11 coaches online • Server time: 04:58
Forum Chat
Log in
Recent Forum Topics goto Post Claw/MBgoto Post 90+ Custom Rosters!goto Post Designer's Comm...
SearchSearch 
Post new topic   Reply to topic
View previous topic Log in to check your private messages View next topic
Nelphine



Joined: Apr 01, 2011

Post   Posted: Apr 04, 2023 - 15:24 Reply with quote Back to top

After discussions and considerations, I have updated the tiers.

Originally, the tiers were based on overall win rates. (These tiers are still at the bottom of this post).

The new tiers are based on win rates against rosters that have between 48 and 56% overall win rate (which I believe are the 'expected' competitive rosters). I have not included wins against vampires yet (despite sneaking into the 48% win rate in 2024), as the new roster is still too new, and the stats are messy (and the games against from other races mostly includes animal savagery vampires).

Teams are in order within a tier, so Undead is better than Orc for instance, despite both being tier 0.5

Updating tiers as of Apr 20 2024 C stats

Tier 0 (absurdly good): 59-63%
- Underworld
Tier 0.5: 55-58% (too good)
- Amazon (down), Undead, Orc


Tier 1 (competitive): 51-54%
- Dwarf, Human, Norse, Chaos Dwarf, Snotling, Skaven
Tier 1.5: 47-50% (competitive if you know exactly how to play them)
- Necromantic, Dark Elf, Khemri


Tier 2 (challenging): 43-46%
- Lizardman (down), Wood Elf, High Elf, Chaos, 2024 Vampires, Pro Elf, OWA
Tier 2.5: 39-42% (bad)
- Nobility, Nurgle, Khorne, Black Orc, Pact


Tier 3 (joke): 35-38%
-
Tier 3.5: 31-34% (truly awful)
- Halfling, Ogre

Tier 4: below 31% (beyond awful)
- Goblin (down, had to make a new tier for them)


Updating tiers as of Dec 20 2023 C stats

Tier 0 (absurdly good): 59-63%
- Underworld, Amazon
Tier 0.5: 55-58% (too good)
- Undead, Orc


Tier 1 (competitive): 51-54%
- Dwarf, Chaos Dwarf, Norse, Human, Skaven, Snotling
Tier 1.5: 47-50% (competitive if you know exactly how to play them)
- Necromantic, Dark Elf, Khemri, Lizardman


Tier 2 (challenging): 43-46%
- High Elf, Wood Elf, Chaos, Pro Elf, OWA
Tier 2.5: 39-42% (bad)
- Nobility, Khorne, Nurgle, Black Orc, Mixed Vampire, Pact


Tier 3 (joke): 35-38%
-
Tier 3.5: 31-34% (truly awful)
- Halfling, Ogre, Goblin


July 27 2023 C stats:

Tier 0 (absurdly good): 59-63%
- amazon, underworld
Tier 0.5: 55-58% (too good)
- orc, undead


Tier 1 (competitive): 51-54%
- dwarf, human, norse, chaos dwarf, skaven
Tier 1.5: 47-50% (competitive if you know exactly how to play them)
- necromantic, snotling, khemri, dark elf, high elf


Tier 2 (challenging): 43-46%
- lizardmen, wood elf, pro elf, chaos, OWA, nobility
Tier 2.5: 39-42% (bad)
- khorne, nurgle, pact, black orc


Tier 3 (joke): 35-38%
- animal savagery vampires, fling
Tier 3.5: 31-34% (truly awful)
- ogre, goblin


April 2023 C stats:

Tier 0 (absurdly good): over 59%
- amazon, underworld
Tier 0.5: 55-58% (too good)
- undead, orc


Tier 1 (competitive): 51-54%
- chaos dwarf, dwarf, human, Norse
Tier 1.5: 47-50% (competitive if you know exactly how to play them)
- skaven, necromantic, snotling, khemri, dark elf


Tier 2 (challenging): 43-46%
- lizardmen, high elf, pro elf, wood elf, chaos
Tier 2.5: 39-42% (bad)
- nobility, OWA, nurgle, khorne, black orc, pact


Tier 3 (joke): 35-38%
- fling, animal savagery vampires
Tier 3.5: less than 35% (truly awful)
- ogre, goblin



notes:
- orcs and khorne are both played by new coaches a lot, and may be lower than they should be (probably still tier 0.5 for orcs, and possibly tier 2 for khorne)
- snotlings get played by strong coaches a lot, and may be higher than they should be (probably still tier 1.5)
- orcs, chaos dwarves, ogres, and underworld are played noticeably longer by the same coach than any other teams
- vampires, high elves, nobility, and pro elves are played noticeably shorter by the same coach than any other teams


Obviously, at high tv (above 1600) things may start to change, but presumably with seasons this isn't really the 'intended' tv for the game, so I'm happy going with these tiers for now.


--------------------------------------------------
April 2024
---
tier 0 (over 60% win rate): underworld, amazon
tier 0.5 (57-60%): undead, orc
---
tier 0.75 (54-56%): snotling, dwarf, human, skaven
tier 1 (51-53%): chaos dwarf, norse, dark elf, khemri, necromantic
tier 1.5 (48-50%): wood elf, high elf, lizardmen, pro elf, 2024 vampires
---
tier 2 (45-47%): OWA, chaos
tier 2.5 (42-44%): mixed vampires, nurgle, black orc, nobility, pact, khorne
---
tier 3 (39-41%): halfling
tier 3.5 (below 39%): ogre, goblin

December 2023
---
tier 0 (over 60% win rate): underworld, amazon
tier 0.5 (57-60%): undead, orc
---
tier 0.75 (54-56%): snotling, skaven, human, dwarf
tier 1 (51-53%): norse, chaos dwarf, dark elf, necromantic, khemri
tier 1.5 (48-50%): lizardmen, high elf, wood elf, pro elf
---
tier 2 (45-47%): OWA, chaos
tier 2.5 (42-44%): nobility, black orc, nurgle, khorne, mixed vampires, pact
---
tier 3 (39-41%): halfling, ogre
tier 3.5 (below 39%): goblin

July 2023
---
tier 0 (over 60% win rate): amazon, underworld
tier 0.5 (57-60%): orc, undead
---
tier 0.75 (54-56%): skaven, dwarf, human, snotling
tier 1 (51-53%): norse, chaos dwarf, khemri, dark elf, necromantic
tier 1.5 (48-50%): lizardmen, high elf, pro elf, wood elf
---
tier 2 (45-47%): OWA, chaos
tier 2.5 (42-44%): nobility, nurgle, black orc, pact
---
tier 3 (39-41%): fling, khorne, ogre
tier 3.5 (below 39%): goblin, animal savagery vampires



Below are my originally proposed tiers based on C stats (Apr 2023):
---
tier 0 (over 60% win rate): amazon, underworld
tier 0.5 (57-60%): orc, undead
---
tier 0.75 (54-56%): skaven, snotling, human, chaos dwarf, dwarf
tier 1 (51-53%): norse, dark elf, khemri, necromantic
tier 1.5 (48-50%): lizardmen, high elf, pro elf, wood elf
---
tier 2 (45-47%): OWA, chaos
tier 2.5 (42-44%): nobility, black orc, nurgle, pact
---
tier 3 (39-41%): fling, khorne, ogre
tier 3.5 (below 39%): goblin, animal savagery vampires



original definitions of tiers:
tier 1 - competitive, this is where team should be, but these are the better ones
tier 1.5 - competitive, these are the challenging ones, this is where teams who should be challenges are
tier 2 - noticeably worse, but still playable
tier 2.5 - pushing into not competitive, do not play these unless you don't mind not winning
tier 3 - not competitive
tier 3.5 - what do you mean there's someone worse than tier 3??
tier 0.75 - these teams are noticeably too good, but still within expectations
tier 0.5 - ... these teams are way better than tier 0.75, so we need a new tier!
tier 0 - wat
tier 4 - ... how bad ARE goblins?

Comments and jeers welcomed!


Last edited by Nelphine on %b %26, %2024 - %23:%Apr; edited 34 times in total
MattDakka



Joined: Oct 09, 2007

Post   Posted: Apr 04, 2023 - 15:30 Reply with quote Back to top

Are these tiers based on Box Trophy games or are they based on overall C games?
Normal C or Box-only C stats?
How many games played per team? 15 or more than 15?
In my opinion EU, HE and WE are better than 48-51% if they can play more than 15 games.
Nelphine



Joined: Apr 01, 2011

Post   Posted: Apr 04, 2023 - 15:40 Reply with quote Back to top

based on all C stats. the average games per team for the 4 'high' ones are 12-13. the average games per team for the 4 'low' ones are 6-7
MattDakka



Joined: Oct 09, 2007

Post   Posted: Apr 04, 2023 - 15:44 Reply with quote Back to top

Thanks! For what is worth, I found that Orcs and Undead are slightly better than Chaos Dwarfs and Dwarfs, according to my own personal and particular experience.
So, those 2 tiers of yours seem right to me.
Not implying that the rest of your tier is wrong, just, I didn't play those races, so I have no data about them.
Imperial Nobility and Black Orcs seem to be in the right tier as well, although I played very few games with them.
Carthage



Joined: Mar 18, 2021

Post   Posted: Apr 04, 2023 - 16:01 Reply with quote Back to top

Spent a few minutes thinking if anything on here looked weird to me but no, everything looks pretty much exactly right in my opinion.
Wolbum



Joined: Jul 14, 2010

Post   Posted: Apr 04, 2023 - 17:07 Reply with quote Back to top

Good stuff Nelphine. Not surprised to see amazon and UW up top. How did you calculate total team win %? As in what was the source?

Despite your notes I am surprised to see Khorne under Nurgle, I would have expected vice versa. OWA and Chaos being competitively viable is a pleasant surprise.

Keep up the good work
Nelphine



Joined: Apr 01, 2011

Post   Posted: Apr 04, 2023 - 17:22 Reply with quote Back to top

This is based on Christer's stats for games from the C division. I would love to do slightly different stats (for instance, instead of taking overall win rate, i would take 'win rate against other rosters' and aggregate them, in order to try to account for any possible lopsided cases where a specific roster, such as OWA, seems to play an exceptionally high number of low quality rosters; or to do things where you only count a subset, such as Star or higher coaches, that should be more indicative of roster strength without the noise of brand new coaches) but I don't have enough access to do so, and if I do, then no matter what I pick, it will be controversial to someone.
Carthage



Joined: Mar 18, 2021

Post   Posted: Apr 04, 2023 - 17:33 Reply with quote Back to top

Controversy drives discussion though. All for a bit of it when it comes to stat interpretation!
mekutata



Joined: May 03, 2015

Post   Posted: Apr 04, 2023 - 18:05 Reply with quote Back to top

I think 9 tiers is not enuff. 27 tiers would allow a more detailed ranking.

I agree that Halflings don't deserve to be tiered.

Khemri should be probably lower. I assume their fumbbl win% rely a lot on decent coaches playing them, they really need a coach who can read the game and understand positioning well.
Dark Elf and Necro, who you tiered at same position, seem much friendlier to beginners.

_________________
Image
Sp00keh



Joined: Dec 06, 2011

Post   Posted: Apr 04, 2023 - 18:12 Reply with quote Back to top

Something that puzzles me on this: https://fumbbl.com/p/rosterstats?y=2023
How come High Elves are so high

Or, how come the rankings are SO different from the previous year https://fumbbl.com/p/rosterstats?y=2022
JackassRampant



Joined: Feb 26, 2011

Post   Posted: Apr 04, 2023 - 18:19 Reply with quote Back to top

We should be thinking of tiers in CR bands if at all. I think this whole thing is kinda suspect, track the stats and all as much as you want to or feel you need to, but as far as roster tiering, there are so many factors that go into each match: 1M Orcs are awesome vs 1M Chaos Chosen, but maybe at 2M the shoe is on the other foot. Lizardmen are a rough go until they find their footing, and then look out! Amazons are great in [C], but in NBFL they're a hard row to hoe. If your league is all bash teams and they're all gearing up against each other, and you get an extra 100k in starting cash, maybe Wood Elves are a good option, what with styles making fights and all, but if it's dominated by Humans and Amazons with some bash teams, and you start with 1M exactly, you're in for a ride. [C] will change in nature when seasons are implemented: a lot of teams that look like they age well now won't be so jolly when AV and bash skills are no protection against Father Time and his Agent Fees.

Rough tiering is okay: "this team tends to be easier to win with or have a higher coaching ceiling than that one," okay. But even that is highly suspect, and is only good for tweak-level balancing in tournaments. For one thing, what do you do with the more technical teams, where the best coaches need no help but the overall win rate is not so hot? It's not even a case of worse coaches bringing the CR down, a coach who can win with the easy teams might have a harder time with the hard teams. So this kind of ranking always has a teleology to it, and in an environment as massive as [C], whatever teleology seems most useful to you is likely not to jibe with the subconscious processes of other coaches.

So yeah, I appreciate the instinct that inspired the effort, and I think the underlying math does have legitimate application, but I don't think you'll get anything terribly valuable from this exercise.

_________________
Lude enixe, obliviscatur timor.
Nelphine



Joined: Apr 01, 2011

Post   Posted: Apr 04, 2023 - 19:03 Reply with quote Back to top

hey now.
i got to define a tier as 'wat'.

COMPLETELY valuable use of my time
Java



Joined: Jan 27, 2018

Post   Posted: Apr 04, 2023 - 20:42 Reply with quote Back to top

Sp00keh wrote:
Something that puzzles me on this: https://fumbbl.com/p/rosterstats?y=2023
How come High Elves are so high

Or, how come the rankings are SO different from the previous year https://fumbbl.com/p/rosterstats?y=2022


The sorting algorithm on the y= and the &m= tables is a mystery to me

The undated table makes sense though

Nelphine wrote:
hey now.
i got to define a tier as 'wat'.

COMPLETELY valuable use of my time


Orc apologist! *hides from thrown slippers*

_________________
Vlad Von Carstein's door-to-door evangelist
Nelphine



Joined: Apr 01, 2011

Post   Posted: Apr 04, 2023 - 21:00 Reply with quote Back to top

Orc apologist? KILL ALL ORCS
JackassRampant



Joined: Feb 26, 2011

Post   Posted: Apr 04, 2023 - 22:18 Reply with quote Back to top

Heeeeeeyyyy! I'z a orc! Wuzzup widdat!

Waaaaaaaaagh!

_________________
Lude enixe, obliviscatur timor.
Display posts from previous:     
 Jump to:   
All times are GMT + 1 Hour
Post new topic   Reply to topic
View previous topic Log in to check your private messages View next topic