Why ClawPOMB is broken

[Darkson]

This is why when you and Darkson say things like "no player gets more than 55% with ClawPOMB" it rings alarms.

And so it should, so it's just as well I, and I believe Dode, have never said that.

As you are well aware, the 55% (which was never a specific % anyway) is the win rate for the whole roster across all players, not one specific player. Of course there are players that can/do have a 70% win rate with Chaos, just as there are players that have a less than 50% win rate with Wood Elves - that doesn't make the Wood Elf % less than 50%.

[dode74]

We would measure it by demonstrating that the effect of player skill was significantly reduced by the introduction of that mechanic. This is why when you and Darkson say things like "no player gets more than 55% with ClawPOMB" it rings alarms. Surely there should be some players that have a 70% win rate?

Turn the alarms off, nobody has said that and "some players" do win lots with it. So what measure would you like to move onto now?

The data I used to determine base cas levels (one that I felt would have no ClawPOMB but was still likely to have blodgers) was taken from TT tournaments

So not necessarily relevant to leagues then. Regardless (because we can go around this bouy all month), the cas base rate as calculated is as per design, as Tom has said. It is therefore not relevant that you find it to be too high, because the game designers knew what it would be and accepted it.

Mostly you are saying you just don't accept the conclusions of my analysis.

No, I am saying I don't accept the assumptions you are making. I don't accept your assumption of what "broken" means, I don't accept your assumption of the effect of a CPOMBer on cas rates, and I don't accept your assumption that "position within the T1 band" is of any relevance at all. Maybe it's you who needs to reread.

Why is that relevant?

Because it is what the game is balanced around. Why is position within tier 1 relevant if everything which should be is within tier 1?

If the addition of a skill results in a 50/50 broken mechanic (as per First Blood Chess), but 50% is within your definition of Tier 1 balance then why does that mean it is not broken?

Let me get this clear: you're using the example of a mechanic which gives White the win every time in chess, which is therefore not a 50-50 mechanic, as a "50-50 broken mechanic"? HUGELY flawed analogy.
That said, if a skill is added or taken away, and the 50-50 results remain, then it is not, by definition, broken. It might be unliked, despised, hated, whatever, but it's not by definition broken.

So you accept that CDs have improved by a significant amount? (You can choose what level of uncertainty you want significant to mean). And this is with the inclusion of lower TV CD results in there that would be expected to perform closer to the initial level, dragging the total results back to a lower average?

CDs are very likely better in leagues than in tournaments. That's the conclusion I am willing to draw from the data. They also can't be said (to even 90CI) to be operating outside the tier 1 band, so they are not broken in leagues. You're bringing up frame dragging yet again when I've already stated that tiers were based on lifetime performance?

[Wulfyn]

This is why when you and Darkson say things like "no player gets more than 55% with ClawPOMB" it rings alarms.

And so it should, so it's just as well I, and I believe Dode, have never said that.

From page 1:

As Sann said, clawpomb teams don't win more than they should (given the BBRC's tier definition)

So as I understand it the Tier 1 definition to which CDs belong is 55%. And you are saying that ClawPOMB teams, of which CDs are one, don't win more than they should (55%).

So exactly what have you not said here?

[Wulfyn]

Turn the alarms off, nobody has said that and "some players" do win lots with it. So what measure would you like to move onto now?

Ok so to start with we had teams are not winning a lot with it - which has been your main point for the last 9 pages. And now when I point out that this is alarming you change your reply to say that what you have been saying all along is not true? Seriously?

So not necessarily relevant to leagues then. Regardless (because we can go around this bouy all month), the cas base rate as calculated is as per design, as Tom has said. It is therefore not relevant that you find it to be too high, because the game designers knew what it would be and accepted it.

That doesn't mean that it is not broken. I stated this point in my last post, specifically that just because someone intentionally designed something to act in a certain way does not mean that it is not a broken mechanic.

Are you saying that broken mechanics do not exist because game mechanics are designed?

I don't accept your assumption of what "broken" means, I don't accept your assumption of the effect of a CPOMBer on cas rates, and I don't accept your assumption that "position within the T1 band" is of any relevance at all. Maybe it's you who needs to reread.

You have stated that you don't accept my explanation of what broken means but you have made no attempt to demonstrate why my definition is flawed or what your alternative would be. You are just, so far, saying "I don't accept it". Right... why?

Because it is what the game is balanced around. Why is position within tier 1 relevant if everything which should be is within tier 1?

Because I have shown you multiple times that games with a specific result can still be broken.

Let me get this clear: you're using the example of a mechanic which gives White the win every time in chess, which is therefore not a 50-50 mechanic, as a "50-50 broken mechanic"? HUGELY flawed analogy.

I am glad you have asked for clarity, because I imagine that the point where I wrote the person who plays white is decided by a 50/50 coin toss was lost for you. I know you are struggling with the statistics, so let me be clear.

50% * 100% = 50%.

That said, if a skill is added or taken away, and the 50-50 results remain, then it is not, by definition, broken.

Let us remove all blood bowl rules other than the winner is decided by a coin toss. According to your definition we have not broken the game.

CDs are very likely better in leagues than in tournaments. That's the conclusion I am willing to draw from the data. They also can't be said (to even 90CI) to be operating outside the tier 1 band, so they are not broken in leagues. You're bringing up frame dragging yet again when I've already stated that tiers were based on lifetime performance?

Very likely? What evidence do you have for this? I mean I am perfectly clear that this is the conclusion that you want to draw from it.

And to top this off you are refusing to accept that maybe even (say) 1% of teams within a race that have a fully ClawPOMBed up team could be game breaking because once you blend their results with a ton of other teams from that race that don't have ClawPOMB you get the result you want.

You're not a scientist are you?

[Moraiwe]

This is why when you and Darkson say things like "no player gets more than 55% with ClawPOMB" it rings alarms.

And so it should, so it's just as well I, and I believe Dode, have never said that.

From page 1:

As Sann said, clawpomb teams don't win more than they should (given the BBRC's tier definition)

So as I understand it the Tier 1 definition to which CDs belong is 55%. And you are saying that ClawPOMB teams, of which CDs are one, don't win more than they should (55%).

So exactly what have you not said here?

You've failed to understand the meaning of the quote from page 1. When Darkson said 'clawpomb teams' he was referring to all the data from matches played by teams with clawpomb... not any one specific coach who uses such a team and wins above 55% of his games. The page 1 quote is very different from what your assumption that people have said 'no player gets more than 55% with ClawPOMB' - nobody has.

[dode74]

Ok so to start with we had teams are not winning a lot with it - which has been your main point for the last 9 pages. And now when I point out that this is alarming you change your reply to say that what you have been saying all along is not true? Seriously?

There is - as you well know - a big difference between "nobody wins more than 55% of their games with it" and "the race does not operate outside the tier 1 range". The latter has been my main point for the last 9 pages, the interpretation being as Moriawe has stated.

That doesn't mean that it is not broken.

Doesn't mean it is, either. You have to show it is.

Are you saying that broken mechanics do not exist because game mechanics are designed?

Obviously not. I'm saying that if the criteria of "balance" is met then the mechanic isn't broken even if you don't like it, and for whatever reason.

You have stated that you don't accept my explanation of what broken means but you have made no attempt to demonstrate why my definition is flawed or what your alternative would be. You are just, so far, saying "I don't accept it". Right... why?

Because you have no way to measure "broken" given your definition. Because your definition of "broken" is arbitrary. Because there is already a definition for balance within BB.

Because I have shown you multiple times that games with a specific result can still be broken.

They can, but this is not one of them. Position within a tier is not a specific result.

I imagine that the point where I wrote the person who plays white is decided by a 50/50 coin toss was lost for you.

But the game is still biased to white. You appear to be struggling with game concepts.

Let us remove all blood bowl rules other than the winner is decided by a coin toss. According to your definition we have not broken the game.

Let's make stupid slippery slope arguments about non-existent situations (which will likely be cloaked in the respectable term of "thought experiment" at some point) in order to make our point, eh? Because that sounds like fun!
You appear to be having difficulty understanding the difference between "game concept" and "game balance".

Very likely? What evidence do you have for this?

ANOVA between CD league results and CD tournament results.

And to top this off you are refusing to accept that maybe even (say) 1% of teams within a race that have a fully ClawPOMBed up team could be game breaking because once you blend their results with a ton of other teams from that race that don't have ClawPOMB you get the result you want.

Anything could be, but they are not by the definition of the tiers. If you want to show that they are then you need to demonstrate it.

You're not a scientist are you?

Ironic, given the above.

[adhansa]

I don't get why people are keeping to refer to the 45-55% intervall. That is used to balanse if a team is in the tier 1 bracket, =/= how strong a skillcombo or gamemachanic is. It is just measuring the wrong thing.
And then you take all the teams of a race as the sample. Is there any proof that that sample is the same sample we are trying to evaluate, teams with enough of this skillcombo to have a major influense on the game. But this point only bugs me to a lesser extend, hard to say how big the difference is, my first point is my main one.

What do I we then should be measuring instead? I am not certain, i am not a stastican, what we want to measure is how teams with enough of this skillcombo is doing and how teams that have come up against it develop afterwards in a representative environment. But that is not all the problems that i and people who look at the problem as i do blame on this skillcombo. There is also problem of the games to often becoming decided to early on and the problem of that some teams are not worthwhile playing at certain TV-intervalls because of it. I find it hard to believe that theese two effects are in the spirit or the intention of the game.

To me it is not worthwhile to try to gather this data unless someone has a brilliant idea of how to easally aquire them. Referring to irrelevant data is even worse. And if we don't have any relevant data should we then just asume that everything is hunkydory, no clouds on the sky, go on as usual? I have a little more trust in human logic than that. Even though our judgement are sometimes clouded by emotions, we as a group in general have pretty good capabilaty to see patterns, identify problems and make conclusions on the observations we have made. Therefore it is problem for me try to fix problems identified without any support of concrete metrics, if theese metrics are to problemtic to produce. And i think this is such a case.

Of cause, in a game like BB, if you start changing the rules you will be drawing a one thread without being able to fully estimate what the effects will be on other threads. And you will never come up with "a perfect game". But if you come up with a game that is generally perceived as a better game, that is a good enough goal to make the effort, as long as you stay aware of new problems shoving up in horizon and are ready to parry for them or modify/retract the changes if they are not in total desirable.

[dode74]

That is used to balanse if a team is in the tier 1 bracket, =/= how strong a skillcombo or gamemachanic is. It is just measuring the wrong thing.

If a game mechanic isn't so strong it pushes the lifetime performance outside the T1 bracket then it's not objectively broken.

And then you take all the teams of a race as the sample

Because that's the definition from the BBRC.

Anything else you want to measure, including all the other stuff in your post, will come down to whether you like it or not. Putting numbers to it without a reference frame to say how much is too much or too little is merely a numerical way of expressing a preference.

[Darkson]

You've failed to understand the meaning of the quote from page 1.

I don't think "failing to understand" is the problem, it's more a shining example of "politician talk" where you take one word/sentence and use it to base your entire reply about, ignoring everything else that was said.

[Moraiwe]

I was trying to be polite. (I hope none of my league-mates see this, they'll die of shock)

[VoodooMike]

You have stated that you don't accept my explanation of what broken means but you have made no attempt to demonstrate why my definition is flawed or what your alternative would be. You are just, so far, saying "I don't accept it". Right... why?

Let me lay this out for you since dode seems intent on hitting you intellectual softballs lest you get your sensitive bits bruised... it'll save time.

Your assertion is that CPOMB is "broken" because it creates a situation in which the match will be won by the first person to remove a piece, at which point the teams will take turns removing each other's pieces until one side runs out and loses. This is the foundation of your "it's random" complaint. So, ignoring the fact that it's not true and you have not support for the idea that it's true, lets just take a look at the basic idea.

Your claim, at best, only applies to CPOMB vs CPOMB

The complaint doesn't really explain why non-CPOMB teams quite regularly wipe the floor with CPOMB teams. Based on your theory of piece removal all non-CPOMB teams should lose every match against a CPOMB team. We know they don't. We being the rest of us, not you, apparently.

You're defining "broken" as "I don't find it fun"

Games of pure chance are not inherently broken - in fact, they're pretty widespread. People flip coins to decide things all time time.. kids play "rock, paper, scissors" despite there being no skill involved. Hell, they play "War" which is almost exactly the same thing as flipping a coin 52 times. Certainly they're not nuanced games, but they're not "broken"

You're ignoring the positioning aspect of Blood Bowl

You know which game involves every block removing the defending piece, but allows both players the ability to position their players to make or avoid hits on subsequent turns? Chess. According to your claims that makes chess the ultimate broken game of chance, not skill - chess blocks are even more likely to result in a casualty than the blocks thrown by the worst CPOMB team of all time.

Now, take a brief look at the idea that the ideas apply to Blood Bowl itself.

CPOMB vs CPOMB casualties average far below 11

Even if we skip running the full data (which we both know only dode and I will ever do, you'll just assume it supports your feelsies) we can head over to FUMBBL and check out the match results for well known CPOMB teams when they face other CPOMB teams....

BillBrasky's WMDs in the Box is quite well known among FUMBBLers. If we start at the bottom and work our way up (most recent matches) we find the most recent CPOMB games:

vs. Nurgle 4 cas caused vs. 3 cas taken
vs. Chaos 1 cas caused vs. 4 cas taken
vs. Chaos 4 cas caused vs. 3 cas taken
vs. Chaos 4 cas caused vs. 6 cas taken
vs. Nurgle 5 cas caused vs. 6 cas taken

And these are high TV Box teams... an environment where far, far more games are being played by a single team than you're ever likely to see in any other environment. Brasky's team has played 2,885 (!) games. Many coaches won't play that many games in their lifetime, much less with a single team.

Even high development CPOMB facing high development CPOMB is nowhere near being a coin flip, and it's not guaranteed piece removal like chess. We know this not only by knowing how Blood Bowl is played, but also from looking at actual Blood Bowl being played. There's no situation, theoretical or actual, where what you're suggesting is true.

That's probably why you're several pages into arguing game philosophy rather than supporting your actual assertions (aka mistakes).

It is just measuring the wrong thing.

Is it? What are we supposed to be measuring, and who has done that measuring to date? I mean... plenty of people piss and moan about CPOMB, but if nobody has ever named a metric or measured it then we know right off the bat they're basing their bitching on nothing but "feeling".

The people who wrote the rules used the 45-55% range as their balancing metric, so it is exactly the thing we should be looking at if we're asking if certain skills make a team "too good". If we can't see any significant positive effect in the area the game is balanced around then the relevant response becomes "too good at what?, and why do we care?".

Imagine a professional hockey player who is an incredible juggler. He juggles during games with amazing panache. What we're saying is similar to us saying that he's "too good at juggling". We check and see that his team isn't winning more games then the teams that lack a top tier juggler.. and the juggling player isn't scoring more goals than non-jugglers... and then we ask "who the hell cares how good he is at juggling?" and we possibly tell the concessions stand to stop selling beer to the guy who declares the juggling to be "broken".

And if we don't have any relevant data should we then just asume that everything is hunkydory, no clouds on the sky, go on as usual? I have a little more trust in human logic than that.

No, you have a desperate hope that people will put faith in human intuition, because you're certainly not holding out for genuine logic. Logic says that if there is an obvious, relevant effect then it will be detectable in a way beyond you feeling "a disturbance in the force". If you're using your brain and logic to tell what is out of alignment, isn't right, etc... then it should be very easy to know what to measure. It's when you're feeling your way through with your intuition that you won't know what to measure because you're following a nebulous itch in your brain, not reason.

[plasmoid]

Hi Dode,
not a big deal for the rest of the discussion, but I still disagree a lot with your interpretation of Galaks quote.

It is unwarranted to claim that you have any shirt of any colour other than red, since the statement provided does not provide sufficient information for you to make that claim. Any suggestion that you think that is silly is based on the contextual argument that (in general) nobody only owns red shirts, the context being your experience of the shirts people wear.

I'm glad you agree that language is understood through context.
Human communication does not rely on formal logic alone.
You've managed to poke a hole in my example. Fair enough. I have more. I think the point remains: Any understanding of a statement is contingent on the context in which the statement was made.

For example:
If you asked me: "Say, do you like apples?"
I might say: "Sure. In fact I bought an apple a few minutes ago"
Given the context of the statement, you'd have no good reason to infer whether I had bought something else that day.

But if you asked me: "Have you done any shopping today?"
I might say; "Sure. In fact I bought an apple a few minutes ago"
And in that context it would make more sense to infer, that an apple was indeed my only shopping today.

So, in a similar vein, had someone asked:
"So Galak, how did you and the BBRC define game balance when you made the CRP", then it would be reasonable to infer, that Galak was giving an adequate description of balance in the CRP design process.

But that is not what was asked. The question was "How are the teams tiered and which team is in which tier?"
And Galak stepped in and explained "how the tiers were defined". Yes, he used the word balance.
But nothing in the context indicates that he was intending to give a full account of balance criteria in CRP.

...
I'm not saying that we should act on any other balancing criteria. Obviously. Since we don't know any.
And I (now) get, that any other talk of balancing something, like a skill, would not be an acceptable criteria for balance for you, because no numbers were associated with it. So even if described as balancing, it is just opinions.
So for all intents and purposes, the 45-55% mark is all we have. Imperfectly described as it may be.
Fair enough.

My only point with this rant is that you're taking the quote out of context.
And that the context is not "an exclusive description of CRP balance".
Cheers
Martin

[voyagers_uk]

It doesn't read like a rant Martin

[voyagers_uk]

Oh and Martin, could you add to my thread in General Chat about your tabletop league please

[plasmoid]

Will do.