Narrow Tiers and Galaks Wish List

[legowarrior]

That something I could get behind.

[garion]

The CRP data I have from FUMBBL (B, R and L) during CRP for about 90,000 games puts the win%s as:

Code: Select all

Race        Played    Win% bounds (95CI)
Amazon       4866     58.22 - 55.44
Wood Elf     7921     56.37 - 54.18
Chaos Dwarf 12534     55.59 - 53.85
Skaven       8782     55.74 - 53.66
Dark Elf    10128     55.37 - 53.43
Lizardmen    6822     55.35 - 52.98
Undead       7817     54.79 - 52.58
Elf          3753     54.06 - 50.87
Dwarf       10408     52.31 - 50.39
High Elf     5628     52.51 - 49.90
Norse        9380     52.16 - 50.14
Necromantic  9559     51.74 - 49.74
Human        8740     50.83 - 48.73
Chaos       15358     50.49 - 48.91
Chaos Pact  10158     49.44 - 47.49
Nurgle      11178     48.34 - 46.48
Slann        5441     48.40 - 45.75
Orc         12413     47.82 - 46.07
Khemri       5386     47.90 - 45.23
Underworld   4060     44.84 - 41.79
Vampire      3319     42.41 - 39.06
Goblin       2105     32.71 - 28.77
Halfling     2138     29.23 - 25.45
Ogre         2294     27.89 - 24.29

Which suggests to me that Amazons and WE may be more worthy of being looked at.

Agreed, but he has already looked at wood elves and amazons and nerfed them and Chaos dwarves are third ion that list without any nerf. Removing mutation access is a very minor nerf. But one that i personally think is definately needed.

As for ogres, you can try all you like messing with that team but while they have St1 players playing they will always suck.

[dode74]

If CD need looking at then so do Skaven and Dark Elves, surely? They have almost identical win% brackets.

[garion]

I don't think there is anything you could do to those rosters to balance them without changing the team quite dramatically or losing the teams character really.

But I personally believe Chaos Dwarves win percentage is also being severly hampered as it is a complete noob team now, they are the second most popular race for obvious reasons. Just behind chaos.

There were similar problems with Orcs data in lrb4, all the new coaches and poor coaches picked them to learn with and as a result their stats were worse than they otherwise would be, when you looked at leagues and tournaments in lrb4, Orcs performed to a very high level dominating in both areas.

Dark Elves i agree are just as good as CDs but they aint great at a low TV so it would be very hard to not mess them up low TV but make them weaker high.

Skaven the same really, imo slightly worse than those two teams because they arent consistently great at a high tv, but you couldnt change them without ruining their character.

Plus even purplegoo has said on many occassions he thinks giving M access to CDs was possibly a mistake, and he never shares opinions lol So something must be wrong.

This was my only play testing experience with them in CRP - http://fumbbl.com/FUMBBL.php?page=team& ... _id=656217

But i found them to be ridiculously good and just retired them through boredom in the end, and I am by no means a great coach.

[DoubleSkulls]

Removing mutation access is a very minor nerf. But one that i personally think is definately needed.

Do you really need that and weakening the CLPOMB? (which is the main benefit of mutation access right?)

[Tourach]

The CRP data I have from FUMBBL (B, R and L) during CRP for about 90,000 games puts the win%s as:

Code: Select all

Race        Played    Win% bounds (95CI)
Amazon       4866     58.22 - 55.44
Wood Elf     7921     56.37 - 54.18
Chaos Dwarf 12534     55.59 - 53.85
Skaven       8782     55.74 - 53.66
Dark Elf    10128     55.37 - 53.43
Lizardmen    6822     55.35 - 52.98
Undead       7817     54.79 - 52.58
Elf          3753     54.06 - 50.87
Dwarf       10408     52.31 - 50.39
High Elf     5628     52.51 - 49.90
Norse        9380     52.16 - 50.14
Necromantic  9559     51.74 - 49.74
Human        8740     50.83 - 48.73
Chaos       15358     50.49 - 48.91
Chaos Pact  10158     49.44 - 47.49
Nurgle      11178     48.34 - 46.48
Slann        5441     48.40 - 45.75
Orc         12413     47.82 - 46.07
Khemri       5386     47.90 - 45.23
Underworld   4060     44.84 - 41.79
Vampire      3319     42.41 - 39.06
Goblin       2105     32.71 - 28.77
Halfling     2138     29.23 - 25.45
Ogre         2294     27.89 - 24.29

Which suggests to me that Amazons and WE may be more worthy of being looked at.

Agreed, but he has already looked at wood elves and amazons and nerfed them and Chaos dwarves are third ion that list without any nerf. Removing mutation access is a very minor nerf. But one that i personally think is definately needed.

As for ogres, you can try all you like messing with that team but while they have St1 players playing they will always suck.

Well fumbbl does always have scewed data for amazones because of the TV trolling (or whatever its called) where people stay below 120tv with amazons and have a severe advantage against newer teams. Same goes for the myriad of chaos dwarf teams abusing a clawpomb legend at low tv.

[garion]

Removing mutation access is a very minor nerf. But one that i personally think is definately needed.

Do you really need that and weakening the CLPOMB? (which is the main benefit of mutation access right?)

I should say - i disagree with the narrowing the tiers by and large.

If i had it my way they would have a troll and no mutation access.

But for narrowing the tiers I would say yes, they just dont need Claw. Claw was changed and made more readily available to counter high Av teams like Dwarves, Chaos Dwarves, Orcs that had in previous editions dominated perpetual leagues... was it not? None of the other av9 monster teams have access to claw so why CDs? Other than it being a strange throwback to 2nd ed. I can't see any reason for it.

CDs are great because they have block and tackle on 6 player. So next skill guard then MB same as dwarves. They lose a few extra players with St access but gain a big guy and a 2 St4 players over normal dwarves. Do they need the ability to destroy Av9 teams? I dont think so.... they can beat those teams anyway, by out guarding them or by out running them with their St4 block + break tackle ball carrying bulls. They don't need to smash them to pieces. They have a the huge advantage over Chaos and Nurgle early on because of their starting skills and they have further advantage later on because they are the only hyperbash team that has room to take grab and fend, which counters other teams fend and protects them from Pilling On.

I don't know why they were given claw access, other than a nod back to 2nd ed, was it an attempt to balance getting rid of the Troll?

[Tourach]

Removing mutation access is a very minor nerf. But one that i personally think is definately needed.

Do you really need that and weakening the CLPOMB? (which is the main benefit of mutation access right?)

As doubleskulls sais, clawpomb is nerfed a bit in ntbb so this will weaken the long time dominance of chaos teams.

[dode74]

As Tourach says, CDs are an almost-perfect minmax team, which will tend to take the win% up considerably. As you can see from the download here, CDs are good in the low-mid TV ranges, but once past 1800TV their win% drops considerably (which is similar to normal dwarves, which start the drop at a slightly lower TV). I'd like to look at some CD data in league settings before coming to any conclusions about there being a problem with the team rather than the environment.

[garion]

Removing mutation access is a very minor nerf. But one that i personally think is definately needed.

Do you really need that and weakening the CLPOMB? (which is the main benefit of mutation access right?)

As doubleskulls sais, clawpomb is nerfed a bit in ntbb so this will weaken the long time dominance of chaos teams.

Yes I know, but as I already said, they just dont need Claw MB, see my last post. They just dont need to ability to be a team that bashes av9 teams. They are an av9 team. Claw and MB havent been touched in that house ruled league. Why give an Av9 team access to Claw MB when no other av9 teams do? They just dont need it!

And yes Dode as always we need to wait for more data. It would be nice to be able to have discussions about Bloodbowl without having to constantly listen to you saying we need stats, we havent got stats for that etc.. etc... It soo boring.

Discussion can be interesting without these things.

[dode74]

Well in that case the data says there is no issue

A question which perhaps Plasmoid would be better to answer: by "narrow tier" do you intend to reduce the T1 teams to closer than 45-55%?

[plasmoid]

Hi guys,
I suppose one man's data is another man's junk
But seriously, it is always tricky to pull data from just one source - and certainly from a source with known flaws and unique characteristics. Just like it is problematic to brush all data aside and go with gut feeling.

In the data sample I started working from, CDs were dead on 50%.
Some places they perform better, some places they perform worse.
But I haven't experienced (nor have I had lots of feedback stating) that there is a real problem. Top of tier 1? Fine.
Admittedly, I'm probably also biased by the fact that I don't want to mess with more than half of the existing teams.

And Dode,
yep - Ideally I'd love to fit all teams into 45-55 for league play, though I'm pretty sure that it can't quite be done, and we'll never get massive data samples for NTBB anyway.
But that's why I've pulled back orcs, woodies, undead, dwarfs and amazons.
And that's why I'm buffing ogres, halflings, gobbos, vampire, khemri and humans. The tier 3 teams may fall a bit short of the 45% mark, but that's OK too.

Cheers
Martin

[dode74]

And yes Dode as always we need to wait for more data. It would be nice to be able to have discussions about Bloodbowl without having to constantly listen to you saying we need stats, we havent got stats for that etc.. etc... It soo boring.

Discussion can be interesting without these things.

Discussions go nowhere without these things, as they will always be entirely subjective and based on gut feeling without them. It's like discussing politics - everyone thinks their opinion is right. I note, though, that examples are fine by you, but what are stats but an amalgamation of numerous examples?

Martin - thanks for the clarification. What do you consider relevant league data? Cyanide? [L] from FUMBBL?

[plasmoid]

Hi Dode,

Martin - thanks for the clarification. What do you consider relevant league data? Cyanide? [L] from FUMBBL?

Relevant data for the NTBB rules would be league play with exclusively NTBB and no other house rules.
Not a lot of those leagues around I imagine

...but relevant data as a basis for gauging team power?
Hmm...
A little tricky actually, but I'd have to say:
League play with a reasonable spread between low, mid and high TV play. It would have to be coach or fixture matched, rather than TV-matched. And all (or the significant majority) of the games played would have to matter - in terms of a scoring system, championship or whatever.
Also, power expressed as a percentage is a good starting point. But there is a huge difference between a team being (low-mid-high TV) 45-50-55 and a team being 10-50-90 - the latter would need fixing, which is why I've tried to figure out which teams are start-strong and end-strong, and tweaked accordingly.

Cheers
Martin

[plasmoid]

Say, Dode, if you can get over the uglyness of play by email, you're very welcome to join playtest tournament 2, next summer