FUMBBL :: Online Blood Bowl League

mrt1212

Joined: Feb 26, 2013

Posted: Jan 13, 2017 - 22:01

PainState wrote:

pythrr wrote:

18 Ogre W 11
19 Chaos Pact FF 10
20 Underworld S 2
20 Slann S 2
20 Norse S 2
23 Goblin - 0
23 Halfling - 0

All basically trash at high TV (where majors are played)

Just to make my point.

Are those races winning in league play?

***

Winning in this context = Winning their league.

Grues gonna Grue.

licker

Joined: Jul 10, 2009

Posted: Jan 13, 2017 - 22:02

You can define what you mean by tiers however you like.

Just do it using a metric which is (nominally) easy for other people to also look at.

I would certainly not choose to use a metric based off of Majors results myself, but if that's what you want, go for it.

mrt1212

Joined: Feb 26, 2013

Posted: Jan 13, 2017 - 22:10

licker wrote:

You can define what you mean by tiers however you like.

Just do it using a metric which is (nominally) easy for other people to also look at.

I would certainly not choose to use a metric based off of Majors results myself, but if that's what you want, go for it.

This is a side bar to this conversation but...

One thing that pops into my mind with high TV win rates for teams is survivor bias.

I would love to mine the stats for some of these thoughts on how they might affect the data we see.

1. Small subset of coaches accruing majority of games for a given team so that their individual skill with the team and the skill rolls accrued is a bigger factor in the data we see than being a reflection of "How does an average coach fair with this team" Cameronhawkins wasn't an average coach, he was basically Lizardmen and Underworld activity in Box for years.

2. In the case of elves in Box above 1800 tv, survior bias. The elf teams that get up to that level get up there because they have players contributing more than expected TV through mind boggling skill rolls (ahem) and they don't throw in the towel when the team gets knocked down, the coaches are usually dedicated and above average. I wish we could correlate the data back to coaching CR and see what the aggregate Coaching CR is by team by TV.

3. Conversely the opposite effect on bashy teams where coaches with lower aptitude or patience try to shortcut their way wins but wind up just treading water with the variance of the dice.

Last edited by mrt1212 on %b %13, %2017 - %22:%Jan; edited 3 times in total

PainState

Joined: Apr 04, 2007

Posted: Jan 13, 2017 - 22:11

licker wrote:

I would certainly not choose to use a metric based off of Majors results myself, but if that's what you want, go for it.

Well, this discussion has caught my fancy.

Iam doing data research on the long running leagues and the race component of winners.

Iam going in thinking the results of League play will mirror R/B play.

Iam also looking at League winners who have won 8+ games in a season. Who went undefeated in L play. Why?

In order to win a R/B major you have to win 7-9 matches in a row. I suspect I will find many seasons of League play where the winner went 10-0 or 9-0.

I find this interesting because a lot of coaches who do not play R/B majors state that it is because coaches can pimp/prep or have a advantage in team build going in... League play is no different in my mind. Some teams are just pimped/primed and ready to dominate a league.

_________________
Comish of the:

licker

Joined: Jul 10, 2009

Posted: Jan 13, 2017 - 22:19

The reason I wouldn't use Majors is because of the over all TV levels of the teams, it's not really representative of how those teams perform over the entirety of the TV range. Of course if you are only interested in high TV tiers then it's more than reasonable to use that as your data set.

I also would not put a ton of stock in league winners only, because again, there is a pretty significant coaching aspect to that, so you're really skewing your data based off of what races the best coaches chose to play.

Is there any reason not to simply look at total league records for all teams? The win% there is as good as from anywhere else (assuming you can ignore or remove 'non-competitive' games some leagues use for various reasons).

Desultory

Joined: Jun 24, 2008

Posted: Jan 13, 2017 - 23:07

As already mentioned tournaments will have higher TV teams in general so races that do poorly at low TV (Vampires) is irrelevent in race selection.

Another big differnece is that the average win rate will be effected by draws, which can not occur in tournaments.
So all those draw heavy (bashy) and higher AV teams like Orcs may benefit significantly from a tournament set up.

mrt1212

Joined: Feb 26, 2013

Posted: Jan 13, 2017 - 23:17

PainState wrote:

I find this interesting because a lot of coaches who do not play R/B majors state that it is because coaches can pimp/prep or have a advantage in team build going in... League play is no different in my mind. Some teams are just pimped/primed and ready to dominate a league.

It's like complaining that some NFL teams are better at practicing and therefore enjoy an advantage in any given game. Pumping up your team is part of the nuffledamn game and tourney scene in B and R, it takes dedication and effort and risk of loss. Since average coaches like myself can invest their time and effort into building a team that overcompensates for my skill as a coach, I'm pulling one over I guess. I want to slap these complainers in the mouth.

The basis of their opinion comes down to wanting to play a variant of chess within blood bowl.

licker

Joined: Jul 10, 2009

Posted: Jan 13, 2017 - 23:45

I don't really pay much attention to it, but I've not heard those complaints myself.

I don't play in tournaments either though, but certainly not for those stated reasons. I would think that (for R) the reason would be more that certain coaches don't want to pimp their teams (because lack of time, or lack of interest in extreme picking), and so know they are at a disadvantage (they are right about that at least) and so they don't find the format interesting.

That's not really in indictment of the format itself, it's just that some people may not like the format.

PainState

Joined: Apr 04, 2007

Posted: Jan 13, 2017 - 23:49

mrt1212 wrote:

The basis of their opinion comes down to wanting to play a variant of chess within blood bowl.

Wait? what? Hold on?

The #1 difference of League play to open play is this.

You have a top Tier coach in League play, he is top of the league most of the time. Then disaster strikes. His Pimped out team takes massive CAS and his team spirals because it will take him two season just to recover. Next two season he spirals to the bottom of the lists and is just trying to build back up.

In R/B play that recovery time could be reduce to 10 days.

The point I made though does not change...Are there Tier 1,2 and 3 teams. If your sole goal is to win a league or tournament there are tier of teams which will aid you in this cause. Your goal will be very hard to achieve if you play Ogres in any Div.

Lets not leave out:

Human
Slann
High Elf
Khemri

The list goes on and on of Tier 2 teams.

_________________
Comish of the:

PainState

Joined: Apr 04, 2007

Posted: Jan 13, 2017 - 23:59

So if we come full circle back to Chaos in a League like WIL, SWL, NFBL or NWL.

Going in you have to expect at a min. 4 seasons of totally futility, slowly climbing up the ladder and then in season 5-6 you finally have a team that can compete barring any serious set back on your +AG ball handlers.

Chaos is a fringe Tier 1 team and it totally depend's on your few +AG ball handlers.

_________________
Comish of the:

licker

Joined: Jul 10, 2009

Posted: Jan 14, 2017 - 00:02

If you define tiers that way then I think you can come up with a reasonable list. It's still limited (obviously) to being designed with the goal of winning a string of games rather than some general win%.

You could look at R and B data to see what teams have the most X game win streaks (or X-1, X-2, ...) I suppose. Well not sure how easy that is to do given the tools we have (thanks Koadah btw), but it would serve as a decent proxy I guess.

mrt1212

Joined: Feb 26, 2013

Posted: Jan 14, 2017 - 00:03

Painstate, people that don't play B or R tourneys because they feel they'd be disadvantaged due to not putting in work on their team like other coaches do are either locked into a mentality where they value evenness of matchup above all else and want the game to revolve solely around their wits versus the opposition OR they legitimately have no teams within spitting distance of the theorized mean of all entries in the tourney.

I can't imagine many other objections that don't come down to ego preservation like most things in the world.

PainState

Joined: Apr 04, 2007

Posted: Jan 14, 2017 - 00:07

licker wrote:

If you define tiers that way then I think you can come up with a reasonable list. It's still limited (obviously) to being designed with the goal of winning a string of games rather than some general win%.

Ok, lets step back. Win % of the large population of FUMBBL is one thing.

You could argue that the large win % of, lets say, Wood Elfs, comes in a open environment they can pick matches. Further more those matches have no meaning.

League and Tournament play is different. In that environment all you can look at is which races have won the most. Which races, based on stats, go farther in these closed environment's. Then if your goal is to win a tournament/league in a closed environment, which race should I choose? The R/B stats I posted earlier should hold true.

_________________
Comish of the:

licker

Joined: Jul 10, 2009

Posted: Jan 14, 2017 - 00:07

PainState wrote:

So if we come full circle back to Chaos in a League like WIL, SWL, NFBL or NWL.

Going in you have to expect at a min. 4 seasons of totally futility, slowly climbing up the ladder and then in season 5-6 you finally have a team that can compete barring any serious set back on your +AG ball handlers.

Chaos is a fringe Tier 1 team and it totally depend's on your few +AG ball handlers.

Mmmm... I can only really speak for NBFL, but both Brasky and Azure won it with Chaos in much less than that number of seasons (of course NBFL jump starts teams...). Otherwise, and no offense to those other Chaos coaches, Chaos has been a mixed bag in terms of finishing, but certainly Jeffro and Ritzi have made multiple playoff appearances with their Chaos teams historically.

Again, I think part of the issue with looking at just league champions is that the ability of the coaches matters so much, so the limited number of data points per coach affects your results.

Azure for example (one of the best coaches on the site) won in the NBFL with whatever team he chose (WE and Chaos while I was there), Kryten has run his Humans far better than anyone else probably could, and Grue has landed Slann in more super bowls and won more of them than anyone else with any other race.

So yeah, there's a problem, because my guess is that if Grue switched from Slann to basically anything else he'd still be claiming Conference and League crowns. But since he's only played Slann, it makes them look better than they probably are generally.

mrt1212

Joined: Feb 26, 2013

Posted: Jan 14, 2017 - 00:10

Like I said, Grues gonna Grue