43 coaches online • Server time: 17:28
Forum Chat
Log in
Recent Forum Topics goto Post Creating a custom to...goto Post Secret League Americ...goto Post DOTP Season 4
SearchSearch 
Post new topic   Reply to topic
View previous topic Log in to check your private messages View next topic
Endzone



Joined: Apr 01, 2008

Post   Posted: Oct 20, 2017 - 03:23 Reply with quote Back to top

Bonus CR for extra TDs sounds like nonsense and would probably lead to more concessions / 'disconnections' from coaches who didn't want to suffer the big defeat. Crediting a high winning margin just promotes bad play.
The_Murker



Joined: Jan 30, 2011

Post   Posted: Oct 20, 2017 - 03:40 Reply with quote Back to top

"Crediting a high winning margin just promotes bad play.."

.. in people activly trying to manage CR.

_________________
Image
Join the wait-list. Watch the action. Leave the Empire. Come to Bretonnia!
Arktoris



Joined: Feb 16, 2004

Post   Posted: Oct 20, 2017 - 03:51 Reply with quote Back to top

thoralf wrote:

If all you claim is that winning bigly shows greated skill overall than grinding wins, then yes. We agree insofar as we're considering cases where the number of wins is the same.

Pushing for big wins all the time as the most rewarding strategy does not seem to reflect what the top coaches do.

You can argue for any intangible you want, what CR can't capture is beyond the topic of this thread.


Yes, we agree that winning decisively vs. barely means the decisive winner has more skill...provided the wins are the same. Consistency trumps bigly.

Intangibles *can* and often are a reflection of CR. Remember that movie "Thor"? One of the take home messages was "All magic is, is science yet understood."

The intangibles are simply the parameters of the equation that are far too complex for a tiny human mind to derive and accurately assess. That's why we can't grasp them...but we see their affect on the game. So they are quite real, and quite affect CR.

When someone stalls against you thoralf, that's not because of their skill level. We know this because anyone and their dog can do that...it's because of your skill.

You intimidate them.

They've seen what players you have. They've seen your ability to play.

and they are afraid of what you might do with them.

Foregoing a 3-1 and settling for the 2-1 is a reflection of YOUR SKILL, not theirs.

And for that, you shouldn't have to surrender as much CR. You were good enough to stop the 3-1.

And it was the intangible of how much of a Nervous Nelly your were able to make your opponent.

Stalling isn't skill. It's exploiting the flaws of the rules. And the game commissioners are now addressing it.

Seasons (where you buy back players based on the # of TDs and CAS)

spp and skills for TDs (less TDs, less skills)

CR bonus for higher TD differentials

TD differentials in settling ties.

and awarding tournament points based on #TDs and CAS instead of flat wins/losses.

The game commissioners are working on closing this loophole in the rule mechanics that have been encouraging cheesey play.

As for top 100 coaches, I've played quite a number of them and the score isn't always dTD of 1. But more importantly, they play that way because fumbbl hasn't implemented seasons, or the point scoring system to winning a tournament.

Right now, a "C" effort is as good as an "A"...so why waste energy going for the A?

_________________
Hail to Manowar! The latest charioteer to DIE for bloodbowl! - Slain, by Ghor Oggaz
thoralf



Joined: Mar 06, 2008

Post   Posted: Oct 20, 2017 - 04:53 Reply with quote Back to top

Arktoris wrote:
Intangibles *can* and often are a reflection of CR.


If intangibles can reflect, then they're not intangibles, aren't they?

_________________
There is always Sneaky Git.
Arktoris



Joined: Feb 16, 2004

Post   Posted: Oct 20, 2017 - 05:11 Reply with quote Back to top

thoralf wrote:
Arktoris wrote:
Intangibles *can* and often are a reflection of CR.


If intangibles can reflect, then they're not intangibles, aren't they?


That's correct. They are tangible...the math that determines their outcome is simply beyond what our brain can handle. So we perceive them as "intangible".

Kind of like the RNG. It's not random at all. The twister is so complex that the numbers it spits out are perceived as random by our tiny human brains.

_________________
Hail to Manowar! The latest charioteer to DIE for bloodbowl! - Slain, by Ghor Oggaz
thoralf



Joined: Mar 06, 2008

Post   Posted: Oct 20, 2017 - 05:25 Reply with quote Back to top

Arktoris wrote:
When someone stalls against you thoralf, that's not because of their skill level. We know this because anyone and their dog can do that...it's because of your skill.

You intimidate them.

You flatter me, A.

Here's an alternative explanation.

A win by one TD gets you 0.9 point times something.

A win by two TDs increases that parameter by 0.1.

Many BB situations are of the following case:

- either you risk little and get 0.9 or
- you risk more (a draw being 0.5) and get the full 1.0

So it's basically a decision to risk 0.4 to get 0.1.

If you risk 0.4 over and over again to get 0.1, you will lose more CR than you'll win if you don't win these bets more than 80% of the times.

If you do that bet and consistently win, then of course more power to you.

_________________
There is always Sneaky Git.
Arktoris



Joined: Feb 16, 2004

Post   Posted: Oct 20, 2017 - 05:32 Reply with quote Back to top

from a numbers point of view, yes, that's saying the same thing. I agree.

Ok, this topic has caused me to rant. I must stop. I'm retiring from this thread.

_________________
Hail to Manowar! The latest charioteer to DIE for bloodbowl! - Slain, by Ghor Oggaz
Grod



Joined: Sep 30, 2003

Post   Posted: Oct 20, 2017 - 08:11 Reply with quote Back to top

thoralf wrote:
Arktoris wrote:
When someone stalls against you thoralf, that's not because of their skill level. We know this because anyone and their dog can do that...it's because of your skill.

You intimidate them.

You flatter me, A.

Here's an alternative explanation.

A win by one TD gets you 0.9 point times something.

A win by two TDs increases that parameter by 0.1.

Many BB situations are of the following case:

- either you risk little and get 0.9 or
- you risk more (a draw being 0.5) and get the full 1.0

So it's basically a decision to risk 0.4 to get 0.1.

If you risk 0.4 over and over again to get 0.1, you will lose more CR than you'll win if you don't win these bets more than 80% of the times.

If you do that bet and consistently win, then of course more power to you.


Once you are sure to be up by 2 tds, you now have only downside by going for a bigger win (since bonus CR is for 2 or more TD difference). So you should now stall in this case, even if you feel sure your opponent can make at most 1 TD and you will win anyway. Stall! Don't be nice to your to your opponent, you are now penalised for it.

_________________
I am so clever that sometimes I don't understand a single word of what I am saying.

Oscar Wilde
muaddib68



Joined: Apr 27, 2014

Post   Posted: Oct 20, 2017 - 09:15 Reply with quote Back to top

ok i made some calculation my self !This are i think totaly false but was just a first shot uh ; so my ratio win /defeat was 1/4.74 ; and my game number was 1999. ( take all of them)
If i taking the fact i've played more time better player than me , for a win i gain about 0.75 of cr and for a lost ive taken 0.33 for reference number ( totaly false i m sure but was just for making an exemple ..)i absolutly dont take score , just lost and win ok ??Was an average and was arbitrary choice i admit !
If we admit a tide game dont change cr , i dont change cr for a tide ..so can just admit i have win 163 games ...for 1577 defeat

So my lost cr was about 500 cr point ; and my winnin was about 120..?? so if we take this i will have a negative cr ^^ so ...


Last edited by muaddib68 on Oct 20, 2017 - 10:39; edited 1 time in total
tussock



Joined: May 29, 2011

Post   Posted: Oct 20, 2017 - 09:15 Reply with quote Back to top

Grod, there's always been penalties for scoring. People stall because it's an effective tactic that protects your team. They stall for 1-0, they stall for 2-0, they stall for 3-0, because it works, at all times, and all places. Controlling the clock, taking less blocks, forcing your opponent to roll more dice later to score, as long as you're fairly safe, you stall. It's bloodbowl.

Not that it's always safe to stall, nor even the best way for your team to minimise blocks taken on good players, thus your opponent can totally force you to score. This is also bloodbowl.

_________________
ImageImage
Christer



Joined: Aug 02, 2003

Post   Posted: Oct 20, 2017 - 10:23
FUMBBL Staff
Reply with quote Back to top

The amount of arguing over the TD differential effect is staggering, considering the absolutely tiny effect it has on CR. That being said, I have since writing the original post removed that part of the formula. Not because I don't think it has merit, but because I don't feel it's important enough to defend my position on it.

The current formula is in a place where I'm happy with it. There are some minor changes from what I posted at the top of this thread, and once I get some spare time I'll update the post to reflect the differences. In short, it's the removal of the TD differential effect, the removal of racial factors and a modification to the cross-bracket k variation.

Of these, the racial factor is the only really significant change, so let me go a bit deeper into the reasoning behind this change.

The intent of the racial factor was to build a system that would map the win rates of the respective races versus other races. To make this reasonable, it kept track of a Race Ranking, calculated very similarly to CR for each race. In addition to this, the CR was set up in 10 different brackets per race and tracked the individual race brackets against each-other. For example, one ranking was used for TV1000 humans vs TV 1000 orcs, and another was used for TV 1000 humans vs TV1200 orcs.

Now, this week of working with the CR system showed me that there are problems with this approach. The race ratings were very very close to 50% across the board. Almost no variation at all. This meant that the overall win probability of the CR system was being pulled closer to 50% than felt correct. This is basically a result of people who play each race are, on an aggregated level, average players on average (surprise, surprise..), meaning all racial ratings were mostly 50%. Another issue was that some rarer race/bracket combinations were simply too rare to have a meaningful racial rating (e.g. things like low TV stunties vs high-TV dwarves, or even high tv stunties in general).

In the end, the system performs better without it in place, so it has remained disabled and will remain so. The concept of racial ranking could probably be reintroduced; a better way to do it that would suit the overall formula better would be to add another component in the core p formula (something like dRF/RF_div), next to the CR and TV components.

The racial factor was inherently designed to take into account how each race performs against the others, in a way that TV can't. A better way to deal with this would be to improve how the team difference is calculated in terms of the rating system. This, as some of you might realize, is effectively equivalent to re-introducing TS (Team Strength). And that's something I've been considering for a while anyway...
mrt1212



Joined: Feb 26, 2013

Post   Posted: Oct 20, 2017 - 10:30 Reply with quote Back to top

Dude, if you can create a Team Strength that is robustly predictive you might get many more Patreon dollars out of this curious git. But you'll have to show your work because the value is in describing the process of how you did it. Wink
PurpleChest



Joined: Oct 25, 2003

Post   Posted: Oct 20, 2017 - 10:33
FUMBBL Staff
Reply with quote Back to top

Arktoris wrote:


Stalling isn't skill. It's exploiting the flaws of the rules. And the game commissioners are now addressing it.

................

The game commissioners are working on closing this loophole in the rule mechanics that have been encouraging cheesey play.


And here the obfuscation ends and the truth is revealed.
Arktoris is, and has always been, arguing against stalling as a valid tactic.

BB is not sport but to torture the sport analogy further:

Most leagues of any sport are decided on wins/ties/losses alone. (Except in the USA which doesnt believe in draws)

Some have moved to awarding bonus points for in game behaviors (cricket and rugby, maybe others) to improve the play style by encouraging more scoring. To improve the spectator experience of their paid for product.

So if we do move to a CR system that is designed to encourage a new play style, let's not pretend we are not doing so.

But BB is not sport. Personally when I look at a CR, I expect to see a number indicating how good at winning BB that coach is. Not what style they play.

More importantly I expect someone rated higher to have a better chance of winning game X than someone rated lower (all other factors being equal).

Then one has to factor in the races of BB. If coach A wins 70% with a mainly STR based roster of teams, and coach B wins 60% but almost exclusively with elves, I would expect coach A to have the higher CR. I do not care how those wins were achieved as the two sets of teams require and respond to different play styles. A +2TD game for Orcs/Dwarves is exceptional, but hardly newsworthy for Elves.

Finally I have always thought it polite, when not feeling threatened, to not stall for 2-0, but to let the opponent have another drive and try for 2-1. As it means nothing to me, and might make their game experience better (Not in tourneys, or with coaches that can possibly turn that infinitesimal chance into a draw/win).

Also, when sports have introduced new criteria affecting the points awarded for a game, they havn't (to my knowledge) applied these changes retroactively.

Why not decide that scoring in the central 4 squares of the td zone displays 'positional/field mastery' and award bonus points for that? there are plenty of sports scoring systems as precedent.


[EDIT] Utterly Ninja'd by Christer, ending the debate.

_________________
Barbarus hic ego sum, quia non intelligor illis -Ovid
I am a barbarian here because i am not understood by anyone
Grod



Joined: Sep 30, 2003

Post   Posted: Oct 20, 2017 - 11:46 Reply with quote Back to top

PurpleChest wrote:
Arktoris wrote:


Stalling isn't skill. It's exploiting the flaws of the rules. And the game commissioners are now addressing it.

................

The game commissioners are working on closing this loophole in the rule mechanics that have been encouraging cheesey play.


And here the obfuscation ends and the truth is revealed...


Actually, the new ruleset reintroduces special cards as a standard rule of playing (2 each on top of inducements?). I remember back in the good old days when you wouldn't stall, because your opponent would be looking intently at his dirty trick card and you were sure there was a trampoline trap in the endzone... I feel that the random element coming back by reintroducing special play cards would help go back to this dynamic. Obviously online we won't be getting all these cards anytime soon, but I think with cards is the way it is intended to be played.

_________________
I am so clever that sometimes I don't understand a single word of what I am saying.

Oscar Wilde
Christer



Joined: Aug 02, 2003

Post   Posted: Oct 20, 2017 - 13:58
FUMBBL Staff
Reply with quote Back to top

PurpleChest wrote:
Finally I have always thought it polite, when not feeling threatened, to not stall for 2-0, but to let the opponent have another drive and try for 2-1. As it means nothing to me, and might make their game experience better (Not in tourneys, or with coaches that can possibly turn that infinitesimal chance into a draw/win).


This brings up a very interesting point when it comes to the ranking system. Given that there is that tiny chance to come back (think riots and blitz turns), giving anyone a tiny chance will, in the long term, backfire on you.

The distinction here is that no ranking system could possibly differentiate your "true" ability versus the level you play at. Explicitly "playing worse" for whatever reason means that your results will be worse than they could be. Over time, this will translate to a lower rating than you would end up having if you played at your full potential.

I'm not saying this is a bad thing to do. It's great for anyone to try to make the community a better place.

The point here is that the rating system does its best to indicate what level you are playing at, and not necessarily what level of coach you are. It also inherently puts more weight into recent results than older ones, which makes sense as the option would be that the system would assume you couldn't improve as a player.

In addition to that, another thing many people overlook when talking about the CR system is that the design of it will average many many games over time. With more games played, the inherent randomness of the dice will have a lower and lower significance to the end result. Sure, a single match can be lost or won purely because of dice rolls. Even two or three. However, after 10 matches, you've simply rolled too many times for the statistical distribution of them being anything but average. And at that point, it's your ability to utilize the good dice rolls well, and reduce the negative effects of the bad ones. Thus, over time, the rating system will converge your rating towards the rough place indicating what level you play at. The individual luck and bad luck streaks will make you move up and down the rating somewhat but in the end everyone has the same dice with an equal amount of lucky and unlucky streaks.
Display posts from previous:     
 Jump to:   
All times are GMT + 1 Hour
Post new topic   Reply to topic
View previous topic Log in to check your private messages View next topic