10* system if everyone is effectively rated from 6-10*? Why not just use a 5* system if noone is ever given 1-5*s?

this is what i thought also.

[hak]

Well, for starters it takes 2 players x5* to equal 10* but it takes 5 players x1* to equal a 5* player. This doesn't invalidate your question, but it shows at least that 1-5* would need a complete rebalancing.

[qan]

^

6-10 is also currently broken in terms of being a proper skill estimation. If it weren't broken, 4x6* players vs 3x8* players should result in each team winning roughly 50% of the time. It's a good enough approximation that we've been able to work with it, though, and I think enough time's been dropped into it to make it fair in an approximate sense. To pretend that it's a good system where every player is always worth playing at their value is absolutely absurd, though. It's a junk system that we've just become very good at adapting to and compensating for over years.

[olde]

this whole rating system is biased. Because from week to week there are only a handful of players who actually care and can play without distractions.

Example: I have beat every 10star wb a million times, does that make me better then them? No. But maybe I cared that day and didn't have distractions from irl shit and they didn't care and/or were distracted.

Because heres my point there are a ton of 7/8 star players who can beat 9/10 players in any givin match and you can never fix that.

But they stay rated at low stars probably because they don't care as much? Or can't lock them selfs in a room to play 4/5 hours on Sunday or log on much during the week to play with out distractions but once they are warm they can slam anyone...

Most of the player base are in there 30s with families.... the rating system will always be broken when it comes from 7-10 star.... Never gonna get it correctly.

Maybe start doing twdt with tiers instead of *ratings.
Line maxed out: consists of 2 elite tier players 1 high tier, and 2 average tier. Or anything lower.... 10star would be elite tier, high would be 9,8,7, average would be anything lower?
I took a poop typing this but one of u fks who care get the idea and could prefect it..

[ogron]

this whole rating system is biased.

Example: I have beat every 10star wb a million times

olde your level of self-assessment is interesting. you're playing with distractions but others aren't. you weren't trying but they were.

of the 1047 TWDT-D player seasons on record, your best season ever was #368.

of the 418 TWLD semis/finals player seasons on record, you've never played a single game.

the claim of "beating 10* players" is basically meaningless TWDD matches where you take a small sample size of games where you did well vs. elite players, then ignore the other 80%, and assume you could do that all the time if you "tried". you may want to consider that those 10* warbirds don't value or care about random TWDD matches and they're the ones distracted and not trying.

yet there isn't a single relevant sample of your "elite" play on record. there are players like Ro and Uprise who were misrated as 7*s and 8*s and completely dominated the league and everyone took notice of them, because playing as a low-star is a massive advantage as you get way more talent around you. yet you play as a low-star every year and go unnoticed.

like 40% of the players in this zone think they could be in the top 5% if they wanted to be. it is statistically impossible.

Coined in 1999 by then-Cornell psychologists David Dunning and Justin Kruger, the eponymous Dunning-Kruger Effect is a cognitive bias whereby people who are incompetent at something are unable to recognize their own incompetence. And not only do they fail to recognize their incompetence, they’re also likely to feel confident that they actually are competent.

[...]

The irony of the Dunning-Kruger Effect is that, Professor Dunning notes, “the knowledge and intelligence that are required to be good at a task are often the same qualities needed to recognize that one is not good at that task—and if one lacks such knowledge and intelligence, one remains ignorant that one is not good at that task.

[...]

One study of high-tech firms discovered that 32-42% of software engineers rated their skills as being in the top 5% of their companies. A nationwide survey found that 21% of Americans believe that it’s ‘very likely’ or ‘fairly likely’ that they’ll become millionaires within the next 10 years.

you may want to consider the possibility that your self-evaluation tools are failing you, and that you're at best an average player who consistently puts in average performances and that your lifetime 7 to 8* rating is stable and accurate.

[hak]

Jikes, you just ended this man's whole career..

[Poseidon]

first for the record, and not sure why ppl dun seem to grasp this, but 6-10* ".5" system is functionally the exact same as a 1-9* system

the only thing broken in this situation is the Idea that stats or even (*)'s are going to be able to accurately represent what happens when u get 5-8 players all with their own unique style together on a team. You simply can't accurately quantify it.

the best way to have accurate * ratings is to find the person in each league who proves themself best at assessing it. for instance, im willing to bet that if u took MY star ratings from last TWDTB which I posted on forums and then analyzed all the basing games that season against them, the team with the most stars from my rating would be SIGNIFICANTLY more likely to win any given game than if you compared it with the * ratings that we actually used that season.

what im saying is, my ratings were quantifiably much more accurate last TWDT than the official ones, and the post is still there for anyone who doubts it. the same would hold true this season.

ill repeat my suggestion from last season, which beam half-utilized:

find the 3 best ratings assessors for each league and have them, in their own private arena, sit down for 1-2hr and discuss and arrive at a conclusion. the shit show w the stream and all that this year was not good and you could tell it was definitely influencing the ratings all over the fkn place.

[Rab]

Dazz
Jav: 6.5 => 7.5

Kess
Jav: 9 => 9.5

Ro
Jav: 9 => 9.5

[Dwopple]

olde your level of self-assessment is interesting. you're playing with distractions but others aren't. you weren't trying but they were.

of the 1047 TWDT-D player seasons on record, your best season ever was #368.

of the 418 TWLD semis/finals player seasons on record, you've never played a single game.

the claim of "beating 10* players" is basically meaningless TWDD matches where you take a small sample size of games where you did well vs. elite players, then ignore the other 80%, and assume you could do that all the time if you "tried". you may want to consider that those 10* warbirds don't value or care about random TWDD matches and they're the ones distracted and not trying.

yet there isn't a single relevant sample of your "elite" play on record. there are players like Ro and Uprise who were misrated as 7*s and 8*s and completely dominated the league and everyone took notice of them, because playing as a low-star is a massive advantage as you get way more talent around you. yet you play as a low-star every year and go unnoticed.

like 40% of the players in this zone think they could be in the top 5% if they wanted to be. it is statistically impossible.





you may want to consider the possibility that your self-evaluation tools are failing you, and that you're at best an average player who consistently puts in average performances and that your lifetime 7 to 8* rating is stable and accurate.

link to spreadsheet plz

[MHz]

Coined in 1999 by then-Cornell psychologists David Dunning and Justin Kruger, the eponymous Dunning-Kruger Effect is a cognitive bias whereby people who are incompetent at something are unable to recognize their own incompetence. And not only do they fail to recognize their incompetence, they’re also likely to feel confident that they actually are competent.

[...]

The irony of the Dunning-Kruger Effect is that, Professor Dunning notes, “the knowledge and intelligence that are required to be good at a task are often the same qualities needed to recognize that one is not good at that task—and if one lacks such knowledge and intelligence, one remains ignorant that one is not good at that task.

[...]

One study of high-tech firms discovered that 32-42% of software engineers rated their skills as being in the top 5% of their companies. A nationwide survey found that 21% of Americans believe that it’s ‘very likely’ or ‘fairly likely’ that they’ll become millionaires within the next 10 years.

Uncanny, I was just about the invoke the Dunning-Kruger effect to describe Poseidon.

[Poseidon]

Uncanny, I was just about the invoke the Dunning-Kruger effect to describe Poseidon.

ya that would be real a touche rebuke from u there, bud, after me citing what is certainly amongst the most impressive subs in the history of TWLB TWDTB, etc. playoffs.

and in reference to this thread abt ratings, after me citing quantifiable data.

derp

[MHz]

ya that would be real a touche rebuke from u there, bud, after me citing what is certainly amongst the most impressive subs in the history of TWLB TWDTB, etc. playoffs.

and in reference to this thread abt ratings, after me citing quantifiable data.

derp

You are currently ranked 114th out of 116 players in TWDTJ. You've gone 2-15 across three rounds, with one team kill. You have a KPD ratio of 0.13. There are probably different species of animal that could do better than that.

In your preferred ship of WB, you are -11 across 7 rounds, averaging 4.8 kills per round.

And that's fine! I'm actually quite impressed on some level. You must have one of the most shocking usage:skill ratios in the game. That's an achievement of sorts.

But yet you reliably come on to the forums to give wrong opinions on what constitutes 'a good player', and give stinky ratings that look way off. Behind all your rants is the palpable sense that you've been wronged by the 'powers at be' for your middling ratings, which are actually quite generous.

You also claimed that 1-9* is "functionally the exact same" as the 6-10* with .5. Which it isn't. Under a 6-10 system, a 6* player has 60% the value of a 10* player. Under a 1-9 system, a 1* player has 11% the value of a 9* player. The former system works better, because even the worst players have played a few years, and have a kind of "starting point" value. In any case, they aren't the same - so to claim you're talking "quantifiable data" exquisitely proves my original point re: Dunning-Kruger.