dode74
Joined: Aug 14, 2009
|
 Posted:
Jul 11, 2014 - 15:29 |
 
|
| bghandras wrote: | | I fundamentally disagree. The rules should cover bbox too. |
From CRP page 34:Not every league will find the same set-up as enjoyable as another league. The League Commissioner’s role is to find the best match of rules for the coaches in his league. It's explicitly acknowledged within the rules that some leagues may like other things. The rules are not based around B but around TT. You are encouraged to house rule for your league, with the same page giving ideas of how to do that (for those without the experience to house rule). If B isn't working then it should be house ruled, and people here have more than enough experience to say how. It is, though, Christer's call as commissioner. If he says jump...
| Quote: | | Bbox may not be the in the heart of game design, but it is based on TV, which is a fundamental part of the design process. Bbox shows immediately if there is any balance issue. It does not create the problem, but it makes it visible. |
The only race which shows balance issues in B (pre-change) iaw the BBRC definition (as opposed to plasmoid's reworked definition) is Zons (based on koadah's data). All the other races show a win% within the tier set for them. So house rule Zons.
If you want to start redefining the tiers, though, then you're into another set of house rules with different aims. |
Last edited by dode74 on %b %11, %2014 - %15:%Jul; edited 1 time in total |
|
koadah
Joined: Mar 30, 2005
|
| Posted:
Jul 11, 2014 - 15:29 |
|
| bghandras wrote: | | dode74 wrote: | | Quote: | | I've not examined the latest Box-tweaks, but they are Unique to Box, and hence of less interest to me |
I think that's a mistake. If there are differences between metagames which change the performances of certain races then it becomes clear that the environment is influencing things and that the environment should be altered rather than the rules, or at least rules set to fit each environment. |
I fundamentally disagree. The rules should cover bbox too. Bbox may not be the in the heart of game design, but it is based on TV, which is a fundamental part of the design process. Bbox shows immediately if there is any balance issue. It does not create the problem, but it makes it visible. |
Who are you disagreeing with? 
Sure look at the data but you need to take into account how the environment is affecting it. |
_________________
O[L]C 2016 Swiss! - April ---- All Star Bowl - Teams of Stars - 2 more teams needed |
|
plasmoid
Joined: Nov 03, 2009
|
| Posted:
Jul 15, 2014 - 01:49 |
|
Hi Dode,
| Quote: | | I think that's a mistake. If there are differences between metagames which change the performances of certain races then it becomes clear that the environment is influencing things and that the environment should be altered rather than the rules, or at least rules set to fit each environment. |
I find that in the current discussion climate, it would be a Waste. Any thoughts on why the data behaves in a certain way would just be shot Down for lack of evidence.
It is impossible to take all metagames into account, so instead I wanted to look at the 3 biggest ones (Resurrection, TV-Matched and League). With FUMBBL making changes to pure TV-Matched, their new metagame is rather insignificant in size compared to Cyanide. Which in turn Means that the data sample is so small that it will take a long time before anything tangible can be said about the new meta.
Cheers
Martin |
|
|
sann0638
Joined: Aug 09, 2010
|
| Posted:
Jul 15, 2014 - 08:15 |
|
My opponent just played Claw in scrabble. W on a triple letter. Claw is broken. |
_________________
NAF Ex-President
Founder of SAWBBL, Wiltshire's BB League on Facebook and Discord |
|
dode74
Joined: Aug 14, 2009
|
| Posted:
Jul 15, 2014 - 09:00 |
|
| Quote: | | their new metagame is rather insignificant in size compared to Cyanide |
Relative size of data samples is less relevant than absolute size.
| Quote: | | Which in turn Means that the data sample is so small that it will take a long time before anything tangible can be said about the new meta. |
There's certainly enough data to run ANOVA for overall data for most teams. Unless you do so then your thoughs on why the data behaves a certain way could just be shot down for lack of analysis... |
|
|
Garion
Joined: Aug 19, 2009
|
| Posted:
Jul 15, 2014 - 09:31 |
|
He could and should be shot down for lack of quality data . However this is only because he is claiming that he is making these changes based on data in the first place. Which to me is rather nonsensical in the first place. After all we are using a rule set where the majority of people feel cpomb is a problem in some way or another yet there is no data from any division that consistently shows it is actually good for winning. Personally I just don't believe it's possible to gain any solid information from bb stats because of the amount of variables. Things like rubbish coaches that use cpomb while most of the better ones seem to shy away from it. Or rookies using orcs to such an extreme their low tv win%.has never in any edition represented their true power. To get a real sample I think someone would need to get say the top 500 coaches and get them to play thousands of games in every format with every race. But that is impossible to achieve.
Now I don't agree with narrow tiers bb to the extremes Martin is trying to. But I respect what he is doing and they could be fun house rules. Personally I think it's all being handled in a bit if a daft way though.
First of all I would just look at player pricing across the board and correct things where needed. For example orc blitzers and human blitzers should be the same price. They have the same value stat line. Whether that is bump them up to 90k or reduce the human Blitzer to 80k. Either is fine. Next would be look at blitz ras and throw ras both of which are too expensive compared with wights and pestigors. So they either need a price reduction or another skill added so they represent their value. I'd give them thick skull as they are skeletons and it fits thematically. Then you can use data sparringly. Check if khemri are one of the best teams in the game in any format in terms of win percentage. If they are performing average or worse then you can justify removing decay, as it has no in game effect but basically makes it even harder to skill the hardest player s in the game to skill up. Then look at big guys. Apart from treemen and trolls none of them are even close to representing their true value. Compare them to a mummy who has no negatrait and regen and the mummy is far far better value. So decrease the cost of all big guys and possibly increase the mummies cost too. Then look at zons linewomen and human linemen are they really the same value or does dodge make them better. Obviously dodge makes them better. Add 10k to their price. And so on and so on. Obviously it would have a lot of opinion but that's fine as these are house rules.just flesh our your opinion. But to claim it is using data to make these changes is daft because that was the same sort of flimsy data that gave flesh gollems their 10k price hike and now forces many coaches to leave the over priced zombies at home. |
_________________
 |
|
cameronhawkins
Joined: Aug 19, 2011
|
| Posted:
Jul 15, 2014 - 09:38 |
|
| koadah wrote: |
I suspect that was old data from the old Box. Lizardmen are No 1. team at under 1500 Amazons are neck and neck with undead and not that far ahead of the rest of the pack.
|
I maintain the utmost respect for your skill and dedication to putting together valuable data.
However, I must take issue with this specific item: using data regarding the win % of Box Lizards to make generalizations about the team's strengths.
The problem is that there are a subset of teams that typically attract highly experienced coaches, and therefore often seem to perform better than they should be able to–– I would put Lizardmen here, along with Halflings, Dark Elves, Slann, and Vampires. (Orcs have this problem in reverse)
And, more importantly, I think the sample size is deceptively small. There are about 6 highly experienced coaches who are responsible for most of the Lizard Box wins–– Tarabaralla, Hitonagashi, and fxii all come to mind, and (although I will never claim to be a good Blood Bowl coach) I have also played a lot of Box Lizards, and I have a pretty good notion about how to win with them! I mean–– for cryin' out loud–– just one of my Lizard teams is responsible for more than 25% of all the wins in your calculations. If you remove Hanson's Roughriders from the equation, the Lizard win % drops from 57.56 to 53.7–– from 1st place to 5th.
Of course, you could prove me wrong by finding a bunch of low-CR coaches whose Lizard team has a 70% win-rate. I see this situation all the time with Chaos Dwarves, Undead, and Wood Elves, but I have never seen it with Lizards. Instead I just see very few Lizard teams, with good coaches being very successful and bad coaches being very unsuccessful |
|
|
koadah
Joined: Mar 30, 2005
|
| Posted:
Jul 15, 2014 - 09:45 |
|
| cameronhawkins wrote: |
However, I must take issue with this specific item: using data regarding the win % of Box Lizards to make generalizations about the team's strengths. |
I don't really see what you have to take issue with.
I am quite clearly talking about the Box after the scheduler change.
| koadah wrote: | | Sure look at the data but you need to take into account how the environment is affecting it. |
|
_________________
O[L]C 2016 Swiss! - April ---- All Star Bowl - Teams of Stars - 2 more teams needed |
|
|
|
| |
Roster Tiers
Search
