Talk Fantasy Football

MattDakka wrote:I believed Khemri teams sucked too, but I changed my mind after seeing their BlackBox win rate : http://fumbbl.com/FUMBBL.php?page=blackbox.
I guess they suck in private leagues, but not in bash-oriented perpetual leagues.

Unless things have changed dramatically since the last time someone scraped the B data, Khemri don't have a very good win rate at all. In fact, they had the lowest win% of any tier 1 team. I'll whip up a list from Koadah's scrape (luckily, since you get to read this post after it was written, the time it takes to do so is zero as far as you can see!).
Well, lowest until I filter out same-race matches which push the % toward 0.5 without any actual utility, at which point Nurgle sucks worse. Not sure what FUMBBL is basing its BWR on, exactly, but it doesn't tell the same story the per-match data does.

VoodooMike wrote:Not sure what FUMBBL is basing its BWR on, exactly, but it doesn't tell the same story the per-match data does.

Yes, I would like to understand how BWR is calculated, there is a great difference between win rate data.

From what I've gleaned it's an Elo-type score with time limits (last 30 days). A bit like CR but with time limits.

http://fumbbl.com/help:Ranking

http://fumbbl.com/index.php?name=PNphpB ... 942#470942
"It only counts teams that have played in the last 30 days, so i would imagine that losing teams are set aside in favour of teams that win."

BWR is a poor measurement of objective racial success, then, since it is taking out how the races have done in the past, and weeding out the teams that don't last. It certainly works as an "average coach rating of active coaching of this race in box" measurement though.

It occurs to me, however, that if we want to look at how races do without letting the box racial composition skew affect the results, we should probably control for that racial composition as well as removing the same-race matches (which always push the win% toward 0.5 and give no actual information... when Slann plays Slann, slann gets 1 win and 1 loss or 2 ties... herp derp). So, here's the racial win%, using box data, controlling for racial composition and only using games against Tier 1 opponents:
Poor 'ol Khemri are still second from the bottom as far as Tier 1 teams go. Too bad there's no good way to control for TV level as well without the error getting too high to make the results useful =\

Thanks for the reply, interesting stats.
It's a pity that BWR is not accurate.

One of the most unbalancing things to me is Guard. Especially as it's only readily available to the teams that are already stronger. If you meet a team with four ST4 guys and four blitzers and all of them have block and guard, it's not much you can do to defend yourself against it. To me that is maybe the biggest issue with higher TV bashy teams.

I would perhaps make guard only available on a double for all position. Or maybe it would be only readily available to linemen.

What are your views on guard? I think it's the most boring skill in the game. (after the claw/pile on/MB combo).

Mass guard is hard to bash against, but 4 or so can be dealt with. Even mass guard can be effectively stalled.

The negation to guard is, oddly, mass horns. I think that's an issue because the teams which come with mass guard tend to have high AV, while the teams with mass horns have easy access to CPOMB...

Our league in Daventry starts on Monday. Currently there's 24 teams signed up, including 10 new teams.

http://www.level10.org/BloodBowl/index.php?section=main

Hi Dzigany (and all)

What are your views on guard? I think it's the most boring skill in the game. (after the claw/pile on/MB combo).

In case this was directed at me, then I'm sorry for the delay.
IMO, Guard is a very powerful skill, it's just been somewhat overshadowed by CPOMB.
"Wall of Guard" has been one of the trademark tactics of bashy teams ever since 3rd ed - and as such I think it has been established that it is powerful, but not broken.

Oh - Dode74 - I respectfully disagree that horns really counters guard. On the blitz, sure, but if you really get into it with a guard team then there'll be guard against you on every block you're trying to make.
The real hard counter to guard is guard. Now, something countering itself has always been a sticky game-mechanic, but I digress...

Dzigany, I'm quite fond of this rule, cooked up by someone else.
I do to some extent wish it was part of NTBB, but the consequences are so har to predict that I just didn't dare.

In short, 1 skill in each category costs 10K more to acquire.
The skills are: Block (G), Dodge (A), Leader (P), Guard (S) and Claw (M)

Cheers
Martin

Hi VoodooMike,
I must say that is a very ingenious way to calculate win-percentage.
Certainly gives a rather interesting and valid number.
Good thinking!

Cheers
Martin

Hi Matt,
I really look forward to following your season. I'll definately be mining your forums for comments and thoughts. I might even post there a few times, but for now: Good luck monday

Cheers
Martin

When trying to improve the Goblin team, instead of tinkering with the rules, the Cockerel ot the Trolls isn't it a better idea to simply improve the Goblin. Make him AV 8 and the goblin team gets a simple substansial boost. I really don't think it will take the Goblin team out of tier 3 but with an increased chance of keeping their players on the pitch they will be able to steal victories a little bit more often.

I don't have any numbers to back me up except for the vague memory that almost every time stunty teams have bin able to cause me any problems is when i haven't bin able to take out their players.

Hi Adhansa,
I actually tend to agree. I even Think it would be in perfect compliance with their warhammer statline.

However, I'm dedicated to reaching stability with the NTBB rules. So unless the 3 troll fails to improve the team, then I won't go looking for a different solution. By the end of the year I'll have the data to make that decision. So far All I Can say is that our first gobbo team did not do well.

Cheers
Martin