Bashing Elf

tenwit · Post by **tenwit** » Tue Apr 22, 2008 3:34 am

PuBbowler addressed that point a few days ago. If you think it wouldn't change things too much, house-rule it in in your next league. You can decide if it makes the turnover rate too low.

You might have a better chance at arguing for a runner-specific skill, especially one that has a greater turnover risk associated. Sprint, maybe.

SillySod · Post by **SillySod** » Tue Apr 22, 2008 12:03 pm

I'm quite suprised at this 43% win ratio.... exactly how big a sample size is that and what are the obvious factors which might bias it one way or the other?

At high end play dark elves become progressively better in relation to their tier one counterparts, several people have commented that they just deplore their low end play. Tournaments would seem to be the main focus for low end play.... and at the Grand Tournament I'd probably rather have a dark elf team than a high elf team (chose high elf comparison as the classic elf team).

Is it just that elves have become fundamentally worse than some of their counterparts? Possibly, but not by much at all.

The agility across the board still gives elf teams a huge selling point although I suspect that its one which is harder to get full use out of than a team that starts all with block. Practice.

plasmoid · Post by **plasmoid** » Tue Apr 22, 2008 1:00 pm

Hi SillySod,

I'm quite suprised at this 43% win ratio.... exactly how big a sample size is that and what are the obvious factors which might bias it one way or the other?

Sample size is 340 games(!)
It's a 100% open league - don't know what kind of bias that could create.
Except perhaps that coaches can have 5 teams, and most coaches have some rotation, so quite a few teams only see their early 10-15 games (AFAIK). And IMO early games probably push dark elf win percentage down.
But IMO that is the reality of LRB

For comparison,
wood elfs are 61.5% (best team), Elfs are 57.49 (third best) and high elfs are 49.16 (middle of the pack).
Cheers
Martin

SillySod · Post by **SillySod** » Tue Apr 22, 2008 2:25 pm

Personally I've played over 600 games and counting so I think the sample size could potentially be bigger, its not too bad at all though and there is a fair difference between the races

I dont really know the community that the games were played in and I cant see the precise statistics for team values for the matchups so I wont be able to pinpoint any bias. I can take a guess or two though

Firstly the whole damage aspect of BB games tends to weigh on peoples minds a bit. In open leagues you tend to find that elf coaches are naturally sought after for games because they do less damage when they play. Equally fragile teams tend to be a bit more picky about who they play because they have more to lose.

Can the elf and wood elf teams get games against weaker (rather than just equally fragile) opposition or are they more motivated to try and get it due to their potential to die?

Secondly I'd want to look at any kind of variation in coaching standards between players or over time. Do new coaches get pointed towards particular teams which will boost their chances as a new player? Do all the cool kids play with a particular race?

Finally I'd look at the style of play which you favour and the general culture of whats right and whats wrong. Is either stalling or fouling especially frowned upon? Dark elves are probably the least well suited to the traditional throwing game (especially in LRB5) so tend to struggle a bit more when the focus is on scoring fast and reliably rather than on controling the clock. They are a little more durable too so can withstand a drawn out stall or survive pressuring a cage a bit more than other elves. Their blitzer cadre is a bit more effective in securing the ball pre-cage too.

I'm not sure exactly how relevant those factors are to you but hopefully they go some way to explaining the figure.

PubBowler · Post by **PubBowler** » Tue Apr 22, 2008 7:08 pm

[quote="plasmoid"
It's a 100% open league - don't know what kind of bias that could create.
Except perhaps that coaches can have 5 teams, and most coaches have some rotation, so quite a few teams only see their early 10-15 games (AFAIK). And IMO early games probably push dark elf win percentage down.
[/quote]

I think this is a valid point.

I also think that due to playtesting lots of coaches were inclined to take suboptimal roster choices (assassins & runners) before people had a handle on tactics for them.

If I recall Dark Elves have a very good record in FUMBBL (excepting it's not LRB5) where such playtesting is not an issue and teams tend to play more games.

Jural · Post by **Jural** » Tue Apr 22, 2008 8:37 pm

You can't explain away statistics with guesswork. Either identify why the sampling is invalid, provide alternate data, or just live with the results.

SillySod · Post by **SillySod** » Wed Apr 23, 2008 4:18 am

Well they arent just vague guesses they are based mostly on the potential biases of an open league. Obviously each league is different though so I cant be sure to pinpoint stuff (and no TV stats). What I was hoping is that those points might help plasmoid identify potential biases.

If you want me to provide a stronger set of stats then take a look at these. You'll notice immediately that Amazon teams are far and away the most horrifically borked team out there but dark elves dont do too badly at all.

It should be easy to get statistics for the number of darkelf teams and games etc but most of that kind of stat page are only accessible if you know the URL now, nearly all of the directly accessible stats are coach, player, and team related.

Daefaroth · Post by **Daefaroth** » Wed Apr 23, 2008 4:33 am

SillySod wrote:If you want me to provide a stronger set of stats then take a look at these. You'll notice immediately that Amazon teams are far and away the most horrifically borked team out there but dark elves dont do too badly at all.

Isn't FUMBBL still running on LRB 4.0? Were there any appreciable team/rule diferrences between 4 and 5? Is this data really better?

Jural · Post by **Jural** » Wed Apr 23, 2008 7:39 am

Sorry, I wasn't talking to just you Sod. As a matter of fact, I respect that you simply stated factors instead of making conclusions on no data.

I'm really tired of people looking at data and then adding in guesses to explain it (orcs are broken but the data doesn't indicate it because mostly idiots play orcs... dark elves are a great side but people choose bad teams with them and can't use them correctly... Khemri look great in winning percentage and casualties, but it's all because of a few good coaches...)

Giving feedback like that is fine, as an opinion, but hand-waving over stats annoys me

By the way, I don't agree with your (And PubBowler's) FUMBBL Comparison (although FUMBBL certainly gives a definitive answer under LRB 4.0, I'd say.)

LRB 4.0 and 5.0 have massive differences, most notably in the handicap system. Now it really is critical how well a team handles chainsaws, how vulnerable they are to fireballs, etc. Of less importance is nigglers, aging, etc. Also of importance is the caliber of star players a team has access to.. something that doesn't really play as large a role in LRB 4, to my knowledge.

Aliboon · Post by **Aliboon** » Wed Apr 23, 2008 10:46 am

If you want me to provide a stronger set of stats then take a look at these. You'll notice immediately that Amazon teams are far and away the most horrifically borked team out there but dark elves dont do too badly at all.

Interesting stuff, can you get that data for all TRs? The 'zons are great as a rookie team, but if you look to the 250+ range, they are the second worst team! Obviously different teams have different development curves, so i'm not too sure how you could work it all out sensibly.

PubBowler · Post by **PubBowler** » Wed Apr 23, 2008 10:50 am

Jural wrote: LRB 4.0 and 5.0 have massive differences, most notably in the handicap system. Now it really is critical how well a team handles chainsaws, how vulnerable they are to fireballs, etc. Of less importance is nigglers, aging, etc. Also of importance is the caliber of star players a team has access to.. something that doesn't really play as large a role in LRB 4, to my knowledge.

Very much agree with this.

And I'm more than happy to work from MBBL data, it's my go to source.

But in this topic we have very experienced coaches holding these opinions:

ianwilliams wrote: DEs are a great team. They are difficult to play well, but in the hands of good coaches I think they are probably the best team in the long term.

Joemanji wrote: I like the DE team as is. I struggled with them initally, but they develop well.

Cramy wrote: The Dark Elves are a great team. They start a bit slow, but once they develop a bit, they are a menace. I have had lots of success with my Dark Elf teams both in leagues and in tourneys. And I have heard of many coaches doing the same.

These are very out of kilter with the stats.
I don't believe considering reasons for this discrepancy is out of line.

None of this is meant to detract from the MBBL stats.
And I don't think it does.

But I don't want to ignore bias in a data set either.

whitetiger · Post by **whitetiger** » Wed Apr 23, 2008 5:24 pm

There's a commercial running on Amercian Television right now wherein an actor states, "90 percent of all statistics can be made to say anything, 50 percent of the time."

Since I don't play FMMBL, I'm interested in stats from 5th edition LRB. From what I've heard and seen, Dark Elves seem to be in the lower rung of things. Is that true? And if so, what can be done to bring them up to par with the other teams?

Jural · Post by **Jural** » Wed Apr 23, 2008 6:40 pm

I would think the most important thing is just to look at the MBBL data and see if there is anything "Strange" about it. Are most of the Dark Elf teams low TV? Were 50% of the results submitted by 1 or 2 coaches who are poor?

Regarding "experiences coaches" and their "opinions"... meh. I think a coach like Ian or Galak or Joe can probably win with just about any roster. Is it really relevant where they put a team on the continuum?

I personally believe that Dark Elves are OK, and that coaches concentrate too much on using their crap positionals (runners and assassins) instead of their top notch blitzers and Witches. I also believe Dwarves are easily the best team in the game, and their stats are dragged down by people who play them 10 times because they like them from D&D or LoTR, give up on the team, and never learn to play them...

But until there is some data backing that up, I'm not going to be going around refuting league data with my opinions!

SillySod · Post by **SillySod** » Wed Apr 23, 2008 8:37 pm

From playing FUMBBL alot as well as playing in a few LRB5 home leagues I'd say that LRB4 and LRB5 tend differ mostly at high levels of TV/TR. Due to the team strength system on FUMBBL most of the games played are between teams which would have a fairly similar TV in LRB5.

Pretty much all the strongest coaches on FUMBBL start a darkelf team the same way. They take 2 blitzers, 9 linemen, FF7, and 2 Re-rolls. It seems to work pretty well... throwers are usually ignored until a bit later on and witch elves are recomended only when you have a full cadre of blitzers. This setup seems to translate into LRB5 pretty neatly because the key change when starting the team will be the oportunity to start with a lower fan factor and more blitzers, the option for journeymen is handy but not huge and shouldnt play against the dark elves favour. Whether the thrower has been downgraded or not I'm doubtful that it matters until the team starts to fully develope.

PS, I dont think that those figures are straight win percentages.... they might be but I'm not sure. They are far from meaningless in any case

(edit: this means the FUMBBL figures)

PubBowler · Post by **PubBowler** » Thu Apr 24, 2008 9:53 am

SillySod wrote: PS, I dont think that those figures are straight win percentages.... they might be but I'm not sure. They are far from meaningless in any case

If you're referring to the MBBL figures, it's wins plus draws as half a win.