Long term success in a league

[Hitonagashi]

It's not formatted, but this is the Box data I grabbed a while ago and made look similar. I didn't care for draws, so the data looks only at who won each matchup....possibly should have weighted draws as half of wins.

http://hitonagashi.github.com/fumbbl_games/

[Chris]

The target was for a tier 1 team to have a floor of around .300 even if it were a massive underdog - so underdogs always had a chance of winning a game.

It seems there is a bit of noise about wizards being a bit too effective. Plasmoid wanted to drop lightning bolt and matt you think the fireball is too good for skaven (and presumably other fast teams). It makes me think perhaps wizards should be unchanged but 200k.

Or given that different teams have different utility with the wizard move it to the team roster to be listed at different costs depending on teams alongside their star players.

So Orcs have it listed for 150k, wood elves for 200k.

[the.tok]

Thanks very much for the data, really helpful

I'll study it with great interest.

Some things surprise me, like goblins having favourable match-ups against orcs. Wouldn't have thought that. Maybe coach skill is to blame for that anormality

[dode74]

It's not formatted, but this is the Box data I grabbed a while ago and made look similar. I didn't care for draws, so the data looks only at who won each matchup....possibly should have weighted draws as half of wins.

http://hitonagashi.github.com/fumbbl_games/

I've already got data for FUMBBL and looked at from 20 Jan to 1 Oct 11 thanks to Koadah, although I'll take a look through this as well as there may be fields which weren't included by Koadah. That's ~57k matches. I have earlier data, and also data from other formats, but I looked specifically at and plotted TV difference against win% (which does include draws - they should be weighted for half a win as Hito says). The plotted results look like this:
http://i1013.photobucket.com/albums/af2 ... FUMBBL.png
The widening margins of error show that as TV difference increases the data becomes less reliable. This is due to the 15% rule on FUMBBL which means that it's not the best place to test how win% differs with TV difference, since we never actually reach 30% with any reliability - 85% of the games were played with a TV difference of 100 or less.
It is markedly different to FOL though, which looks like this (disconnections and concessions were removed for this):
http://i1013.photobucket.com/albums/af2 ... /FOLS4.png
Noticeably, the win% for the low TV team on FOL reaches 30% much sooner than the FUMBBL data would were it extrapolated. I suspect this is due to low cost stars being available, and possibly also due to the overall better coaching standards on FUMBBL. Anyone interested in the FOL data can get the spreadsheet here: http://www.mediafire.com/?9ps2pbsufusm127 . The initial data is the first few columns (up to column CI).

I quite like Chris' wizard solution, but it might lead to some difficult decisions regarding who gets it at which price (difficult in that nobody will agree, not that we don't have ideas).

[mattgslater]

Dode,

You seem to misunderstand a couple of things I said.

The "rough picture" was the target.

I understand. That's where you're in error. A "rough picture" is like a view from 100 yards away. You can see that the objects are all there, but you can't make out the details. When you're inspecting your progress, that's a great place to start. But you just can't stop there. You gotta go into the data, a lot deeper than those tables you made.

What you're actually talking about here is the ability of TV to accurately reflect on-pitch capability against a wide range of opponents.

No. I see where you could have made that mistake, as my argument would be a good way to open that awful can of worms. But I'm not going there. I don't mind the presence of sweet spots and leveling mechanisms, and I don't think that we could ever factor the noise out sufficiently to do more than the mildest smoothing of the natural curves. What I'm talking about is this

[T]eams with low sweet spots ... can't be crippled, but teams with high sweet spots, which rely on spread-out skills and don't get good inducement bargains, can be rendered irrelevant for a long time with a streak of bad luck, much like those light teams used to be subject to crippling in the days before Journeymen.

to which you responded

Skaven most certainly can be crippled (literally as well as figuratively) through effective play. Certainly the wizard can be effective, but it can also be a total damp squib. I think the teams which can struggle are those with players such as BOBs which have a lot of potential but take time to make solid.

The second part is your concurrence with my point, great. But the first part shows we need to break in-game considerations from metagame considerations.

At their best, with good luck (a Blitz on the kickoff, then a bad turn on your part, or a good Wizard hit), a Skaven team at any TV level will run all over you, and you just mark the game up as a loss and start bobbing for Cas. With bad luck, they get massacred and maybe players die. Then what happens? Linerats are seamlessly replaced by Journeymen: only if they lose their Kicker or all their Tackle do they have to, gah, skill one of five MA7/AG3 guys. Oh, the horror. Blitzers peak high, but Gutter Runners, for all the nifty combos they eventually get, are just as efficient out of the box. So there are three, maybe four guys on the team who can hurt the team's efficiency by dying.

Orcs need a lot of Block and Guard, and they want Mighty Blow, plus the toolbox skills. That all costs improvements. These improvements are good in TV terms. Therefore, Orcs peak out much higher than Skaven. An Orc team before its peak is not as likely to win as a peak Orc team (by definition). It's unusual that a big Orc team will take a ton of damage, but a) it's a lot more common than it used to be, both in absolute and proportional terms, and b) that damage tends to fall on the BOBs, who get the best mileage from skill-stacking, start off the least efficient as rookies, and are hardest to improve (especially for the first skill).

This didn't used to be a big deal. Way back when, it would be rare that you'd lose two BOBs in five games, because there was no ClawPOMB, or even ClawMB. And Skaven could be killed too, far more easily than Orcs. Take that starting 11, kill both Stormvermin and two GRs in a two-game span, get just one bad Winnings roll, and voila, noncompetitive team. But they fixed that with Journeymen and inducements and divorcing Treasury from TV. So now it's just the Orcs (and Dwarfs, and to a lesser extent Undead, Nurgle, and Khemri) who can be killed.

I agree with DS on the wizard. Currently it's a no-brainer with a 150TV difference, and I know that people will actually cut rookies to ensure they get one.

We're together on that one. We're not together on what to do about it. I think the proper answer isn't in making Wizards worse, but in throwing some crumbs to those once-fat teams that got left out of the new edition's goodie basket.

[DoubleSkulls]

The target was for a tier 1 team to have a floor of around .300 even if it were a massive underdog - so underdogs always had a chance of winning a game.

It seems there is a bit of noise about wizards being a bit too effective. Plasmoid wanted to drop lightning bolt and matt you think the fireball is too good for skaven (and presumably other fast teams). It makes me think perhaps wizards should be unchanged but 200k.

Or given that different teams have different utility with the wizard move it to the team roster to be listed at different costs depending on teams alongside their star players.

So Orcs have it listed for 150k, wood elves for 200k.

Happy to be disproven, but I'm not sure its really that variable. My recollection of advice I've seen over the past few years on inducements almost invariably says take the wizard at 150k or any amount above that. If the wizard is a no brainer if you've 150k or more then, IMO, its probably undercosted for everyone.

[mattgslater]

The target was for a tier 1 team to have a floor of around .300 even if it were a massive underdog - so underdogs always had a chance of winning a game.

It seems there is a bit of noise about wizards being a bit too effective. Plasmoid wanted to drop lightning bolt and matt you think the fireball is too good for skaven (and presumably other fast teams). It makes me think perhaps wizards should be unchanged but 200k.

Or given that different teams have different utility with the wizard move it to the team roster to be listed at different costs depending on teams alongside their star players.

So Orcs have it listed for 150k, wood elves for 200k.

Happy to be disproven, but I'm not sure its really that variable. My recollection of advice I've seen over the past few years on inducements almost invariably says take the wizard at 150k or any amount above that. If the wizard is a no brainer if you've 150k or more then, IMO, its probably undercosted for everyone.

* I think the Wizard is undercosted for everyone.
* I think the Wizard is undercosted for some teams more than others, notably fast teams and medium-weight bash teams.
* I think the Wizard is undercosted against some teams more than others, notably slow teams and medium-weight bash teams.
* I consider a Wizard any time I'm a 150k+ underdog, but situationally I sometimes forgo one, particularly against speed teams.
* I think the primary differences between Fire Ball and Lightning Bolt are situational and coaching-driven. I use the threat of FB to try to loosen the cage (and, if the kick goes shallow or I've got some speed, also to use the LB threat to deter keepaway games), and then either force them to tighten up and 'Ball the action at the start of my turn when the time is right, or force them to leave a route to the carrier (okay if I have to blitz to get there, if a push will do), then 'Bolt the carrier and scoop the ball. Or, if I get control without Wizarding, I save it for the second half, use it on offense to help me reposition if I get stalled out.
* I don't think any of this is a problem.

[DoubleSkulls]

* I don't think any of this is a problem.

In your post on why underdog TV skaven are so good I came away with the impression that one of the key issues was the ability of Skaven to use the threat of a wizard to such good effect - in that it made it harder for bash teams to concentrate their force around the ball carrier, and at the same time were generally in a very good position to exploit the actual use of a fireball.

Although you may not like wizards against speed teams there seem to be a large number of coaches who find the lightning bolt too effective at stopping them.

Regardless if the universal advice is to take a wizard then it probably means its under priced.

[the.tok]

Regardless if the universal advice is to take a wizard then it probably means its under priced.

Totally agree with that.
Also, don't you think it would still be taken as much if it was 200k?
Says something I think

[swilhelm73]

I think there is no question that the wizard is the single best inducement overall. Generally then for the cost of 7 skills you have a fair chance of stopping a TD or turning a stalled drive to a scoring one.

Presumably this would be more of an issue at high TV - the difference between a TV1150 and TV1000 team is much more extreme then TV2150 vs TV2000.

Perhaps this does merit an analysis of win/loss for teams with a difference of 150-190 and those 50-140. If the former perform much better then the latter this inducement creates an odd incentive to weaken you team in terms of players or RRs just to get the wizard.

[dode74]

I understand. That's where you're in error. A "rough picture" is like a view from 100 yards away. You can see that the objects are all there, but you can't make out the details. When you're inspecting your progress, that's a great place to start. But you just can't stop there. You gotta go into the data, a lot deeper than those tables you made.

I disagree. Maintaining the roughness is in keeping with the game. There are no dead cert winners as underdogs, so I don't see it as being too rough. Go in too deeply and you overtinker and remove all the variation.

Reference the skaven issue, skaven lose a lot more players in general than orcs because they are AV7 every time; BOBs are only AV7 when they are actually being blocked with claw. Sure, a BOB or two may die, but the large treasuries normally accrued by orcs (and other bash teams) mean that they can be replaced fairly easily. It may take them quite a while to get to 2-3 skills again, but I don't see anything wrong with that considering the relatively (compared to Skaven) rate of turnover. Where does it say that BOBs are entitled to all be skilled at the same time? Skaven die, as you say, and linerats are easily replaced by journeymen, and GRs skill fast and are decent out of the box. In general, skaven can't maintain the same treasury as orcs. The higher player turnover rate will either force you to spend it (on linerats or on positionals when they buy the farm) or force you to use journeymen (and the inherent unreliability they come with) so you can hold a buffer to replace positionals. When facing higher TV teams, skaven face the additional problem that the opposition will likely have a few tackle players, maybe grab, and the ability to effectively foul (although not as effectively as we might like) any GRs which do get knocked down.
Furthermore, if your Orc team does lose a few hundred TV in BOBs from a particularly nasty game, your TV will drop allowing you to purchase BOB mercenaries if you can't afford to buy real replacements. They'll be more expensive than the real thing, but I see no reason why your efficiency shouldn't be affected when you've taken a beating. Skaven's is affected, but probably by less and probably more regularly. Overall, it evens out.
Personally I don't think orcs take as much of a beating as you suggest. I've not looked at the FUMBBL data in the same way, but in FOL Orcs take the second lowest number of cas per game, just behind dwarves. As a ratio of cas caused:taken per game, orcs also come second to dwarves. That's obviously subject to the limitations of FOL, but the data is there. From personal experience I played Orcs in OCC for 8 seasons and lost 2 blitzers and a few lineorcs in the entire 72+ games (some seasons were 13 games). Certainly there was luck involved, but there was also a very high concentration of Chaos teams due to there only being 8 races for the majority of that time (DE were in for the last two seasons, I think). That, combined with looking at the success of other Orc teams in the league, means my own experience shows there is no real issue for them. If anything, the lack of claw teams may be causing an issue in the short term - my own division has no claw teams and the Orcs are the top 3 teams!

Perhaps this does merit an analysis of win/loss for teams with a difference of 150-190 and those 50-140. If the former perform much better then the latter this inducement creates an odd incentive to weaken you team in terms of players or RRs just to get the wizard.

Already did that and it follows the declining win% line. The problem is parsing out those where the wizard is taken and comparing them to those games where he is not. If we could get a raft of games where the wizard was taken by the underdog and compare the win% of the underdog to games where the same TV difference is available but the wizard is not taken by the underdog then we could assess if it is actually making a big difference.

[mattgslater]

In your post on why underdog TV skaven are so good I came away with the impression that one of the key issues was the ability of Skaven to use the threat of a wizard to such good effect - in that it made it harder for bash teams to concentrate their force around the ball carrier, and at the same time were generally in a very good position to exploit the actual use of a fireball.

It does make them very good, yes. The singular Skaven utility for universal inducements is, along with Journeymen, more than a compensation for the sudden devaluing of MA on linemen. However, it doesn't make them too good, just top-tier. My complaint is not that Skaven should be worse, but that they were saved from the major glitch that deterred people from playing them in the same ruleset that cursed certain other teams with the same major glitch.

Although you may not like wizards against speed teams there seem to be a large number of coaches who find the lightning bolt too effective at stopping them.

Sure. But it gets competition at certain points. If by hiring an expensive Star, I have to forgo the Wizard, but the Star brings something that I really want, then I may not take the Wizard. For instance, if my Humans are down 310k against High Elves, I would likely go for Zug+Babe over Wiz+Helmut+Babe, to grab a convincing edge in the blocking game. But certainly, in the 150k-200k range it's a no-brainer.

Regardless if the universal advice is to take a wizard then it probably means its under priced.

Also, don't you think it would still be taken as much if it was 200k?

Down this road lieth imbalance and sorrow. Wizards are good, they're fun, and maybe they're a tad cheap. There is no balance point on unbalancing elements. Remember trying to find a balance on Khemri, then settling for "good enough to be fun" instead of building a truly competitive team? It's not the same balance game (horseshoes instead of blackjack, in that your mythical balance point is the midpoint of your range rather than the high point), but some of the same considerations are there that you'll never hit a ringer from that far away.

[mattgslater]

Wow, Dode. Whack-a-Mole? Really?

I disagree. Maintaining the roughness is in keeping with the game.

We're not "maintaining" anything whatsoever by collecting less useful data. Working blindly is not in keeping with the game.

There are no dead cert winners as underdogs, so I don't see it as being too rough.

Define "dead cert" please. Which of these, if either, do you mean?
a) You can't ever simply assume an underdog has an overwhelming advantage; or
b) You can't infer that one team race or coaching style or whatever gets more from being an underdog than another one gets.

I agree with the first, not with the second. Underdogs are almost never prohibitive favorites, unless there's a huge coaching discrepancy. "Favorites" are seldom prohibitive favorites, either. But some teams just like to get handicaps, while others don't.

Go in too deeply and you overtinker and remove all the variation.

No, that's absurd. You can't "reason out" all variation in Blood Bowl. We're coaches, not supercomputers, and by collecting better data, we'll never approach a sense of collective BB-omniscience, I promise you. Even if we use supercomputers and eventually build a Deep Blood to take on all the world's coaches, it will only ever give Malmir or RO a pretty good fight, and maybe teach us a few new tricks. A more likely upshot is a set of best practices, which will help poor coaches climb the curve, and turn a few average coaches into good ones.

Reference the skaven issue, skaven lose a lot more players in general than orcs because they are AV7 every time; BOBs are only AV7 when they are actually being blocked with claw. Sure, a BOB or two may die, but the large treasuries normally accrued by orcs (and other bash teams) mean that they can be replaced fairly easily. It may take them quite a while to get to 2-3 skills again, but I don't see anything wrong with that considering the relatively (compared to Skaven) rate of turnover. Where does it say that BOBs are entitled to all be skilled at the same time? Skaven die, as you say, and linerats are easily replaced by journeymen, and GRs skill fast and are decent out of the box.

That's my point. A Skaven team can survive any loss whatsoever, with no decrease in its competitiveness, if it has the Treasury to replace dead GRs. This is fortunate, because they must do it frequently. An Orc team cannot. This used to be fortunate, because it was rare. Now, it's not rare; just a relatively ordinary exception to the rules, like a streak of failed KO recoveries, only on a metagame scale. But now it's a problem that applies to the Orc and not the Skaven.

In general, skaven can't maintain the same treasury as orcs. The higher player turnover rate will either force you to spend it (on linerats or on positionals when they buy the farm) or force you to use journeymen (and the inherent unreliability they come with) so you can hold a buffer to replace positionals.

You mean Journeymen to replace linerats, and occasionally a positional. A Journeyrat is a better value for the TV than a rookie Stormvermin, and not that much worse, gold-for-gold, as a badass TCPOMB Stormvermin (the difference in TV is most of a Wizard, or a Bribe/Fezglitch/2xBabe plus 30k, or a DP Merc Lino). Oh, and that badass SV is the only guy on a rat team who clearly justifies his replacement value! (Okay, him and the oh-so-easy Block GR.) Yes, you have to spend money with Skaven more than you do with Orcs, but not generally faster than you make it, except maybe over a game or two.

When facing higher TV teams, skaven face the additional problem that the opposition will likely have a few tackle players, maybe grab, and the ability to effectively foul (although not as effectively as we might like) any GRs which do get knocked down.

Yes, yes, paper-scissors-rock. Let's keep it to what happens between games or over a span of several games: Skaven Win% is not a major concern, either way. These huge advantages just balance them out as big underdogs, which is fine.

Furthermore, if your Orc team does lose a few hundred TV in BOBs from a particularly nasty game, your TV will drop allowing you to purchase BOB mercenaries if you can't afford to buy real replacements. They'll be more expensive than the real thing, but I see no reason why your efficiency shouldn't be affected when you've taken a beating.

You can't take Merc BOBs as a policy and expect ever to get back to competitiveness. They're not a 0-16 position, you want to max out. Merc BOBs don't improve, they eat SPP, they can't get the all-important two-skill stack (B/G), they cause turnovers if you give them Guard, and cost you Guard if you give them Block. You need those BOBs on the roster, even when they're tub-of-lard rookies costing you games, because you have to advance them twice if you want to hit peak ever again. I've taken my share of Merc BOBs, and I love them as fill-ins, but I know better than to just run fewer rostered BOBs, unless I had a MNG BOB, or I was down a BOB and didn't have the cash. Otherwise, no way. Especially since you really only need three skilled ones. If one BOB goes down, you hire another one. But if two go down, you just keep losing until the MVPs fall right, and you're back on track. In a seasonal league, that's a lost season, no matter how good a coach you are.

Skaven's is affected, but probably by less and probably more regularly. Overall, it evens out.

No, it doesn't. Saying it evens out because you refuse to look closely is just wrong and daft.

Personally I don't think orcs take as much of a beating as you suggest. I've not looked at the FUMBBL data in the same way, but in FOL Orcs take the second lowest number of cas per game, just behind dwarves. As a ratio of cas caused:taken per game, orcs also come second to dwarves. That's obviously subject to the limitations of FOL, but the data is there.

This jives with my experience but is not relevant to my point. If you play five games and take 3 Cas, that looks great. If all 3 Cas happen to be dead/gimped BOBs, and your two Apo attempts both failed, that's the end of the road for you.

From personal experience I played Orcs in OCC for 8 seasons and lost 2 blitzers and a few lineorcs in the entire 72+ games (some seasons were 13 games). Certainly there was luck involved, but there was also a very high concentration of Chaos teams due to there only being 8 races for the majority of that time (DE were in for the last two seasons, I think). That, combined with looking at the success of other Orc teams in the league, means my own experience shows there is no real issue for them. If anything, the lack of claw teams may be causing an issue in the short term - my own division has no claw teams and the Orcs are the top 3 teams!

Maybe that's what's going on. Certainly, Orcs were very competitive in LRB4 and before, and while they're riding high in the new rules, they're at least as good. It's just that their odds of crashing, while still low, are much higher than they used to be. I wouldn't be surprised if tweaking the killstack solved this problem, nor would I be surprised if an indirect route worked just as well, or even better. Maybe the solution is allowing coaches to hire Mercs like JMs, just without any skills paid-for in the game. (Amnesia?) Maybe it's in allowing coaches to induce MVPs.

[dode74]

Wow, Dode. Whack-a-Mole? Really?

I don't know what you mean by this. Care to explain?

We're not "maintaining" anything whatsoever by collecting less useful data. Working blindly is not in keeping with the game.

Define "useful". In terms of maintaining the aims of the BBRC (tiers etc) then the aggregated data is more useful than data which is parsed out in a manner which was not intended.

Define "dead cert" please.

Option a. Obviously some teams have a lower TV sweet-spot than others, and I've never said otherwise. Quite the opposite.

No, that's absurd. You can't "reason out" all variation in Blood Bowl. We're coaches, not supercomputers, and by collecting better data, we'll never approach a sense of collective BB-omniscience, I promise you. Even if we use supercomputers and eventually build a Deep Blood to take on all the world's coaches, it will only ever give Malmir or RO a pretty good fight, and maybe teach us a few new tricks. A more likely upshot is a set of best practices, which will help poor coaches climb the curve, and turn a few average coaches into good ones.

Non sequitur. I was talking about delving beyond the aggregate. What you may achieve by that delving is an insight to how a team may best be played on average (the sweet-spot, for example), but breaking it down in order to prove an imbalance redefines balance.

That's my point. A Skaven team can survive any loss whatsoever, with no decrease in its competitiveness, if it has the Treasury to replace dead GRs. This is fortunate, because they must do it frequently. An Orc team cannot.

I don't see that as a problem, possibly because the data and experience I have says that Orcs lose that kind of player rarely enough to deal with it just fine.

You mean Journeymen to replace linerats, and occasionally a positional. A Journeyrat is a better value for the TV than a rookie Stormvermin, and not that much worse, gold-for-gold, as a badass TCPOMB Stormvermin (the difference in TV is most of a Wizard, or a Bribe/Fezglitch/2xBabe plus 30k, or a DP Merc Lino). Oh, and that badass SV is the only guy on a rat team who clearly justifies his replacement value! (Okay, him and the oh-so-easy Block GR.) Yes, you have to spend money with Skaven more than you do with Orcs, but not generally faster than you make it, except maybe over a game or two.

A journeyrat is more TV efficient than a rookie stormvermin? How do you get that?

Yes, yes, paper-scissors-rock. Let's keep it to what happens between games or over a span of several games: Skaven Win% is not a major concern, either way. These huge advantages just balance them out as big underdogs, which is fine.

Why keep it to that span? It's the metagame where things balance, and in the metagame orcs lose few expensive players rarely and skaven lose a lot often.

You can't take Merc BOBs as a policy and expect ever to get back to competitiveness. (snipped the rest)

Who said policy? I said if you can't afford replacements, which is rarely the case as Orcs.

No, it doesn't. Saying it evens out because you refuse to look closely is just wrong and daft.

It evens out in the win%, which is the metagame. Calling it daft doesn't make it so.

Maybe that's what's going on. Certainly, Orcs were very competitive in LRB4 and before, and while they're riding high in the new rules, they're at least as good. It's just that their odds of crashing, while still low, are much higher than they used to be.

Again, I don't see those odds of crashing as "too high". I suspect that's where we differ.

[mattgslater]

Yeah, I think we're going around on three points of departure.

1) You seem to believe that statistical analysis can in itself be vicious. I think that may be true in some aspects of life (too much math brought us nuclear meltdowns and climate change, after all), but not in board games, where the consequences seldom transcend the game and the environment in which it is played. I don't think you can crash a game that way, even, unless it's a very simple one and you figure out a way that always works. Children do that to Tic-Tac-Toe, supercomputers do it to Chess, but Blood Bowl is vastly more complex, operationally, and the data you can mine with any population of BB games whatsoever can only go so far, no matter how many lenses you put on it. That's pretty far, mind you... but it will never dilute the magic of the game, just deepen your understanding of its vastness.

2) I think the numerous incremental increases in the game's deadliness have mostly been resolved by the underdog mechanics, and in fact that the only losers were the teams the LRB6 rules set out to tone down in the first place; I just think they went a step or two too far. It's no longer possible to kill the teams that used to get killed the most. But the teams that used to get killed the least now get killed more, and other teams get killed less, so now they get killed the most, and their coaches are the ones who go through spans of irrelevance, rather than the other one. The point of game design isn't to put the un-fun shoe on the other foot from time to time, it's to make the game fun for everybody, all the time.

3) You put a lot of faith in aggregates. I don't. I think that aggregates are great tools for game design, but like most tools they make lousy masters. You need to look at scenarios, especially when you're happy with the overall product and just want to tinker to make it perfect. (Must ... make ... perfect ... game ... ! Sure, you'll never get there, but that doesn't make it okay to stop trying.) IMO, if somebody enjoys a game of BB, then he should enjoy the leagues he plays in, almost every time. It's very hard to have a lot of fun if your team has been killed. It may not mean much in the aggregate if 5% of Orc teams get wrecked through no fault of anybody's and then can't compete, but that this happens to 0% of elf teams. But for 5% of Orc coaches at any given time, it's really obnoxious. If you can cater to that 5% without getting in the way of the other 95%, then that's what you do.