Why ClawPOMB is broken

Don't understand a particular rule or just need to clarify something? This is the forum for you. With 2 of the BBRC members and the main LRB5/6 writer present at TFF, you're bound to get as good an answer as possible.

Moderator: TFF Mods

Locked
plasmoid
Legend
Legend
Posts: 5334
Joined: Sun May 05, 2002 8:55 am
Location: Copenhagen
Contact:

Re: Why ClawPOMB is broken

Post by plasmoid »

Hi Mike,
No. In fact, you even saying that is a prelude to scumbag dishonesty whereby you use more unfalsifiable shit and then demand it be falsified or accepted as true.
Actually, I don't know whether such a poll would fall out one way or the other.
I figured that if there was not a majority experiencing that CPOMB makes the game less enjoyable, then the poll would be a lovely thing to show to anyone using the "I don't like it" argument against CPOMB.
Genuinely curious.
And not quite sure whether it can't be done. Or just that nobody who can wants to.

Cheers
Martin

Reason: ''
Narrow Tier BB? http://www.plasmoids.dk/bbowl/NTBB.htm
Or just visit http://www.plasmoids.dk instead
plasmoid
Legend
Legend
Posts: 5334
Joined: Sun May 05, 2002 8:55 am
Location: Copenhagen
Contact:

Re: Why ClawPOMB is broken

Post by plasmoid »

Hi Dode,
For Pro Elf teams, for example, a net cas of -2.3 to -1.0 was the 50% mark
Those are some very interesting stats! Might be interesting to see at what point each race drops below the 45% mark, and how much damage a CPOMB team averages against each race, for comparison.
I'm guessing your stats don't take into account how many players were on the starting roster(?). It seems to me that dealing with 2 casualties must be easier when you start the game with 13 players than with 11.
And Wifflebat is talking about leagues. It's also not necessarily true in MM, because your TV drops and you play against a lower TV team.
OK, but Wifflebat aside, the really big data we have are from MM Leagues. Right? And in those data, dropping a really banged up team will prevent losses(?) Did you ever happen to do some stats comparing banged up teams to undamaged teams at the same TV?

Cheers
Martin

Reason: ''
Narrow Tier BB? http://www.plasmoids.dk/bbowl/NTBB.htm
Or just visit http://www.plasmoids.dk instead
User avatar
Wifflebat
Emerging Star
Emerging Star
Posts: 476
Joined: Sun Feb 02, 2014 5:56 pm
Location: Ohio, USA

Re: Why ClawPOMB is broken

Post by Wifflebat »

VoodooMike wrote:You will never get the data you're "if"ing about here... not about league play.. so this is a dead-end tangent. My answer to your "ifs" is that once the data existed then we'd have a different discussion.
Wifflebat wrote:I'm not in opposition to you here; just genuinely interested in how we could mine some data on the "problem." If we really care to...
I don't consider you in opposition, this is just how I talk. "Data mining" is not about collecting data, btw, its about extracting information from data that has already been collected.

My ultimate viewpoint is this: on anything that is subjective preference, the answer is house ruling not altering the base rules.
Well, that's it, then. If I'm the only one this deep in the thread who's discussing this based on Wulfyn's "broken in all formats" premise, then I'm just a fading echo, I guess. As far as I'm concerned, this is the only place I'd care about ClawPOMB being "broken." Perpetual MM is as much about manipulating your schedule and TV as it is about the game being played, and we'd be better off investigating the format. I was indeed assuming that we were allowing for the possibility of sifting through league data for something useful, (including some kind of record of dropped teams).

Reason: ''
I was Puzzlemonkey, but now I'm Wifflebat. Please forward my mail...
User avatar
Wifflebat
Emerging Star
Emerging Star
Posts: 476
Joined: Sun Feb 02, 2014 5:56 pm
Location: Ohio, USA

Re: Why ClawPOMB is broken

Post by Wifflebat »

dode74 wrote:
Wifflebat wrote:If, in such a case, we saw that ClawPomb teams did significantly more TV damage to their opponents than other bash teams (choosing, perhaps, TVs where four ClawPOMB players were possible on such a team--or even only matches which featured a certain number of ClawPomb players), and that other bash teams were more or less on a par with each other, would that be data worth considering?
If that extra damage isn't leading to them winning more matches then is it a problem?
Well, that's the question, isn't it? It doesn't appear that they're winning more games, so if there is a problem, does it lie in the ongoing effects of the injuries they cause? Or is it just that they're a pain in the ass to play?

Reason: ''
I was Puzzlemonkey, but now I'm Wifflebat. Please forward my mail...
dode74
Ex-Cyanide/Focus toadie
Posts: 2565
Joined: Fri Jul 24, 2009 4:55 pm
Location: Near Reading, UK

Re: Why ClawPOMB is broken

Post by dode74 »

plasmoid wrote:Those are some very interesting stats! Might be interesting to see at what point each race drops below the 45% mark, and how much damage a CPOMB team averages against each race, for comparison.
I'm guessing your stats don't take into account how many players were on the starting roster(?). It seems to me that dealing with 2 casualties must be easier when you start the game with 13 players than with 11.
Nope, which is why the conclusion I drew is as far as I am willing to go.
OK, but Wifflebat aside, the really big data we have are from MM Leagues. Right? And in those data, dropping a really banged up team will prevent losses(?)
If that were happening with any regularity then we'd see significant differences between MM and scheduled leagues. And we have lots of data from many sorts of league, not just MM. Certainly enough to make a comparison between them.
Did you ever happen to do some stats comparing banged up teams to undamaged teams at the same TV?
No, and until such analysis is carried out and a difference can be shown to be there any conjecture based on the concept that there is a difference is just that: conjecture.
Wifflebat wrote:if there is a problem, does it lie in the ongoing effects of the injuries they cause? Or is it just that they're a pain in the ass to play?
What "ongoing effect" do you think there might be? If it's not showing up in win percentage drops in the other races then there doesn't seem to be an ongoing problem.
As for the second question, that's purely subjective. It also falls foul of the fact that if it wasn't CPOMB then the next thing would be the "pain in the ass": people don't like losing players, so if it wasn't CPOMB under the spotlight it'd be the next skill or combo.

Reason: ''
User avatar
Wifflebat
Emerging Star
Emerging Star
Posts: 476
Joined: Sun Feb 02, 2014 5:56 pm
Location: Ohio, USA

Re: Why ClawPOMB is broken

Post by Wifflebat »

dode74 wrote:
Wifflebat wrote:if there is a problem, does it lie in the ongoing effects of the injuries they cause? Or is it just that they're a pain in the ass to play?
What "ongoing effect" do you think there might be? If it's not showing up in win percentage drops in the other races then there doesn't seem to be an ongoing problem.
As for the second question, that's purely subjective. It also falls foul of the fact that if it wasn't CPOMB then the next thing would be the "pain in the ass": people don't like losing players, so if it wasn't CPOMB under the spotlight it'd be the next skill or combo.
I meant to suggest that if it's just a pain in the ass, or something that people don't like to play against, then it's decidedly not a problem. There are enough people in Blood Bowl who hate one team/play style/skill or another that it's important to not listen to any of them. 8)

As far as the win percentage being the only metric for overpowered-ness, again, the only other stat that I would look for would be teams damaged so badly that they never recover. That is the only thing that those who think ClawPOMB is broken could possibly have going for them--that teams are so devastated by ClawPOMB that they're no longer viable, regardless of wins or losses. And since they don't seem to be so destroyed that they're not able to win their fair share, that can only be measured in teams that quit, right? But, yeah, unless some evidence of that presents itself, we don't have any facts that suggest a problem...

Reason: ''
I was Puzzlemonkey, but now I'm Wifflebat. Please forward my mail...
Wulfyn
Emerging Star
Emerging Star
Posts: 323
Joined: Mon May 19, 2014 9:33 pm

Re: Why ClawPOMB is broken

Post by Wulfyn »

Darkson wrote:Win % IS the most important factor here, because its what the game is balanced on.
Oh dear no, that's not right at all. Did you not see my examples specifically pointing this out? In the 'first blood' chess variant you can get a perfect 50/50 result based upon randomly deciding who is white and who is black. White wins every time but if the colour selection is random then it is fair. Similarly the shooty army of death comes down to who wins the roll for the first turn. Again this is a 50/50 result. Both of these games have a broken mechanic (feel free to counter this point of you disagree) and both have a 50% win rate. So we know for certain that win% is not a necessary factor (it is sufficient but it is not necessary).

Consider rock paper scissors. Let's say we invented a new mechanic whereby you could not pick rock. With only paper and scissors left you have no chance of winning with paper, so everyone picks scissors and draws. Again a broken mechanic with a fair win %.

Reason: ''
Wulfyn
Emerging Star
Emerging Star
Posts: 323
Joined: Mon May 19, 2014 9:33 pm

Re: Why ClawPOMB is broken

Post by Wulfyn »

dode74 wrote:Again, as I said (and you appear to have ignored): not every block or blitz can be done with your CPOMBer (particularly if he has used PO last turn, limiting his movement and therefore target options), your CPOMBer can be injured himself rendering him ineffective for the remainder of the match, or your opponent may position well denying you the chance to make effective use of your CPOMBer, perhaps by forcing him out of position or into a position where PO would result in a nasty foul. Solid play can counter the ability to use skills as effectively, meaning the simple mathematical increase in effectiveness does not translate to actual effect on the pitch. I'm not denying it's effective as a method of removing players, simply not as effective as the simple maths suggests.
The average 2 cas from my example is not dependent upon the 50 blocks. They were independently selected from different data sources. For the 2 cas I used the NAF TT history which came to 1.93 cas per game average and I rounded up to 2. I did not calculate it based on assumed number of blocks, successful knockdowns, etc. I used actual data for this number.

I then needed to work out what proportion I thought represented the ClawPOMB attacks. This was more subjective. I felt that it was very likely that you would get 16 attacks out of the ClawPOMBer(s). Whilst what you say is true I think it is mitigated by other factors (such as having more than one ClawPOMBer, or it being Skaven Blitzers instead of Chaos Dwarf Blockers, or taking Jump Up, or being left in contact with enemy, etc.). For this reason I felt that 16 was a sensible middle ground. But I needed to know 16 out of what. Again this was subjective based upon my fumbbl stats for more bashy teams getting 40 to 50 blocks per game. I chose the higher of these two numbers to lower the percentage of blocks. This allowed me to split the 2 cas into the group to be extrapolated by the ClawPOMB killstack escalation and the one to keep the same. I also made no allowances for the team being more likely to make further 2dbs and expose vulnerable players through going numbers up.

Overall I think that this presents a reasonable minimum for how effective the killstack difference is. Do feel free to suggest your own numbers and more realistic effects from things like extra 2dbs from the numbers change. Ultimately I don't think it alters the argument at all.

Reason: ''
Wulfyn
Emerging Star
Emerging Star
Posts: 323
Joined: Mon May 19, 2014 9:33 pm

Re: Why ClawPOMB is broken

Post by Wulfyn »

dode74 wrote:Image
OCC (perpetual scheduled league) data for seasons 4-28, over 24k matches at all TVs. Plenty, but not masses, of Chaos (certainly nowhere near MM levels).
So right off the bat I am far more convinced by using scheduled leagues over free choice MM. In the latter people are free to pick their opponent and my bet is that ClawPOMB gets avoided except by other ClawPOMB killer teams. As such we would expect far ore mirror matches and thus an average win rate much closer to 50%; which is unhelpful.

So I look at those stats and I compare them to the TT win rate (https://docs.google.com/spreadsheets/d/ ... =392950278) and we get some interesting differences.

* Chaos Dwarfs improve from 9th to 1st, whilst Dwarfs fall from 8th to 14th.
* Necromantic improve from 10th to 4th.
* High Elves improve from 12th to 3rd.
* Nurgle improves from 18th to 11th.
* Vampires improve from 19th to 8th.
* Chaos improves from 21st to 10th.

So what is going on here? Well it won't slip you by that 4 of those teams have pretty easy access to ClawPOMB. Of the two that do not we have a common theme in Blood Bowl advancement which is that teams with great base stats and poor starter skills tend to be weak, but will overtake others once they get those core skills. That can also be said of Chaos, but not of the other 3 teams.

What is the explanation for why teams like Chaos Dwarfs are now the top team? Dode are you able to provide TV splits for this?

Reason: ''
User avatar
VoodooMike
Emerging Star
Emerging Star
Posts: 434
Joined: Thu Oct 07, 2010 8:03 am

Re: Why ClawPOMB is broken

Post by VoodooMike »

plasmoid wrote:My question is whether they could be having more fun playing the game.
No doubt I'd have more fun playing the game if I was getting a blowjob from a supermodel at the same time. I'm not sure that suggests it's a necessary addition to the game. Additionally, as you wanted to quip about opinion rather than fact, any response given to such a poll would only be opinion about one's own feelings, not fact about them - they'd be guessing at their fun as they'd be theorizing about something they had not experienced.
plasmoid wrote:Calling that fact "stupid beyond belief" doesn't make it any less true.
That's not what I called stupid. You're now throwing random things in front of my statements to deflect them away from what you've actually said.
plasmoid wrote:They're also fairly good at stomping the non-bash teams that don't have a lot of blodge.
Do we have a "stomping" metric I'm unaware of? Certainly they're not "fairly good" at winning matches against the non-bash teams.. unless we're talking about goblins or halflings?
plasmoid wrote:Actually, it was a way of saying that the more you're beat up, the less likely you are to outscore your opponent.
That's only true if the method of matching two teams for a game doesn't take it into account, hence the "plasmoid's imaginary matching system". As I said, there is no known environment in which that hurt elf team ends up playing against the chaos team again. If we get to manipulate the matching system then hey, we can make it support any point we want!
plasmoid wrote:That's opinion, not fact. For the record, I disagree that we'd have the same discussion.
Heh, it's entertaining to see the guy who is sitting here masturbating about using opinions to support his position suddenly think only objective facts matter. Hopefully that fence you rock ain't topped with those pokey bits.

There will always be differential outcomes between different rosters. That's an intentional part of the design, and it doesn't seem to be an issue for people outside of perpetual open play environments. That this particular differential outcome is important while others is not seems questionable.
plasmoid wrote:The apoth and the game winnings define what is a tolerable level of attrition. If you consistently suffer more than (roughly) 1,5 lost player (pre-apoth), then you'll be in a perpetual downward spiral. Less than that, and you can cope.
And have you used the data to calculate which rosters do or do not cross this metric of yours, or are you planning to run an opinion poll instead? To claim this is your metric is deeply disingenuous if you don't already have that information handy as it means it is something you thought up after deciding what your position is, rather than it being what your position is founded on.

As it happens, the effect you're talking about will still be environment-dependent, heavily affected by racial demographics. Even if a CPOMB team results in a net loss to your treasury that is only going to be a problem for you if you play more games against teams that produce that net loss than you do against teams that produce a net gain that outweighs those losses.
Wifflebat wrote:If I'm the only one this deep in the thread who's discussing this based on Wulfyn's "broken in all formats" premise, then I'm just a fading echo, I guess.
You're not, though. You've specifically stated you're only talking about leagues. We're all discussing the concept of "broken in all formats" but I'm not sure there's more to be said than the fact that all data that exists fails to show that it is broken in any format outside completely subjective views.

Reason: ''
Image
Oventa
Veteran
Veteran
Posts: 246
Joined: Tue Apr 01, 2008 7:58 am
Location: Munich, Germany

Re: Why ClawPOMB is broken

Post by Oventa »

Ok, so probably regretting to post in this thread, but feeling reckless today ...

I would like to highlight that I feel not all are on the same topic.

Darkson, Mike, dode, argument that clawmbpo is not broken,as it is not visible from win statistics.
And I agree to that, and wulfyn probably might as well.
That is not his topic, as I understand him.

I read from his comments, that his focus is the variance related due to clawmbpo, which He calls the "fun".
Something that straume picked up in an earlier post, but that did not get much attention.

the theory (as I understand it) is:
You don't automatically win more often, but it is less skill related, but even more based on dice luck than bloodbowl already is.
So if I am a great player and play against a not so great clawmbpo player, I still probably win more games than I loose, but those that I loose I probably feel helpless, as those are the games on which I loose 3 players by turn 2 and from there on get bashed upon.
Now one can argue it is same with playing against wood elves.
Even a bad player who never rolls a one with elves will be tough and sometimes you also just watch elves action.
The difference to clawmbpo is, that you at least have 11 players to move around and hope for a 1 and feel like you are participating at least a bit in this game. In a bash game where you are 4 players down and have no ag4 player, hope is not really in reach ...

Now is this subjective right now? Yes
Do we have any proof / statistics on this? No
It is just a theory right now.
And justifyications against or for it can not be made based on win statistics because it does not matter for the theory. That is my key point I wanted to add to this thread.

To proof this theory we would need numbers on inflicted and caused CAS and change of TV in teams and coaching skill data.
So I agree to wulfyn theory, and "feel" the same, but don't think any data right now presented proofs it or against it.

Cheers

Reason: ''
Image
dode74
Ex-Cyanide/Focus toadie
Posts: 2565
Joined: Fri Jul 24, 2009 4:55 pm
Location: Near Reading, UK

Re: Why ClawPOMB is broken

Post by dode74 »

Wulfyn -
For the 2 cas I used the NAF TT history which came to 1.93 cas per game average and I rounded up to 2. I did not calculate it based on assumed number of blocks, successful knockdowns, etc. I used actual data for this number.
So the assumption is that all the cas come from blocks? Do you have the source for this data? I'm not suggesting 2 cas per game is an unreasonable average, btw. I'm saying that "add a CPOMBer" does not automagically increase that number to 4.

As for the "comparison" of positions on your win rate chart, are those positions on the NAF table based on inferences of the various races' performances on TT, or are they observational statistics on their actual performances? Because if the latter then you making inferences about them seems methodologically off. We can't actually say CDs are "first" in terms of league data because the sample size means there's no statistically significant difference between their performance and that of, say, Lizards. What might be a valid comparison to assess whether a team is "better" in TT than in league is to look at the win rates of those individual races in TT and in leagues. I don't doubt teams like Chaos and Nurgle (traditional slow-starters) would improve their win percentages, but does that matter? I would say that as long as their win percentage remains within the Tier 1 boundary in leagues then no, it does not.

Reason: ''
straume
Emerging Star
Emerging Star
Posts: 364
Joined: Fri Mar 28, 2014 9:21 am

Re: Why ClawPOMB is broken

Post by straume »

@Oventa: Yes! This!

Absolutely subjective, but my personal opinion is that anyone who understands the game understands this. You create a different game where positioning matters less, and using your killer piece every turn to perform what would be less optimal choices (without the kill stack) is the "better" way to play. Less fun!

Reason: ''
dode74
Ex-Cyanide/Focus toadie
Posts: 2565
Joined: Fri Jul 24, 2009 4:55 pm
Location: Near Reading, UK

Re: Why ClawPOMB is broken

Post by dode74 »

straume wrote:You create a different game where positioning matters less
Interesting. What is the point of positioning if not to increase the odds of something you want to happen happening (usually scoring)? You're not arguing against CPOMB per se here, but against player removal.
perform what would be less optimal choices (without the kill stack) is the "better" way to play
You could replace "kill stack" with any skill or combination of skills in that comment. Obviously picking the best odds to do something is a large part of the game.

Reason: ''
straume
Emerging Star
Emerging Star
Posts: 364
Joined: Fri Mar 28, 2014 9:21 am

Re: Why ClawPOMB is broken

Post by straume »

dode74 wrote:You could replace "kill stack" with any skill or combination of skills in that comment. Obviously picking the best odds to do something is a large part of the game.
Valid point, there. I guess what I am saying is that the kill stack draws you to focusing more on player removal than actual positioning. Use 4-6 turns to kill stuff, and then go and score instead of trying for a steady advance, switch sides, battle for every yard gained up the pitch, screen up and all that fun positioning stuff. Now is this a problem? I guess not (and the win rates suggest it isn`t), I simply argue it makes for a less interesting game than it could be (as per Wulfyn "shooty killy army death" analogy).

Reason: ''
Locked