For and Against

[eloquentEEL]

For next season, I propose a revision to the For and Against system.

I understand the need to include against points, because different refs mark in slightly different ranges, so you need some form of comparison. However, as a team has no control over the against points, especially whether or not the opposition gets 3, 4 or 5 articles in, I think the for and against should be worked out on averages rather than totals. Ie. if a team only gets 4 posts in, then divide the total score by 4 for the for and against.

Any thoughts?

 

[El Coconuto]

It sounds good in theory, although the way it is now serves as incentive for teams to get five posters in each round and rewards teams who do get in five posters. I do see what you're saying though, but I'll leave this one up for the proper channels to think about.

 

[Anonymous]

On the surface it sounds like it would accomodate those teams who don't get all 5 posts in.

 

[antonius]

I think it's a good point, if a team is lucky enough to meet a team that only has four posters then the team with four posters has given the team with five a big advantage, that could affect another team later in the comp ie:- lions gained a superior for and against this year because they played a team with less posters, that has effectively meant that the Panthers have to get a huge win to overcome that now. When the original for and against difference was out of their (Panthers) control.
Does that make sense?

 

[eloquentEEL]

thanks antonius,

that's a perfect example. i think team's have enough incentive to get 5 posters in, because failing to do so is almost an automatic loss of 2 competition points.

 

[Anonymous]

I think we should do whatever we can to encourage a full compliment of players from every team, every round.
I'd like to hear a proposal which rewards teams for consistently getting all five posts in. Just thinking aloud now but perhaps a bonus point every three rounds would see the for-and-against even out.

 

[rabs]

perhaps just work on the for points only

 

[antonius]

On second thoughts dividing by the number of posters you could have the team with four all score say 8.8, and the winning team score an average of 8.6, so they (winning team) end up with a minus for and against?? which seems unfair also.

 

[Anonymous]

By coincidence, I was thinking about this last week when I saw how important the F&A was going to be.
It seemed to me that the teams who didnt get all players in had effectively 'disadvantaged' other teams in the comp.
At the time I couldn't think of a solution and figured it was good enough the way it is.

Nevertheless, if we have all 10 posts in from every match in every round, then it wouldn't be an issue.

I came to the conclusion that we should be working towards a more solid foundation so a full compliment of players are happening every week and that this becomes the norm.

For the record, 2004 is marked improvement on 2003 with all teams fielding a full side for most rounds and in a couple of cases, every round.

Having said that, any new ideas should be put forward.

 

[eloquentEEL]

perhaps just work on the for points only

as i see it, the main reason for keeping Against points is to correct for different marks ranges for the referees.

if we only work on the for points, then one team might get the highest scoring ref for a lot more games than another team, who might get the lowest scoring ref for a lot of games.

i think one solution that might be acceptable to everyone, is to only use the For points, but at the end of the season, correct for the reffing via a process of normalisation, ie. by looking at the statistical distribution of marks that each ref has awarded over the whole season, and how many times each team has been reffed by each ref, and correcting appropriately.

this would require some statistical gymnastics, but would probably the fairest system for everyone

 

[antonius]

Seriously I think the refs are pretty much marking within the same peramaters I know when I started I was marking lower than the others, so I adjusted my start point (for want of a better word) to bring mine in line with the more experienced guys. I always look at all the games each round and the marks in them are on par with each other. If you check the tables I think you'll find they bear witness to that. Apart from the game where there wasn't a full side.

 

[LeagueNut]

Just throwing this in - a match played 5 vs 4 could be seen in a similar light to an NRL match played 13 vs 12 ... it's normally reasonable to expect the 'full' team to win, and win well.

The Sharks/Eels match (NRL, not F7's!) from late 2003 is a good example ... the Eels gave their for-and-against a massive boost with their 74-4 win and there wasn't a hell of a lot that the other teams could have done about that!

 

[eloquentEEL]

In the NRL, there are floggings with 13 vs 13 matches. Either way, both teams influence the points for and against. In F7's, there's no way to influence points against.

Regarding the actual stats, i've collected some stats over rounds 6, 7, 8 and round 1.

Ref - Low score - High score - Average score

Rnd 8
Salivor 8.1 8.7 8.35
Mystique 7.8 9.5 8.8
Antonius 8.1 9.8 8.74
Antonius 8.0 8.8 8.54

Rnd 7
Mystique 7.7 8.7 8.37
Antonius 8.2 8.7 8.47
Salivor 8.2 8.8 8.47
Antionius 8.3 8.7 8.52

Rnd 6
Mystique 8.3 9.3 8.66
Antonius 6.5 8.7 8.27
Antonius 8.4 8.7 8.49
Salivor 7.9 9.0 8.59

Rnd 1
Backpacker 7.9 8.8 8.45
Salivor 8.0 8.8 8.18
Mystique 7.7 8.5 8.10
Antonius 8.0 9.0 8.41

there is a bit of variation between ref's, but there seems to be a bigger variation over time (although that may be due to improvement of articles), so i don't know where all of that leaves us, especially since one or two refs are coming/going for next season

 

[antonius]

I think that shows what i'm saying the refs are scoringpretty much the same each round. Ther have been the odd stand out posts in some games but generally I think you'll find the mean average is pretty close.