Cyanide releases their Khorne Roster

[dode74]

It's not political

It seems to me that Dode and Megr1m and Dode tacitly admit that Khorne is a weaker than average team. And that pricing could have something to do with that.

It is a mid-T1.5 team. Pricing is always something to do with things, but that doesn't mean I agree with your concept of how to alter the roster, let alone whether it should be improved.

This really didn't have to go into the philosophy of 'balancing', but I'll happily explain my view Again:
I don't think an isolated Group of weak teams does the game any favours.

I know. I disagree. This is not NTBB, though, so your philosophy on that isn't hugely relevant. The tiers are as they are by design whether we like it or not, and house ruling is available if we don't. As it goes, should narrower tiers be your thing then NTBB probably works pretty well. It's not my thing, though, and it's not the game designers' thing.

But with the numbers you've quoted, there's a fairly high risk that they're not within tier 1.
Pushing their mean towards 50% would make it a lot more likely. Now that we don't have enough data to know for certain.

We can't know "for certain" for any team, only with a certain level of confidence. Your own NTBB didn't alter teams where the 95CI suggested that they could be within the relevant range (your selected ranges), so why alter this team where the 95CI suggests it could be within the relevant T1 range? It's a reversal on the restraint you showed when you were trying to narrow the tiers, and now you're not...

How about 'beyond reasonable doubt' then?

No. Because I don't agree with your argument about Blitzing skills. Lots of skills have limitations, so an argument could be made for many skills being 10k on start, but that's not how the guide is written. Only horns (and a couple of other skills) have that reduced cost, and that is likely for a good reason. I imagine that considerable thought went into that guide, so claiming it was an oversight simply because Juggs isn't mentioned (neither are a whole host of skills) seems odd, and without some sort of confirmation from the author I reject the idea. It comes across as an attempt to validate your conclusion rather than coming to your conclusions through looking at the evidence. If you do look at the evidence it's clear that Khorne are a mid-T1.5 team and don't need a buff. If you do want to buff them as a house rule then I've suggested a couple of ways here, and there have been many suggestions made elsewhere by many people. Frankly, you can obviously do what you like as far as that goes: house rule away. Just please don't post hoc justify it on what appears to be spurious reasoning.

[plasmoid]

Hi Dode,

It comes across as an attempt to validate your conclusion rather than coming to your conclusions through looking at the evidence.

I may look that way to you, but that's not how it happened.
I worked through the pricing of all the teams, and Khorne stood out. I'll send you the table, if you're interested.

I don't agree with your argument about Blitzing skills. Lots of skills have limitations, so an argument could be made for many skills being 10k on start, but that's not how the guide is written. Only horns (and a couple of other skills) have that reduced cost, and that is likely for a good reason. I imagine that considerable thought went into that guide, so claiming it was an oversight simply because Juggs isn't mentioned (neither are a whole host of skills) seems odd, and without some sort of confirmation from the author I reject the idea.

Fair enough. To me this comes across as shooting for plausable deniability.
The guide only mentions skills that are already on existing roster. (One reason is that a fair bit of reverse Engineering went into it).
So a skill like Juggernaut isn't an oversight. It's just not there, because no team has it.
When the Khorne team was created, a price had to be agreed on by the designers. Jugger on a starter was a first.

Standard pricing is 20K for a skill. (But in reality, looking at the actual rosters, a lot of skills are 10K - are 20K with a 10K discount if you will).
And overpricing is a fine Tool for making something tier 1.5 rather than tier 1.
And as a team designer, that was/is your prerogative.

But compared to horns, the parallels are abundant:
Horns (10K): Used exclusively when throwing block dice on a blitz action. This reduces the risk of a turnover and increases the odds of a knockdown.
Jugger (?K): Used exclusively when throwing block dice on a blitz action. This reduces the risk of a turnover and negates 3 decent skills.
How Jugger could be a 20K skill while Horns is 10K is beyond me.

Cheers
Martin

[plasmoid]

Hi Smeborg,
If one were to improve the team - and I understand that you wouldn't want that - then, IMO, Price tweaks are surely the least invasive.
It is still the same team and players, performing in the same way.

Now your first 5 games will just be a touch less horrible.
And a more developed team will have room for an extra lineman on the bench.

Cheers
Martin

[Smeborg]

Hi again Martin - I don't have a problem with house rules for leagues, provided "only" that the house rules do not cause coaches to leave the league, nor deter newcomers.

I don't want to engage in lengthy debate, but I suggest the designers got the costs about right for the Khorne roster. Most of the player-types look somewhat under-priced in themselves, this is offset by the repetition of 6 or 7 pairs of Jugs+Horns, skills that can only be used on the 1 blitz per turn.

I have started with 2 quite different rosters, and have theorised happily about tourney rosters for Khorne (they are fine at, say, TV110 plus skills). While they cannot have everything they want in a starting roster, theirs is better than some other teams (e.g. most of the Elves, Nurgle). The choices are quite neat: Bloodthirster or not? 2 or 3 Rerolls? You pays your money and you makes your choice.

I don't feel the roster is greatly under-powered, it's meant to be a "challenge".

All the best.

[Megr1m]

Hi again Martin - I don't have a problem with house rules for leagues, provided "only" that the house rules do not cause coaches to leave the league, nor deter newcomers.

I don't want to engage in lengthy debate, but I suggest the designers got the costs about right for the Khorne roster. Most of the player-types look somewhat under-priced in themselves, this is offset by the repetition of 6 or 7 pairs of Jugs+Horns, skills that can only be used on the 1 blitz per turn.

I have started with 2 quite different rosters, and have theorised happily about tourney rosters for Khorne (they are fine at, say, TV110 plus skills). While they cannot have everything they want in a starting roster, theirs is better than some other teams (e.g. most of the Elves, Nurgle). The choices are quite neat: Bloodthirster or not? 2 or 3 Rerolls? You pays your money and you makes your choice.

I don't feel the roster is greatly under-powered, it's meant to be a "challenge".

All the best.

I concur. The overall balance of the roster feels about right, and the design team I think did a good job. In my mind, I classify them as a 'tier 2' team and thus they seem costed correctly. Its possible to tweak them so that they are more smooth to build up, but that would only be the case if one were interested in tweaking the roster. I don't have a problem with a tiered team structure, so for me this isn't a concern.

[dode74]

I may look that way to you, but that's not how it happened.
I worked through the pricing of all the teams, and Khorne stood out. I'll send you the table, if you're interested.

Can if you like, but we worked out the pricing and I'm content with it. We know that not all the teams are worked out the same way in the guide, and we know that tweaking prices for overall team balance is acceptable in accordance with that guide, but I disagree on your analysis of the skill costing. I actually think mass horns is a very powerful skill, since it allows you to basically choose to 2d blitz wherever you want with any one player, freeing up another player for an assist elsewhere. Mass horns is more useful than horns on only one or two players, especially when it's gained at 10k. Similar is true of mass juggs, especially when combined with horns: it's a free, low-risk 2d blitz with any of those 7 players, making marking out blocks very hard to achieve for the opposition. 30k a player for those two skills works just fine and maybe it'd help you to think of it as a premium, just as you pay a premium for blodge.

That said, you've ignored a major point I made: the team's 95CI actually puts them within the tier 1 bracket. Why are you wanting to tweak them upwards when you didn't do so for teams in the same situation in NTBB?

Two pretty experienced Khorne players have told you the team doesn't need a tweak. Maybe there's a time to not tinker?

[Jimmy Fantastic]

Horns isn't that good, neither is Juggs.
No Regen on the Herald is the worst thing about the roster cos how is it even meant to be a daemon?

[plasmoid]

Hi Jimmy - it's a typo. It's really a Khorngor

[Fassbinder75]

Horns isn't that good, neither is Juggs.
No Regen on the Herald is the worst thing about the roster cos how is it even meant to be a daemon?

Dode makes about the best case he can for Juggs, but it isn't worth the TV. If it beat Side Step it would be almost worth it. Horns is ok. Agree with the inconsistency on the Herald as well.

[dode74]

it isn't worth the TV

The same is true of many, many skills. And yet...

The point is that the guide says what the guide says and the team is designed as it is designed. Change the parameters of either and you're into house rule territory, which is fine.

As for regen on the heralds, we didn't want to add a 30k skill for free, and didn't want the herald to cost more for the sake of 1kTV team options. Adding it was an option should the team prove to miss the design performance, which it didn't.

[plasmoid]

Hi Dode,
first Things first:

That said, you've ignored a major point I made: the team's 95CI actually puts them within the tier 1 bracket. Why are you wanting to tweak them upwards when you didn't do so for teams in the same situation in NTBB?

Well, I'm still thinking about whether to tweak them or not. But 2 reasons.
1) There won't be many changes in NTBB2015. 2 or 3 as far as I can tell. I intend this to be the year that wraps it up. So, I want to consider Khorne now, if at all. I can't wait 10 years for the data to come in. So I either do a tweak and list it as iffy, or I ignore the Khorne team, since it isn't an official team anyway - (NAF ignores it, GW is unlikely to make it official, and we don't even know that Cyanide will use it as DLC for BB2).

2) Did all this data really get me anywhere? I used it for NTBB2014, but it is easy enough to just shrug it all off by saying that box data isn't representative of anything other than the box.
I was originally told to follow the Mountains of data, since that would always trump anecdotal data. But when I did follow the data, it didn't show me much. In many cases it didn't show me that anything was wrong. But it didn't show me that anything was right either. It just showed me that these piles of data wasn't enough to say anything. I get the null hypothesis. And I get 'don't fix it if it ain't broken'. But for the most part I can't tell that it ain't broken. Just that we can't prove that it is. Which is not the same thing.
So if the overwhelming ammounts of data isn't worth following, I guess I should just return to second rate data: I don't think anyone has said that Khorne is clearly tier 1. Certainly not in early Development.

Bah, humbug.

Oh - on a very different note: What data do we have on Khorne?
I'd love some on popularity in MM. Is it more or less popular than CDs and Underworld, released at the same time. And does popularity perhaps change after the first year? After all, a new team has to have some extra attraction at first.

As for winstats - what do we have?
I suspect it isn't tracked for TV of the team. I think I remember OCC not tracking that. Does MM?
I'd suspect their low TV performance to be quite weak, improving (but how much) as the team builds.
I'm curious.

I disagree on your analysis of the skill costing. I actually think mass horns is a very powerful skill, since it allows you to basically choose to 2d blitz wherever you want with any one player, freeing up another player for an assist elsewhere. Mass horns is more useful than horns on only one or two players, especially when it's gained at 10k.

However powerful you think it is, the formula already prices it at 10K on a team with minimum 7 horns. If it gets weaker on a team with maximum 7 horns Guys, then that's just even worse.

Similar is true of mass juggs, especially when combined with horns: it's a free, low-risk 2d blitz with any of those 7 players, making marking out blocks very hard to achieve for the opposition. 30k a player for those two skills works just fine and maybe it'd help you to think of it as a premium, just as you pay a premium for blodge.

Frankly, I'd think you were kidding.
I'd say it was a skill weaker than the 10K horns. And it doesn't even combo all that well. Most of the time, Horns gets you 2 dice - lowering the turnover rate considerably. Lowering it further is neat, but to combo well it would have to improve the knockdown chance. And it only Works on the one blitz. Heck, I'd much rather have block (which would have been 20K, BTW).

Two pretty experienced Khorne players have told you the team doesn't need a tweak. Maybe there's a time to not tinker?

They told me that don't agree with philosophy of improving weak teams, so they'd prefer for it to stay a weak starter.
I agree. If I wanted them tier 1.5, then they wouldn't need a buff.
(but I'd be interested to know just how bad they are over their first 6-8 games (TV 1000-1400-ish).

The point is that the guide says what the guide says and the team is designed as it is designed. Change the parameters of either and you're into house rule territory, which is fine.

Except the guide doesn't say anything about jugger.
But yeah, the team is designed as it is designed. No argument there.

BTW:

Yes, purists, it's a house ruled roster anyway, but it is an official roster in other places.

Official? Don't you mean legal?

Cheers
Martin

[dode74]

So your two reasons are a self-imposed deadline and the fact that the data didn't support your assumptions. Not particularly robust. If I were feeling a little less charitable I might suggest that it was dishonest.

it is easy enough to just shrug it all off by saying that box data isn't representative of anything other than the box.

Actually it's not. We have data for R which shows similar results, and data for various leagues, including OCC, which does the same. With reference to Khorne specifically we have both OCC and FOL showing very similar results as well. If you want to "shrug off" results then you need to show there is a reason to do so.

It just showed me that these piles of data wasn't enough to say anything. I get the null hypothesis. And I get 'don't fix it if it ain't broken'. But for the most part I can't tell that it ain't broken. Just that we can't prove that it is. Which is not the same thing.

The data shows you that you cannot say it is broken. If you can't say it is broken then you don't fix it. Imagine if you took your car for a checkup to a mechanic who said "I can't say that the gearbox is broken, but I'm going to fix it anyway." You don't know it's broken, he doesn't know it's broken, but he's decided to change something anyway!

if the overwhelming ammounts of data isn't worth following

It is worth following. It just doesn't say what you want it to say. That doesn't mean you get to ignore it if you want to claim to have NTBB (or whatever) based on data.
The null hypothesis is a valid result - it's a rejection of your hypothesis, and your hypothesis is "it's broken". Since the result was the null then we reject it until the data supports the hypothesis (if it ever does). Your unwillingness to wait for further data is irrelevant to the method.
Of course, you could just abandon all use of data at all and simply try to sell NTBB on opinion and anecdote alone. That's fine, but not particularly convincing for many people, and you want to convince as many as possible.

What data do we have on Khorne?

I've given you plenty before. What more do you need? I've not done a scrape of FOL data in a little while and I need to do that, so I can give you some more data when I've done that.

I suspect it isn't tracked for TV of the team. I think I remember OCC not tracking that. Does MM?

TV-based winstats are not a requirement for balance outside of NTBB. If you want to tweak Khorne for NTBB then I suggest you take the discussion there. But yes, I can track for TV bands with FOL data. I could for OCC with a bit of work, too.

Except the guide doesn't say anything about jugger.

Regardless of your opinion on the value of Juggs, this is what the guide actually says:

• All skills add 20k to the player's price other than the following:

Since Juggs is clearly a member of the set "all skills" and is not mentioned in the exceptions then it is quite clearly a 20k skill whether you like it or not. There are many, many skills not mentioned by name but which are members of the set "all skills".
Now, the guide also says that player costs can be adjusted by 10k either way for balance purposes, but that the total discount should not exceed 40k. So you could reasonably within the guide reduce the cost of either the Heralds or the Letters, but not both. That assumes that it is needed for balance purposes, of course, which it is not outside of NTBB.

Official? Don't you mean legal?

On Cyanide it's official. Any team is legal on Cyanide if you're willing to create it and play over LAN or hotseat, but only 23 races are playable in all Cyanide leagues which don't house-rule for races. That doesn't mean I wouldn't rather have had Slann or Pact in Cyanide; I'm simply describing the situation as it is rather than how we want it to be.

[MKL]

It just showed me that these piles of data wasn't enough to say anything. I get the null hypothesis. And I get 'don't fix it if it ain't broken'. But for the most part I can't tell that it ain't broken. Just that we can't prove that it is. Which is not the same thing.

The data shows you that you cannot say it is broken. If you can't say it is broken then you don't fix it. Imagine if you took your car for a checkup to a mechanic who said "I can't say that the gearbox is broken, but I'm going to fix it anyway." You don't know it's broken, he doesn't know it's broken, but he's decided to change something anyway!

Hi Dode74,

I respect your intellectual rigour, but comparing Blood Bowl to a car that doesn't need to be fixed is a fascinating yet misleading analogy. It can be argued that it's more comparable to a game of make-believe: it's a game devised [or pulled out of the ass] by the guys of GW, a bunch of people noted for their fertile imagination and lack of flair for rules and balance. These first rules had been patched beyond recognition by successive bunch of peoples equipped with just gut feelings and limited experiences. If these many rulebooks had been perfect, we didn't had to give them progressive numbers, and I have still to see evidence that this haphazard patching system gave us a perfect rulebook.
Critics of the current rulebook (and we know there are) can compare it to, say, the slavery in the 19th century USA: maybe it works, but it's awful.
But if you want to keep to your analogy about cars and mechanics, there are a lot of people making a living out after market auto parts: someone bring his car to the mechanic to improve it, not to fix it. Maybe not you or me, but someone else likes to do it.

That just about analogies. As an (ex-)medievalist, I read with interest discussions about the statistical usefulness of limited or biased data

[dode74]

I totally agree with you on GW and the methods of patching their rules. But another patch is exactly what is being suggested here.

No analogy is perfect as they all break down at some point (otherwise they'd be the actual thing), but the car/mechanic thing works as far as I am taking it.

But if you want to keep to your analogy about cars and mechanics, there are a lot of people making a living out after market auto parts: someone bring his car to the mechanic to improve it, not to fix it. Maybe not you or me, but someone else likes to do it.

Which is what house rules are for. And in many cases we don't know if it's going to be an improvement or simply a cheaper replacement for an original part (which is by far the largest portion of the after-market, but then I'm breaking the analogy again with no real relevance ). The point, though, is that what is being said is broken hasn't been shown to be broken, certainly not with any data. If Plasmoid (or anyone) wants to change it for a set of house rules then nobody could reasonably object. However, claiming that the house rule is better and should be taken up as a main rule will require showing that either the main rule is broken or the house rule actually improves something just like after-market parts need to be able to show a performance improvement over the originals, and that is not what is being done here at all.

[Smeborg]

My recent experience in tourneys where teams are "handicapped" suggests that the balance between existing teams is very fine. One extra skill seems to be enough to make an average team into a winning team, in the hands of a decent coach, of course.

So I suggest that even "fine tuning" a roster can easily create more unbalance than it is attempting to solve.

All the best.