Improvement table awarding a given value of improvements.

[Ghost]

Dode. I think lack of variation is being undervalued by you. It has a direct bearing on win%, which you don't take into consideration.

You also have to explain something to me, on that graph of yours. If I read it correct, very few teams have a win% above 50. And none above 55. It can not be true that almost all the teams are well below 50 (many even below 40). All games added together in a big pool, should always come out as 50 % for all the teams combined. Or are you telling me, that the tier 3 teams that are left out, are the ones who make the math come together, pulling win%s above 60?

About not being from LRB 4, I see the answer to that myself, realising Slann is there.

[dode74]

Ghost -

I care about upper TV. Because that is where the lack of variation is. And it happens for a reason. Not because the other teams (non-bashers) can't go there, but because they make sure not to go there - and that happens for a reason too.

And it is only possible in a TV-based matchmaker. That's why we see it happen. Of course people will stay out of that range if they can win enough at lower TVs and maintain team. It's hardly the fault of the rules when a house rule - matching by TV - allows them to do so. Remove that artificiality and people will have to try to move into the higher TVs to maintain competitiveness with teams they can no longer avoid.

You know my opinion as to why, and the Sprinttables back it up - thus that table holds a lot of value.

Affirmation of the consequent - I think a, this thing b backs that up despite its acknowledged flaws, therefore b is valuable. I's a form of circular reasoning.

In a perpetual league, where you can't manipulate your TV to influence who you will meet, you can't avoid the TV 1800 + teams. You can stay under TV 1800 yourself, if you want, but inducements are overpriced as part of an otherwise wellworking mechanic. What happens above TV 1800 is very important, as you will not be able to avoid it in a normal setting.

Exactly right, which is why matchmaking by TV, and not the rules, are the problem!

Just to be clear. It is only winnings after having played 30 games? The results of the first 30 games are not included? And they are not from LRB 4, right?

First 30 are not included. They are from FUMBBL after the move to CRP, so there are probably some old LRB 4 constructed teams in there, but the rules are CRP.

Dode. I think lack of variation is being undervalued by you. It has a direct bearing on win%, which you don't take into consideration.

Not at all. We don't even know that there is a lack of variation yet.
On the graph the win% was "normalised" so it would fit into the scale. WE have a win% of 66%. I just spotted that it says win%/4 - that should be win%/5

garion -

The stats are rubbish it obvious to anyone.

Not at all. You only say they are rubbish because you don't agree with them. If you have an actual reason for it then say so. Claiming that "it's the noobs" dragging down the win% is pure speculation - feel free to prove it.
As for your stats, you get them I'll happily parse them if you have them. The stuff I have doesn't have coach ratings at all in it.
Taking "good coaches" will reduce the sample size (and therefore the accuracy) and artificially inflate the win%.

Because at high TVs CPOMB will be playing CPOMb teams 8 times out of 10. Also elves etc... will be playing most of their games below the 1800 maybe 1600 mark because every time they go over they usually only last a handful of games before they get smashed back down below the 1800 or 1600 mark

Which is only possible in a TV-based MM environment. There is a theme here...

No we decided looking at polls is pretty redundant. talking to people that know about the game and have played through previous editions, there is a general feeling that this combo is causing big problems, hence the huge threads blogs etc... that pop up everyweek moaning about this topic, the fact that this thread has even moved in that direction as well like so many others before it speaks volumens too.

Squeaky wheels are merely squeaky, not necessarily correct. Just because people moan about something doesn't make them right, and people tend to moan about stuff they feel is bad and stay quiet on stuff they think is fine as it is. As I said before: "I value people's opinions as well, and I am not dismissing them. I am saying that opinion is often coloured by expectation, and data can show surprising variance from that expectation. As for "wrong with team development", wrong is not the same as different. Team development may be different, but that doesn't make it wrong."

It is not necessarily that it is too powerful, but more that it has damaged the game.

So it's not too powerful now? Make your mind up
We don't yet know if it has damaged the game as we're yet to get the data on skill distribution. Unless you are claiming that the TV split is due to CPOMB and not due to people taking advantage of the exploit of staying at lower TVs to avoid killer teams which TV-based MM causes? Because if that were the case we would see the same thing happening in environments where there is no TV-based MM, such as leagues, and we don't.

[Ghost]

I was waiting for you to respond with "it should have been a 5, not a 4". That was what I thought too. Good, that is an honest excuse, from an honest man

But that is all good. As now the graph makes sense and shows what I have said all along. The win% of the bashers levitate towards 50.

It is no surpise at all, as they make up for the vast majority of games played. Even more so at the upper TVs, where they meet each other over and over (and in mirrormatches). Of course then don't run away with win% of 60 or more. Added to that, that bash becomes about rolling dice, whereas finesse teams are about rolling as few dice as possible (thus finesseteams is dependent on having a good coach, that make full use of positioning and aspects without dicerolling), and clearly you can see why the natural outcome moves towards 50. Just like the graph shows.

The only thing the graph shows of value is, that there is lack of variation - the vast majority of games are played by bashers, exactly the problem many of us has. The win% you ca throw to the garbage bin, it holds no value as it is extracted from TV matchmaking.

You are undervaluing the impact of lack of variation. It has a direct bearing on win%, for the reasons I have just given you.

[garion]

Lol Dodge, I cba to go through all of this again with you. I wish you would stay out of threads that are trying to think of possible solutions to the problem we ALL see but you do not. That was the purpose of the thread, not to talk about stats and how you think they hold all the answers. We have all posted our concerns with the rule set. Stats, data etc.. I couldnt give a monkeys about. The feeling amongst all the people in this thread (everyone in this thread but you I believe) is there are problems that we would like to try and address, the problems all revolve around similar themes. If you don't think there are any problems with the rules then why post in these threads at all? It is certainly not helpful to the discussion to have some one posting "there isnt even a problem?" after everybodies post. That's obviosuly your opinion and that is fine, your entitled to it just as much as anyone else is to think the opposite, so stop arguing over opinion! No amount of data will ever be able to conclusively prove or disprove any of this, there are just far far too many variables. Did anyone prove Fang was broken back in the day? No but it was changed. did anyone prove fouling was broken prior to LRB4? No but it was changed. Did anyone prove anything as absolute Fact before and after any of the rule changes from 3rd ed right the way through the LRB's to now? No they didn't they looked at public opinion from people respected in the game. Top coaches, people that have extensive knowledge. So I really do not get where you are coming from, nor will I ever with this unhealthy obsession with statistical anlysis from the current edition. Rule change has never been done that way before so why choose to do it now?

[Ghost]

I must say that Garion is right here. I puzzles me, why you want to make our opinion wrong, when yours is just as much of an opinion.

[dode74]

Ghost -
Mirror-matches were removed from this data to prevent that. I did forget to mention that before and apologise for it, but Koadah, who put the original stats together, may be able to verify if you doubt me.

the vast majority of games are played by bashers, exactly the problem many of us has.

If that is the only thing of value (and I disagree but will take the point on for the purpose of furthering this discussion) the question, therefore, is "why"? Why is this the case in TV-based MM and not anywhere else?

You are undervaluing the impact of lack of variation. It has a direct bearing on win%, for the reasons I have just given you.

As I said, mirror-matches are removed. If Nurgle can play Chaos lots and they are at about 50% then it's all good. Same for any bash team with any other bash team, or indeed any team vs any team.

garion -

I wish you would stay out of threads that are trying to think of possible solutions to the problem we ALL see but you do not.

I'm sure you would: that way you'd get to be right without having to show it - all very nice and pleasant and easy for you with pats on the back all round after you've finished agreeing with each other
Dissenting voices are surely a good thing? If you can't argue your point then why do you hold it? If you could actually show that there was a problem then you'd have a point, but you can't: you give your opinion, claim that there are lots of other people who hold that opinion and, when shown polls otherwise dismiss the validity of polls (quite rightly), only then to go on and claim that the people who hold your opinion are of better quality in some manner (experience, coach ability) and therefore must be right in some misguided appeal to authority.
Besides that, I've not said "there isn't a problem". I've asked what the problem is and have so far received one definition that may yet hold up to scrutiny - that of a lack of variation in skill selection. Once we have that data then we can check it to see if the hypothesis holds up.

Why do I post in these threads? Because you are proposing changes to a game which I play. I have every right to defend my point of view, just as much as you do. If you want it as a house rule then fill your boots, but otherwise you are proposing things which will have a direct effect on me, so I will counter them if I feel the need. Convince me that there is a problem outside of house ruled leagues and I'll happily change my mind.

So I really do not get where you are coming from, nor will I ever with this unhealthy obsession with statistical anlysis from the current edition.

The game is based on dice rolls. The effect of rules on the game can be measured statistically and directly adjusted by adjusting the probabilities accordingly. There are other factors than adjusting probabilities - changing incentives (such as skill costs, in a vague reference to the OP ) works just as well, and certain house rules (such as TV-based MM) change the incentives themselves.

Rule change has never been done that way before so why choose to do it now?

I believe that the tools weren't in place to do it easily before. They are now, and the data is easily collected and manipulated due to the large number of online games taking place. Aeroplanes weren't designed with wind tunnels and computers back in the 30s, but they still came up with the Spitfire. Add the wind tunnels and the computers and you have better, faster, heavier and longer range aeroplanes. That doesn't mean we have to use them, but we can if it makes things better. It's called progress

[Ghost]

Ofcourse bash playing bash all over will levitate the win% towards 50. That is not a surprise to you, is it? It is what happens. Look at the amount of games played. It does not matter that mirrormatches are stripped out, really. Bash bash, and that is what bash does. At upper TVs, there is lack of variation (as shown by your graph) and bash dominate (as shown by Sprinttables). If things were fine, there would be no problems with variation in MMs. Yet there is. How you can go from saying, that with problems in MMs there is no problems with the rules, I don't understand. So Chaos have only won once in OCC? Fine, how many coaches have played Chaos for long enough to make it into div 1 in OCC? I read a prominent member saying, that in Francobowl there was huge problems with Bash, and he welcomed initiatives for ruleschange. In TT it is becoming an issue too, apparantly.

Why is it bash all over? Because they dominate above TV 1800.

I really don't get, why you feel so strongly for being the only one disagreeing to what is obvious to everyone else. I think it is something about personal pride, more than it is about being constructive on the ideas given - which has pulled all others to the thread. For some reason you have a big stake in proving to us all, that everything is fine - and I don't get why.

[garion]

Just leave is Ghost, Dode never ever stops arguing I have presented him with evidence from all the sprints and the Major cups on fumbbl that bash teams have won with ease, as well as numerous team examples with ludicrously high win percentages, I have also shown him many other polls that support peoples feelings that there is a problem in many other threads that have gone the way of this one (down the drain) it all just gets ignored, the thread has been ruined now anyway by this ocd for arguing against anyone that thinks there is a problem.

[Ghost]

Yeah well. I have this slight feeling, that I find myself wasting my time on arguing things, which was not the point I found this thread interesting. I think I have better things to spend my time on, to be honest. But it is a shame, as I think some interesting ideas was made here, and I would really like to see the game improve further.

[garion]

+1 lets at least get this thread back on track by posting the OP's suggestion.

Okay, I hope this appears as simple written down as it seems in my head. The basic idea is that when rolling for an Improvement you would get to choose skills equal to the (current) value of that Improvement instead of a normal skill. So for example, if you rolled double 6 (worth 50K) you could choose a 'normal' skill (20K) and a 'doubles' skill (30K) instead. Alongside this some normal skills would be changed in price, ranging between 10K-30K.

I'll try to address some of the issues this system might bring up later in the post, but for now let's spell out exactly what the table I'm imagining might look like:

*IMPROVEMENT TABLE*
2-9 = A single New skill of any value or new skills with value no greater than 20K.
10 = Increase the player's MA or AV by 1 point or a single New skill of any value or new skills with value no greater than 30K.
11 = Increase the player's AG by 1 point or New skill(s) with combined value no greater than 40K.
12 = Increase the player's ST by 1 point or New skill(s) with combined value no greater than 50K.

Note : the new skill(s) a player takes must always come from the categories he has normal access to unless the Improvement roll is a doubles.

Alongside this some skills would change in the value they add to a player's TV, summarized below:

10,000gc = Fend, Kickoff Return, Pass Block, Pro, Shadowing, Catch, Sprint, Sure Feet, Diving Catch, Hail Mary Pass, Nerves of Steel, Dump Off, Grab, Thick Skull, Strong Arm, Extra Arms, Disturbing Presence, Prehensile Tail, Very Long Legs, Horns, Big Hand.

20,000gc = everything not mentioned elsewhere

30,000gc = Block, Dodge, Leader, Guard, Piling On, Claw

Skills players do not have Normal access to and have to be gained by rolling doubles cost +10,000 instead of the flat 30,000.

As you can see, lots of things go into the 10K list, as this will encourage people to take two good skills instead of the same good (but boring) one or two.

For example, an Human Lineman rolls a Normal Improvement. He could take :

Block, adding 30K to his TV.
Wrestle, adding 20K to his TV.
Fend and Pro, adding 20K to his TV.

To me this perfectly logical. The Block guy is the better player, and the Wrestle guy is probably still better than the Fend/Pro guy. But this gives coaches the option to mess around more without being punished quite so harshly. At the moment there really is a right and a wrong way to build teams, and poor old skills like Pass Block don't get much of a look in.

Note that this doesn't increase the frequency of doubles skill access, and doesn't change how likely a player is to get any given skill. I don't like the stick of traits, that stops you taking certain skills. I much prefer a carrot system like this, that rewards coaches who try to build outside the cornerstone skills such as Block/Dodge/Guard. It also has a built in mechanism for punishing abuse of ClawPOMB. Note that some of the 10K skills are in that list because the only players who want them require a doubles to get them. E.g. NoS is a skill desired by catchers, and isn't worth 30K (the same as Block) IMO. I'd also say stunties don't have to pay the 10K doubles penalty. 40K for a Block Goblin ain't right.

Thoughts?

[dode74]

Ghost -

Ofcourse bash playing bash all over will levitate the win% towards 50.

That would only be the case if you were to aggregate the bash teams data. It isn't aggregated, it is separated by race. As such we can safely say that these teams are balanced well against each other when developed.

At upper TVs, there is lack of variation (as shown by your graph)

My graph doesn't remove teams by TV, only by games played.

and bash dominate (as shown by Sprinttables).

Conflating the issue. Sprint is in TV-based MM and rewards teams which are resilient. It is a house rule and alters both the incentives (to play a team which is able to stay at high TV) and the rewards (rewards for a combination of survival and winning the game).

If things were fine, there would be no problems with variation in MMs. Yet there is. How you can go from saying, that with problems in MMs there is no problems with the rules, I don't understand.

MM is a house rule. It rewards the teams which are most efficient at the highest TVs. Remove the house rule and play the game as intended and we'll see.

So Chaos have only won once in OCC? Fine, how many coaches have played Chaos for long enough to make it into div 1 in OCC?

Quite a few, actually. Roughly half of the teams which have been around for the full 11 seasons are Chaos.

Why is it bash all over? Because they dominate above TV 1800.

But not when the teams are developed but may be below 1800TV, which is what happens in leagues and is what my graph is showing.

I really don't get, why you feel so strongly for being the only one disagreeing to what is obvious to everyone else. I think it is something about personal pride, more than it is about being constructive on the ideas given - which has pulled all others to the thread. For some reason you have a big stake in proving to us all, that everything is fine - and I don't get why.

Attacking my motives instead of my argument now? Why is that? Possibly because you have nothing on my argument?
Seriously, this isn't personal. I see a whole bunch of people saying they see "problems" and saying that it's to do with cpomb, and when I ask then to show how that problem is manifesting it all points towards TV-based MM, which is not the environment for which the rules were intended -> viewtopic.php?p=583648#p583648

But as I said with Cyanide's MM ... the rules were not meant to be used on leagues that did only TV matchings for match-ups. Its not organic and it ignores that certain inducements were added in to help weed down those higher TV teams.

garion -
Personal attacks now? Really? I thought you better than that.

I'm awfully sorry that I don't agree with you because you don't have the data to back up what you are saying, or you are presenting skewed data (sprints, MM) but either not admitting or not recognising it, and I do apologise for not taking at face value the opinion of some very good players who may simply need to adjust to the new ruleset (as they have had to before, no doubt), but the fact of the matter is that I have a dog in this fight. I play this game, and so do you. If you want to change the rules then you are welcome to say so, but I have every right to tell you that I don't like either your proposal or your reasons. I get that you don't like being criticised in this - paradigm shifts can be difficult - but, like I said, show some data that bash is too much and I will happily back down. Otherwise all I see is you tinkering to change a game that I enjoy for no reason other than your own preference, which is just damned selfish. If you want to then house rule it, otherwise you need to be prepared and able to back up your position rather than retreating into appeals to authority, personal attacks, and nebulous claims.

Now if you want to discuss this as a house rule then I will butt out and suggest that you request that it be moved to the house rules forum to prevent confusion. If you are suggesting a change to the base rules then I will continue to defend my position, as is my wont and my right.

[garion]

sigh... lets at least get this thread back on track by posting the OP's suggestion. Again!

Okay, I hope this appears as simple written down as it seems in my head. The basic idea is that when rolling for an Improvement you would get to choose skills equal to the (current) value of that Improvement instead of a normal skill. So for example, if you rolled double 6 (worth 50K) you could choose a 'normal' skill (20K) and a 'doubles' skill (30K) instead. Alongside this some normal skills would be changed in price, ranging between 10K-30K.

I'll try to address some of the issues this system might bring up later in the post, but for now let's spell out exactly what the table I'm imagining might look like:

*IMPROVEMENT TABLE*
2-9 = A single New skill of any value or new skills with value no greater than 20K.
10 = Increase the player's MA or AV by 1 point or a single New skill of any value or new skills with value no greater than 30K.
11 = Increase the player's AG by 1 point or New skill(s) with combined value no greater than 40K.
12 = Increase the player's ST by 1 point or New skill(s) with combined value no greater than 50K.

Note : the new skill(s) a player takes must always come from the categories he has normal access to unless the Improvement roll is a doubles.

Alongside this some skills would change in the value they add to a player's TV, summarized below:

10,000gc = Fend, Kickoff Return, Pass Block, Pro, Shadowing, Catch, Sprint, Sure Feet, Diving Catch, Hail Mary Pass, Nerves of Steel, Dump Off, Grab, Thick Skull, Strong Arm, Extra Arms, Disturbing Presence, Prehensile Tail, Very Long Legs, Horns, Big Hand.

20,000gc = everything not mentioned elsewhere

30,000gc = Block, Dodge, Leader, Guard, Piling On, Claw

Skills players do not have Normal access to and have to be gained by rolling doubles cost +10,000 instead of the flat 30,000.

As you can see, lots of things go into the 10K list, as this will encourage people to take two good skills instead of the same good (but boring) one or two.

For example, an Human Lineman rolls a Normal Improvement. He could take :

Block, adding 30K to his TV.
Wrestle, adding 20K to his TV.
Fend and Pro, adding 20K to his TV.

To me this perfectly logical. The Block guy is the better player, and the Wrestle guy is probably still better than the Fend/Pro guy. But this gives coaches the option to mess around more without being punished quite so harshly. At the moment there really is a right and a wrong way to build teams, and poor old skills like Pass Block don't get much of a look in.

Note that this doesn't increase the frequency of doubles skill access, and doesn't change how likely a player is to get any given skill. I don't like the stick of traits, that stops you taking certain skills. I much prefer a carrot system like this, that rewards coaches who try to build outside the cornerstone skills such as Block/Dodge/Guard. It also has a built in mechanism for punishing abuse of ClawPOMB. Note that some of the 10K skills are in that list because the only players who want them require a doubles to get them. E.g. NoS is a skill desired by catchers, and isn't worth 30K (the same as Block) IMO. I'd also say stunties don't have to pay the 10K doubles penalty. 40K for a Block Goblin ain't right.

Thoughts?

[dode74]

Is this a suggested house rule or change to the base rules?

[Rhyoth]

Regarding Joe's rule, i think it's a fine and simple rule, although i would make some modifications :
_ i wouldn't allow more than 2 skills/roll
_ allow a player to take 30k of skills on any double roll (20k normel skill + 10k romal skill OR 10k double skill + 10k double)
_ can't say i agree on this skill tiers. IMHO, you should at least :
a) include Mighty Blow to the 30k category
b) exclude Sure Feet and Dump-off from the 10 k tier (Sure Feet + Sprint and Dump-off+NoS both worth more than 20k), as well as Fend (would make a Fend "spam" too effective)



Now, i'd like to add my 2 cents to the ongoing debate. I think we forgot the root of the problem here : some skills are simply more valuable than ohers ! (yeah it looks like a bold statement, but just try to prove me wrong )
Team building being too linear and boring are only the symptoms, not the disease. (i would also argue this issue is much older than LRB 4 or CRP, by the way, but that's not the point).
That's why i like Joe's idea to make valuable skills more expensive.

Now, how do we measure skills' value, and more precisely is how to identify those "valuable " skills ?

Although it's a quite comlpicated problem, i'll try a simple approach, with these 3 questions :
_ How good is the skill on its own ? (or, if you prefer, how useful a player would be if he had only this one skill ?)
_ How good is the skill when spammed all over a team ?
_ How good is the skill when combined with other skills ?
(That would define 3 values : a "Basic Value", a "Spamming Value", and "Synergic Value" ; personnally, i strongly suspect the later two are underestimated in the ruleset)

Now, for which skills would you answer "very good" on those 3 questions at the same time ?
Personnally, i can only think of a few skills :
Block, Guard, Dodge, and, to a lesser extent, Migthy Blow.

If you want a (partial) statistical confirmation, try to look at :
_ most frequent picks for first skill, sorted out by skill category
_ most frequent picks for first and second skill combined, sorted out by skill category
_ most frequent picks on double roll, sorted out by skill category

(a more rigouros test is possible : see how well team perform if they don't take any of those 4 skills as improvements)

Now, if those numbers don't prove me wrong, it's pretty safe to assume those 4 skills are more valuable* than others and thus should be taxed ; maybe a 10k raise would be too much, but then, we would still have favored the diversification of skill picks, so would it be so bad ?



So, if some skills should be more expensive, those 4 should probably be the first to be raised, but is there any other skill that deserves it as welll ? And more importantly, how to identify them ? Well i don't really know that, but here are some random thoughts about it :
_ if a skill is very effective on its own, but has almost no added value when spammed (ex : Leader, or Accurate), if you add them to the "30k tier" as a test, even if you're wrong, the damage will be very limited...
_ depending on the environment, you could easily add/remove some skills to the "30k pool" (ex: Piling On, or even Claw, in MM)
_Stand Firm and Side Step are both very good when spammed and combined with Block/Dodge/Guard, but i wouldn't increase their value for 2 reasons 1) they are slightly less effective on their own, 2) we already raised the other part of the "Combo"

[the.tok]

some skills are simply more valuable than ohers ! (yeah it looks like a bold statement, but just try to prove me wrong )
Team building being too linear and boring are only the symptoms, not the disease. (i would also argue this issue is much older than LRB 4 or CRP, by the way, but that's not the point).
That's why i like Joe's idea to make valuable skills more expensive.

Buffing the weaker skills could also be done to achieve a similar purpose though that will be difficult to get them to "block" level
could stack with original topic too, if some skills seem to be between 10k and 20k category

Also, I think a "2 skills max per improvement roll" could be good, avoiding the double 6 roll becoming 5 skills at once, which seems a bit too much to me, not really sure

for example
on a 6spp beastman : roll 6+6 : strong arm + fend + extra arms + pass
16 spp : normal roll : kick off return + big hand

he's already a pretty good player, and can get block next

do you guys think it is better or worse than 4ST + block?

the good point is I really had to think hard selecting the skills, so it is already appealing in a sense