Thinking ahead: NTBB 2013

[dode74]

Example of a match where bought inducements made a difference: FUMBBL Cup Final X

An Apo was bought in the semifinal, too (both were used), so the bank would have prevented the wizard being bought for the final.

[plasmoid]

Right,
as stated, I started from the data here:
http://www.plasmoids.dk/bbowl/LRB6Stats.htm

Note that I used the 2nd table from the top. This combines the data from LRB6 playtest with LRB5 playtest data. This was basically because the LRB6 sample wasn't that big, but also because the rules changes between 5 and 6 were fairly small. Where there were roster changes between the 2 editions, I naturally did not use the LRB5 stats for that race.

So, outside of the 45-55 tier one, we have:
Necro 60.28
Wood 56.52
Undead 56.47
Dwarf 56.16
Elf 55.20

Vampire 42.75
Halfling 37.07
Gobbo 36.81
Ogre 32.42

It should be noted that some of the individual sample sized (especially for the 3 new teams and the 3 stunty teams) are quite small. And also that the whole sample would benefit from being a lot larger, IMO. So - I felt it was best to temper the decisions with discussion and feedback from coaches.

Anyway, Shteve0 and Dode74, you asked about Pact, Skaven, Elf and Necro.
As you can see, Skaven and Pact fall withing the tier 1 zone.
Elfs are at 55.2, and while I know this isn't a proper calculations of "margin of error", I figured that being above by 0.20 was not sufficiently clear to warrant a change.
And then there is necro. It is worth remembering that these data were collected from leagues playtesting LRB6 changes. LRB6 included nerco flesh golems at 100K! And that's the roster that scored so high. In the end, the BBRC voted to not use this change, so necro wen't back to the roster that was just 0.16 shy of 45% - again, IMO, within the margins of error.
[Coincidentally, Halflings were playtested with 50K Master Chef, which was also shot down, inflating their numbers substantially).

I hope that explains it.
I think it is also worth noting that wood elf, undead and dwarfs are (IMO) all of them teams that start strong and finish weaker, so their pure 'early game' stats are probably higher than indicated here, which is further grounds for nerfing them.

Now, of course, several controversial decisions remain:
Nerfing orcs and amazons.
Buffing Humans and Khemri.
Buffing Slann and Underworld.

I hope to be able to get back to that in roughly 12 hours.
Cheers
Martin

[dode74]

Margin of error calculations for WE, Undead and Dwarves put their range at least partially below 55%, with the high end of the Lizard range well above the mean of the WE score. Basically those sample sizes really aren't big enough to draw any real conclusions from.

I'm not trying to put a damper on things, just explaining that I don't agree with the logic.

[plasmoid]

I understand.
But those were the stats I had
And I do think thought/discussion can be a helpful addition to the stats.

(BTW, I've considered nerfing lizards before , but I had to draw the line somewhere)
Cheers
Martin

[dode74]

I understand.
But those were the stats I had

Sure, but if we can agree that the stats don't really tell us anything about which teams might be considered over- or under-achieving then surely it's worthwhile getting together more stats so that we can be a bit more clear.

And I do think thought/discussion can be a helpful addition to the stats.

No question, but without any real statistical backing then thought and discussion (i.e. subjectivity) is all we have.

For me, the first question would be whether there can be some other useful stats which we can use in order to shore up the database and ensure that the thought is at least headed in the right direction. For example, taking FUMBBL B stats and looking at matches where both teams have played less than 10 (or 20, or whatever) games only might be considered useful in the context of what you're trying to do.

On that matter, from what I've read you're trying to use the data to narrow the low development tiers but are being a bit more subjective about the high development stuff, and you've certainly suggested that high development teams (e.g. long-term WE, Dwarf, UD etc) are under review. Perhaps it's also worth backing that up with some data from longer-running leagues?

[huff]

Leave fend for the Bretonians (who look like a fun team BTW), but if you must nerf the War Dancer (one of the most iconic pieces mind you) then you better lower their cost to 110 American gold crowns. That way they can afford a couple more positionals from the start and it goes with the player formula (block and dodge on same piece is an extra 20k or 10k for them being a piece >100).

My 2 cents is that you will loose support changing such an iconic piece, no matter how good some of your other changes may be.

Nerf the killstack, un-nerf fouling and give a bump to the bottom teams and you'll be alright.

[plasmoid]

Hi all,
working on the backlog:

Nerfing orcs and amazons.
Buffing Humans and Khemri.
Buffing Slann and Underworld.
I hope to be able to get back to that in roughly 12 hours.

...and by '12 hours' I mean 'a week'.

OK
*Buffing humans: Around the time of the start of the project, there was a lot of talk about humans as a 'box' team needing to be able to stand up to orcs. As well as about humans being a top team in the fluff, and such an underperformer in the game.
So this buff was done despite the stats putting them in (very low) tier 1.
Galak and Ian (and now Babs) agreed to giving them a minor buff, as has (I think) anybody who has bothered to comment.
Not that controversial I think.

*Khemri: Weak performer, but not below tier 1. This decision was made partly because decay seemed such a hard kick to a weak team. I honestly would have been happy to just remove decay. But prior to CRP there had been a roster in playtest that was shot down in the final BBRC vote, which the playtesters had been very keen on. So was Galak and Ian. So I followed Galaks lead on this.

Though the BBRC is no more, to me it carries some weight that 3 of the 4 BBRC members would actually back the khemri and human change - despite not particularly wanting to narow the tiers.

*Orcs: Their win-stats are not great. In part, there is a theory about orcs being a noob magnet. Who knows. I was thinking more about all previous editions prior to CRP-claw, where orcs have been a top team. With a nerf to claw I think it sound to assume that orcs are back. In many ways this looks controversial on paper. But I have a hard time finding anyone thinking 90K for orc blitzers is unfair.

*Underworld: Admittedly these were only affected because of the originally planned buff to right-stuff (voids tackle on blocks). When that got canned, they got the reroll discount as compensation. I've been seriously considering removing their tiny buff to cut down the number of modified teams. But again, the change doesn't seem controversial to anyone, and being able to afford one more gobbo on the starting roster makes them more viable without quite being up there with the tier 1 teams. Feels perfect.

And then there are the Slann and the Amazon.
Both were originally not on my list of teams to modify. Both changes happened later as a result of feedback.
Slann not for being weak per se, but for being a screwed up list where a key positional was quite unpopular.

And amazons... Amazons are tricky.
As stated earlier, I was pushed from several sources to include them. I wasn't convinced, so I created a thread on TFF for discussion. That somewhat-convinced me, but I'm still not completely sure. I don't remember the source, and maybe I'm dreaming, but I do vaguely remember seeing some extreme short term and tournament stats that made them look overly good, even if their performance takes a nose-dive at higher TV.

Cheers
Martin

[plasmoid]

Hi Garion,
re: The Bank

I don't agree that it is particularly rough to vampires or ogres.
Yes, they'd be hard pressed to replace 2 of their best players in one go.
So would a lot of other teams.
But at least they have cheap linemen fillers to fill the gaps.
And remember, with the nerfed killstack, you'd be less likely to experience that.
Not to mention journeymen, mercs and stars.

But here's the kicker - I just don't feel sorry here.
It was the intent all the way through the vault process that team TV would rise and fall.
And it was never the intent to let coaches have that kind of cash in reserve for replacing players. Never.
There are so many other sources of help.

And a good thing about bank: It meddles with minmaxers.
If you want to run an super tight 11 man team, then it is a great boon to have a full treasury to keep yourself out of trouble.
But if you couldn't have your big stash, then perhaps you should consider adding that (somewhat redundant) 12th guy - because you just can't hang on to the cash.
I like that.

Cheers
Martin

[plasmoid]

Hi Huff,
I'm glad you like the look of the Bretonnians.
And I know that nerfing the iconac wardancer will put some people off.
But within the framework of the NTBB I don't think it is avoidable.

then you better lower their cost to 110 American gold crowns. That way they can afford a couple more positionals from the start and it goes with the player formula (block and dodge on same piece is an extra 20k or 10k for them being a piece >100).

Actually, I'm deliberately avoiding doing things that are ambiguous when it comes to nerf/buff. Would the team really be weakened if it could get those positionals? Hard to tell. I nerf to nerf

Moreover, I think wood elf coaches haven't paid a dime of the blodge-tax that you mention.
8347 (80K) + 3 20K skills = 140K, rounded to 120K. (The whole rounding thing, BTW, is stupid - IMO. Bad rule).
Dodge and leap are still awesome, but even if you were to argue that Fend is a 10K skill, I'd argue that the 115K (after rounding) would then deserve to be rounded up to 120K anyway - as dodge is so awesome on an AG4 player anyway.

Or to put it differently:
7347 + 3 skills = 110K (Witch Elf)
8347 + 3 skills = 120K

It certainly doesn't seem particularly harsh to me.

Cheers
Martin

[plasmoid]

Hi Dode,

Sure, but if we can agree that the stats don't really tell us anything about which teams might be considered over- or under-achieving then surely it's worthwhile getting together more stats so that we can be a bit more clear.

Hmmm..., well.... - we don't completely agree on that.
I think the data in most instances follows what was already expected.
I know that we don't have sufficient hard data - but statistics isn't the only kind of data. The social sciences consider soft data quite valid, especially when combined with hard data. To wit, if I tell you something it's just an anecdote and subjective. But if lots of people tell you the same thing, then it becomes something more.
So I do think discussion and experience has a place in this.

I do agree that some shoring up would be very nice.
But on the other hand any data sample from any source has been attacked rigorously on here. Including mine.
Box data would certainly be both TV-matched and from a skewed environment.

If anything, it would be awesome to do a league data collection, like the one I did before.
...it's just that it was such a lot of work
I don't think I can be bothered to do it again.

Cheers
Martin

[dode74]

I think the data in most instances follows what was already expected.

Well that's just confirmation bias. The data doesn't actually tell us which teams are over- or under-achieving (see my league data as a counter-example), but because it's shaping up to confirm what we think will be true is not a good reason to say "well it'll be fine". If anything I'd be very wary of that and would actively seek harder evidence.

I know that we don't have sufficient hard data - but statistics isn't the only kind of data. The social sciences consider soft data quite valid, especially when combined with hard data. To wit, if I tell you something it's just an anecdote and subjective. But if lots of people tell you the same thing, then it becomes something more.

This isn't the social sciences though: it's a game of probabilities using dice. Lots of people telling you the same thing is just lots of people who are able to tell you about it remembering that thing (perceptual bias) - it says nothing about what the general population thinks of that matter (selection bias).

I do agree that some shoring up would be very nice.
But on the other hand any data sample from any source has been attacked rigorously on here. Including mine.
Box data would certainly be both TV-matched and from a skewed environment.

I don't think there are any sources of data which are totally free from biases, but it's up to you to match the sources with the area of the game you wish to look at and justify it to yourself - they're your rules, after all, and you don't actually have to convince anyone about them. All I'm saying is that the data you're using doesn't say (yet) what is being suggested, so justifying changes based on that data is flawed.
If you're looking at very early game stats then I may be able to help with that as my Cyanide league runs "starter leagues" of no more than 9 games, and we've probably had about 30 (if not more) of these played. Fresh teams, 1000TV, no more than 10 games each. I may be able to dredge up the data from there, but it will be full of all the biases which come both with the Cyanide game and with starter leagues.

[Shteve0]

I think I agree with Dode here, assuming I understand his POV correctly.

Essentially your changelist doesn't entirely tally with even the data you're using, or at least the data seems to be insufficient to support some of the changes (or non changes) selected. If the principle is to narrow the tiers on the basis of boosting underperforming/reigning in overperforming teams, basing it on anecdotal evidence is counterintuitive; assuming we don't think we're smarter than the BBRC, the advantage we have in putting together house rules is that of hindsight and effectively huge amounts of playtest data.

Of course, I too appreciate that these are your houserules and you can do whatever you want, but just my 2c.

[plasmoid]

Actually, Dode, I suppose FOL stats would be good for 'shoring up'.
It's not TV-matched, right?
Was there any BB1 seasons, and if there were, can they be removed.
And are the stats complete? I vaguely remember you saying something about retired teams being excluded from the stats?
Cheers
Martin

PS - though, admittedly, if I change NTBB whenever new stats come in, then NTBB will never reach stability, and I want the NTBB rules to settle.

[dode74]

@ Shteve0 - yes, that's what I'm saying.

@ Plasmoid - FOL is TV-matched, but the stats which I sent you are complete. The mid-season stats I previously published for Khorne weren't, but any on that spreadsheet are. There were no BB1 seasons.
If you're just looking at low-TV, new teams then you can pick records from both FOL and from B - I doubt there's a vast mount of difference between those specific games and early league games.

[spubbbba]

I know that we don't have sufficient hard data - but statistics isn't the only kind of data. The social sciences consider soft data quite valid, especially when combined with hard data. To wit, if I tell you something it's just an anecdote and subjective. But if lots of people tell you the same thing, then it becomes something more.

This isn't the social sciences though: it's a game of probabilities using dice. Lots of people telling you the same thing is just lots of people who are able to tell you about it remembering that thing (perceptual bias) - it says nothing about what the general population thinks of that matter (selection bias).

Well it depends if the changes to NTBBL are concerned solely with balance or are also trying to make teams more enjoyable and interesting to play with and against. For that you do need to gather anecdotal evidence.

Of how hard a team is to hurt and/or beat may well be a factor. I’m pretty sure that’s why dwarfs are seen as the least fun team by many and goblins as the most fun. Sometimes goblins wreck your team but usually even if you lose you still wrack up the spps and do some damage to them.

In my eyes that’s why zons need changing more than any other team. Overall their win rate is about right, but they are easily one of the top teams at 100 TV, unless they are facing dwarfs. Whilst they get steadily worse at higher TV. The blandness of the roster really limits them and they don’t even have a differentiation of stats like other teams do.