46 coaches online • Server time: 14:39
Index Search Usergroups Profile
Log in
Recent Forum Topics goto Post Another I need a log...goto Post SWL Season LXXgoto Post A Very Steely Setup ...
SearchSearch 
Post new topic   Reply to topic
View previous topic Log in to check your private messages View next topic
Arktoris



Joined: Feb 16, 2004

Post   Posted: Oct 17, 2017 - 20:50 Reply with quote Back to top

Bloodbowl, like most games, is about excitement. Stuff happening. The more passing, interceptions, casualties, and touchdowns, the better.

You can strive to "pass the class" with minimum risk and effort. Grind for the 2-1 stall/win, and you'll still win. But that is the C student mentality, and Christer is now recognizing that in the CR.

Same with any other sport.

Team A plays team B in soccer.

team A barely beats B with a score of 1-0.

A > B

but if team A whollops team B 7-0

A >>> B

_________________
Hail to Berserker Eric! The latest charioteer to DIE for bloodbowl! - Slain, by Zufbar Hammerers!
Sp00keh



Joined: Dec 06, 2011

Post   Posted: Oct 17, 2017 - 21:27 Reply with quote Back to top

JackassRampant wrote:
Sp00keh wrote:
A 1-0 win is worth 90% the CR of a 2-0 win.
Only if p=1.00. If p=0.50, a 1-0 win is worth 80% the CR of a 2-0 win. No?

But it caps out at +2 TD, so I don't think it's too bad.


'No' because p is not a multiplier on S
CR' = CR + k * (s-p)

EDIT - BASIC MATHS FAIL. UPDATED:

example results with CR starting on 150, k=1, for either a 1TD win (S=0.9) or a 2+TD win (S=1.0), over a range of P values
Code:

 s     p       CR`      difference
1.0   0.3      150.70   85.71%
0.9   0.3      150.60   
1.0   0.4      150.60   83.33%
0.9   0.4      150.50   
1.0   0.5      150.50   80.00%
0.9   0.5      150.40   
1.0   0.6      150.40   75.00%
0.9   0.6      150.30   


where 'difference' is the CR` comparing S=1.0 with S=0.9 (on the row below)


Last edited by Sp00keh on Oct 18, 2017 - 09:25; edited 2 times in total
Tripleskull



Joined: Oct 12, 2003

Post   Posted: Oct 17, 2017 - 21:55 Reply with quote Back to top

MattDakka wrote:
I agree, there is not necessarily direct and strict correlation between winning with high TD difference and coach's skill (of course, there could be exceptions).


If there is a correlation I would expect it to be inverse if you looked at the average Delta-TD of all wins.
Grod



Joined: Sep 30, 2003

Post   Posted: Oct 17, 2017 - 22:00 Reply with quote Back to top

How likely is a higher ranked coach actually going to win (and not draw or lose) a match? The best coaches on this site very rarely have a win percentage over 75. The game is quite random after all. It's a relevant question particularly in the Box where you can't choose your opponent and may be penalised by matching a lower CR opponent where your expected CR change is negative.

Anyway its a delicate and complicated thing to get just right. Sounds like you're getting on the right track. Must be a pain to have a new idea, but then have to wait hours before you see the result!

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

Oscar Wilde
Tripleskull



Joined: Oct 12, 2003

Post   Posted: Oct 17, 2017 - 22:01 Reply with quote Back to top

Arktoris wrote:
You can strive to "pass the class" with minimum risk and effort. Grind for the 2-1 stall/win, and you'll still win. But that is the C student mentality, and Christer is now recognizing that in the CR.


I am sorry but you could hardly be more wrong. This is the ideal A+ play. A 4-0 win usually tells a story about a lot of luck and not seldom a story of unnecessary risk. Risk you do not want your team mates to take on a team if you are competing at a high level.
Sp00keh



Joined: Dec 06, 2011

Post   Posted: Oct 17, 2017 - 22:03 Reply with quote Back to top

Arktoris' ranked High Elf team Exorcists
average total TDs (both sides): 3.6791
average TD difference: 1.4791
games played: 240

PurpleChest's ranked dwarf team Purple Delvers
average total TDs (both sides): 2.2166
average TD difference: 1.2635
games played: 277

elf games are higher scoring overall than dwarf games, they have more 'TDs scored by both sides' by 66%
that's the obvious one but its not important

elf games have more TD-difference than dwarf games, but only by 17%
Dominik



Joined: Oct 29, 2004

Post   Posted: Oct 17, 2017 - 22:24 Reply with quote Back to top

PurpleChest wrote:
I have always (perceptions aside)defended CR as a reasonable attempt at defining coach rating. Go check, I stand by everything i have ever posted.

I struggle to keep up with the pure math, but can usually follow the reasoning and hence eventually get the math too.

But I am lost as to why a 4-0 win is better than a 1-0 win.
It would seem contrary to your stated aims.
Increasing, as it does, the value of unfair games with high score outcomes.
Promoting certain styles of play and certain races while devaluing others,
Bringing 'perception' of a 'good' win into something seeking a rational outcome.

So I am interested as to why you feel this is an appropriate measure? It feels like a step toward 'personal preference of play style' and away from 'how likely is a win' to me.


It could be to reward those who coach Elves and Skaven, as they frequently win that high contrary to Nurgle, Dwarf and Chaos coaches.
happygrue



Joined: Oct 15, 2010

Post   Posted: Oct 17, 2017 - 22:27 Reply with quote Back to top

It's also worth pointing out that TD differential varys widely by team. A fling team that wins with one net touchdown vs an elf team that wins by three net touchdowns mean very different things.

However, if we're trying to measure coach skill, then net TDs across *all* teams over time says a lot of things. For example: Two coaches, both of whom are ranked at 170, each with 1000 games to their credit, one with TD differential per game at -0.5 and the other with TD differential at 1.5 per game. We can't know exactly what's going on there without more info, but a betting coach would know where to put money...

That's just one part of what's going on, but it would be interesting to find a "long term" metric like TD differential with which to temper the "what is an upset" calculations.

EDIT: I'm not trying to suggest it's super easy, because which races are being played by the various coaches are going to play into that. Just musing.

_________________
Come join us in #metabox, the Discord channel for HLP, ARR, and E.L.F. in the box!
Image
Sp00keh



Joined: Dec 06, 2011

Post   Posted: Oct 17, 2017 - 22:48 Reply with quote Back to top

i'm pretty sure you can't get to CR170 with negative average-touchdowns-difference

i'm pretty sure average-touchdowns-difference would correlate quite closely with winrate (ignoring racial effects)
Arktoris



Joined: Feb 16, 2004

Post   Posted: Oct 17, 2017 - 23:02 Reply with quote Back to top

Tripleskull wrote:
Arktoris wrote:
You can strive to "pass the class" with minimum risk and effort. G But that is the C student mentality


I am sorry but you could hardly be more wrong. This is the ideal A+ play.


it is the ideal play...for those with the C student mentality.

If you are clearly the better coach, I expect you to decisively win, not barely win.

Same if the New York Yankees played my local high school at baseball. I expect the Yankees to win by a big blowout margin. If the score were 2-1, I bet the headline in the papers would be "NY Yankees struggle to beat local high school"

Tripleskull wrote:
A 4-0 win usually tells a story about a lot of luck and not seldom a story of unnecessary risk.


First off, it isn't unnecessary risk if it leads to a touchdown.

For CR we're suppose to look at the average not just one game. Anyone can have a lucker game and beat PeteW 3-0. And this will be balanced off as others have lucker games vs you. What the RNG giveth, it takeths away and the overall CR stays the same.

But if I can consistently beat PeteW 3-0 game after game and you can only barely win 2-1 vs him...I'm clearly the better coach over you and my CR should reflect that.

This is also why some tournaments are now scoring based on TDs and CAS and not just a flat win/lose record. The commissioners are addressing the C student mentality and are building in a factor to counter it.

_________________
Hail to Berserker Eric! The latest charioteer to DIE for bloodbowl! - Slain, by Zufbar Hammerers!
bghandras



Joined: Feb 06, 2011

Post   Posted: Oct 17, 2017 - 23:19 Reply with quote Back to top

Did you consider using different factors for the racial rankings? The theory and the sample i looked at suggests that the same were used for general and racial ones.

_________________
Image
"This is turning into the best thing on FUMBBL" - Seventyone
"THIS LEAGUE IS AWESOME!!!!" - Kummo
"Last league i would leave." - SpecialOne
Sp00keh



Joined: Dec 06, 2011

Post   Posted: Oct 17, 2017 - 23:46 Reply with quote Back to top

Actually I agree with purplechest and tripleskull, sort of

If safe a 2-1 win is available, take it.
Letting the other guy score once means risking less of your own players chasing after the turnover.
Chasing the turnover, or blocking his TD, could change a win into draw.
Stalling out the half is safer than kicking him the ball and hoping to steal it.

However, if TDs are available, if a turnover is gifted to you, if you bashed them off the pitch, then scoring more shows that you had a stronger win
Scoring quick might mean your players are protected from fouls etc

So, 'it depends'. it's not a massive factor, anyway
The_Murker



Joined: Jan 30, 2011

Post   Posted: Oct 18, 2017 - 00:25 Reply with quote Back to top

Sorry, Arktoris.. I love your team and your Blood Bowl spirit on FUMBBL, but your level of Blood Bowl (by choice) just isn't a top level of the game, when compared to the top 100 coaches on the website. Perhaps this causes your view to differ quite significantly from theirs.

Your grade-school analogy is yours alone, and does not fit. Your New York Yankees analogy is completly out to lunch. You don't throw dice in baseball. An 'action packed' game might be better in your eyes, but it is not a more consistant way to win. CR is about wins.. not action, not game 'quality.'

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



Joined: Oct 15, 2010

Post   Posted: Oct 18, 2017 - 00:49 Reply with quote Back to top

Sp00keh wrote:
i'm pretty sure you can't get to CR170 with negative average-touchdowns-difference

i'm pretty sure average-touchdowns-difference would correlate quite closely with winrate (ignoring racial effects)


If the sample size is large enough, of course you could? In fact, I just looked up a coach with negative TD differential (though much less than -0.5 per game) who I know had very high CR a few times - close to number one on the site IIRC - due to the magic of min/maxing. I bet it would not be super hard to find other such examples, as many times such players are not actually playing to score at all, and are quite happy to let the score become zero-to-much in a game that they have already lost as long as it means killing some menz.

Maybe it would be harder these days than it used to be, but sure: change it to 0.5 to 1.5 and the argument is the same.

Though that line of thinking does make me wonder how score-influenced ranking changes when not everyone is trying to play the same game. For example, many coaches will continue to attempt to minimize the opponent's score once the game is "lost". But not everyone plays that way... some coaches would be more interested in preserving or gaining spp. Others would do a bit of both. I don't know how much such playstyle differences mess up CR, maybe it's a trivial amount and nothing we should worry about?

_________________
Come join us in #metabox, the Discord channel for HLP, ARR, and E.L.F. in the box!
Image
Arktoris



Joined: Feb 16, 2004

Post   Posted: Oct 18, 2017 - 01:37 Reply with quote Back to top

The_Murker wrote:
Sorry, Arktoris.. I love your team and your Blood Bowl spirit on FUMBBL, but your level of Blood Bowl (by choice) just isn't a top level of the game, when compared to the top 100 coaches on the website.


and when it comes to those top 100 coaches, the ones that play it safe and only play vs 140s coaches should not get the same CR reward as those who took the "unnecessary" risk to play high CR coaches and won.

Same with team choice. Those top 100 coaches who choose to always play it safe with tier 1 teams against tier 2,3 teams should not get the same CR reward as those who took the "unnecessary" risk and played a tier 2 team against tier 1 and won.

Same with tactic styles. Those top 100 coaches who choose to play it safe for a 2-1 grind win should not get the same CR reward as those who took the "unnecessary" risk of going for a decisive win.

Fame doesn't belong to the ubiquitous wall-flowers of life. It's given to those that have the balls to take the big risks, reach for the brass ring, and make it.


Do you understand now?

_________________
Hail to Berserker Eric! The latest charioteer to DIE for bloodbowl! - Slain, by Zufbar Hammerers!
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
Powered by PNphpBB2 1.2 © 2003 PNphpBB Group
Credits