rugbyunion
Latest News:
New Page 1

Whatever your views on Saracens, whether a Sarries fan or not, leave them here.

To leave a message on this board you must register. To register click HERE, it takes only a minute.
Non-rugby posts are welcome, but please prefix your subject header with "OT" or "Off Topic".


Thought for the Day:
CHAMPIONS AGAIN!

Latest: EXETER CHIEFS 10:27 SARACENS (Aviva Premiership Final)
Audio: Click the link below. If it ain' there, it ain't on!
Upcoming TV:

BBC Online Rugby Union Commentaries

The Fish | Rugby Union News | Fez Boys | Saracens Fixtures | The SSA


Better than Decorating
Sara'sman (IP Logged)
09 October, 2017 15:51
As usual, faced with something I didn't fancy doing, I found a more interesting use of my time and thought I'd share the results for those interested.

I looked back at last year's results having first grouped the table into quartiles - Top (Wasps, Exeter, Sarries), Upper (Leicester, Bath, Quins) Lower (Northampton, Newcastle, Glaws) and Bottom (Sale, Wuss, Bris). I then averaged the scores and rounded to the nearest half giving the following results:

Top v Top 3-2, Upper 4.5-0.5, Lower 4.5-0.5, Bottom 5-0.5
Upp v Top 1.5-3, Upper 3.5-1.5, Lower 4.5-0, Bottom 4.5-0.5
Low v Top 2-3, Upper 1.5-3.5, Lower 2.5-2.5, Bottom 4.5-0.5
Bot v Top 1-4, Upper 3-2, Lower 1.5-3.5, Bottom 3-2

It surprised me that, despite a small sample size, the table had some remarkably consistent patterns, horizontally, vertically and most of all diagonally. And one clear anomaly - Bottom v Upper.

Keeping the groups the same and replacing Bris with Irish so far our results have been: Saints +0.5, Bath -3, Newcastle +1, Sale 0, Wuss 0 and Wasps +2 compared to "expectations", leaving us at +0.5!

I made a couple of very minor changes to the above table to "smooth" the patterns where one more point in 9 games would have resulted in a minor difference. Based on this I looked at each team's results this year, and where it places them in a league table that should, in theory, take account of their fixture list so far.

Saracens +0.5
Saints +9.5
Exeter -0.5

Bath +3
Newcastle +3.5
Leicester +0.5

Sale +5.5
Gloucester -2
Quins -3.5

Wasps -12
Worcester -3.5
Irish -4.5



Edited 3 time(s). Last edit at 10/10/2017 22:05 by Sara'sman.

 
Re: Better than Decorating
F-F-F-FEZ (IP Logged)
09 October, 2017 16:24
You should have done the decorating!! winking smiley

 
Re: Better than Decorating
derbyshire fan (IP Logged)
09 October, 2017 20:23
Another stats fan - you should compare spreadsheets with Superman Wears Schalk Brits Pyjamas

 
Re: Better than Decorating
AP (IP Logged)
09 October, 2017 21:45
Sorry, it needs more explanation of your methods before I can begin to understand it.



Successful hills are here to stay
Everything must be this way
Gentle streets where people play
Welcome to the Soft Parade

 
Re: Better than Decorating
JO'G (IP Logged)
09 October, 2017 22:07
From my reading, the grid represents the average match points for and against for each team in the relevant quartile against teams in the same and other quartile

so Sarries v other top 3 -
expect 3 points at home, 2 away
4-6 expect 4.5 at home 3 away
7-9 expect 4.5 at home 3 away
10-12 expect 5 at home 4 away

so comparing what you would expect a team to get in their matches against what they have this season

Saints doing very much better
Sale, Newcastle and Bath doing better
Quins and Worcester doing a bit worse
Wasps doing really, really badly



Park team from London
Just a park team from London
European Champions
Just European champions



Edited 1 time(s). Last edit at 10/10/2017 08:54 by JO'G .

 
Re: Better than Decorating
flavman 35 (IP Logged)
10 October, 2017 06:41
My head hurts

 
Re: Better than Decorating
nedrichards (IP Logged)
10 October, 2017 10:02
Ah! Good to see some NFL style 'strength of schedule' thinking applied to an 'everyone plays everyone' league. For extra credit I wonder how international periods affect the 'top vs bottom' argument, or do teams like Quins, whose England representation doesn't match their league position, skew it too much.

 
Re: Better than Decorating
Sara'sman (IP Logged)
10 October, 2017 10:15
AP - I hope JO'G's explanation has helped. It was just a bit of task avoidance fun that led on to two other excuses. The task was further avoided by posting on here!

If we take last year's 6 games between the top 3 (Pests, Ex, us) the match points scores were 4-1, 5-0, 3-3, 0-5, 5-0 and 2-2 giving an average score of 3-2. Thus when we played Pests this year we could "expect" a 3-2 win . I then did the same for each group v each of the others.

Next, I assumed the groups to be the same for this year to look at how we are doing against "expectations" - 0.5 up. Against Wasps for example (expect 3-2) we won 5-0 gaining 2 points against expectations for a top 3 team. (Pests also lost 2).

Finally, with my paintbrush inactive, I looked at all the other sides results against their expectations assuming them to start in the same group of 3 as last year.

Hope that hasn't created more headaches. F-F-F-FEZ thanks - made me chuckle, you're right!



Edited 1 time(s). Last edit at 10/10/2017 11:24 by Sara'sman.

 
Re: Better than Decorating
Sara'sman (IP Logged)
10 October, 2017 10:27
Quote:
nedrichards
Ah! Good to see some NFL style 'strength of schedule' thinking applied to an 'everyone plays everyone' league. For extra credit I wonder how international periods affect the 'top vs bottom' argument, or do teams like Quins, whose England representation doesn't match their league position, skew it too much.

ned - I think there are so many variables at play (Internationals, Injuries, Rotation, Ref's, bounce of the ball, targeting, ground conditions, Europe, end of season situation ...) that no method can be taken too seriously, least of all mine!

What set me off was that after round 5 Newcastle and Bath had had very different fixtures in my view. Apart from playing us at home and each other, Bath played Leicester, Saints and Wasps, while the Falcons were playing Sale, Worcester and Irish. I wanted to know which team was doing better.



Edited 1 time(s). Last edit at 10/10/2017 11:27 by Sara'sman.

 
Re: Better than Decorating
OldMarovian (IP Logged)
10 October, 2017 17:02
On a podcast an interesting point was made that if you swap Newcastle and Wasps around the table is pretty much business as usual.

 
Re: Better than Decorating
hasta (IP Logged)
10 October, 2017 20:30
Prompted by Sara'sman's post on my thread along similar lines on the Bath board, here's my version of strength of schedule. Based on current league position, scaled by 1.5 for away games, higher score means harder schedule to date:

Wuss/Irish 53
Bath/Wasps 49
Saints 48.5
Tigers 48
Newcastle 47.5
Saracens 44
Gloucester/Sale 43.5
Quins 36.5
Exeter 33.5
Mean - 45.75

My conclusion is that, yourselves excluded, it's an absolute scrap for 2-6th place. Quins should be seriously concerned given their performance to date with a relatively easy schedule. Exeter may not be as clear-cut a second place as everyone thinks.

 
Re: Better than Decorating
AP (IP Logged)
10 October, 2017 20:34
Sara'sman, I sort of follow what you are doing (with my 40 year old A at Maths A Level) but it's your final league that was confusing me. You say:

"Based on this I looked at each team's results this year, and where it places them in a league table that should, in theory, take account of their fixture list so far."

But what you set out is the present league standings. Quins and Worcester both have a factor of -3.5. Which is doing worse? Where should they be in a league?



Successful hills are here to stay
Everything must be this way
Gentle streets where people play
Welcome to the Soft Parade

 
Re: Better than Decorating
Sara'sman (IP Logged)
10 October, 2017 22:01
AP - you've hit on the last, somewhat dubious feature! I'm suggesting that the order of the teams reflects their performance/position so far but taking account of their fixture list - hard or easy. I'll try to explain my thinking.

Each team "started" in the group of 3 from last season's table. So Quins (7th) started in the Lower group (7-9) but are -3.5 against expectations for the average of this group. They just stay in this group but in 9th place. [sorry I had them 8th originally and will correct]

Wuss (11th last year) start in the Bottom group (10-12) and are -3.5 against the expected average for it, and stay 11th as Irish are performing worse taking account of their fixtures.

So my answer to your question is that both are doing equally poorly compared to expectations. But Quins "expected" to be in a higher group than Worcester so are doing better in absolute terms.

If you're still with me, Sale (10th last year) are +5.5 against the expected average for the Bottom group. This not only takes them up to the Lower group but places them 7th! This is because the averages for each of the four groups last year was 82, 59, 49 and 31 meaning 18 points difference (in 22 games) between Lower and Bottom - roughly 5.5 in 6 games. So Sale would be on 0 in the Lower group, thus ahead of Glaws and Quins!

Just sharing a bit of fun!

 
Re: Better than Decorating
Seany_Boy0511 (IP Logged)
11 October, 2017 17:04
So basically you're saying Saracens and Exeter near the top is how it should be, Bath are doing ok, Sale and Newcastle and Saints are exceeding expectations and wasps and Quins are below par with the others doing as anticipated?

If so that is certainly a lot more interesting than decorating! Although I'd find decorating less mind boggling!

 
Re: Better than Decorating
Sara'sman (IP Logged)
11 October, 2017 18:05
I think JO'G summed it up well in his post Seany, with Irish also doing worse than "expected". Perhaps Wuss and Irish target more points either side of Europe or during international periods.

In summary I offer three stats.

1. A grid showing the (rounded) average match points score of last year's games with teams grouped in 3s by league position. e.g. Bottom 3 hosting Top 3 averaged 1-4.

2. A league table based on performance taking account of the opposition's strength (based on last year's league position). e.g. Saints are second!

3. A number that suggests how well a side is doing based on opposition strength (measured as in 2) and an expectation of finishing in the middle of the same group of 3 as last year. e.g. Exeter are underperforming marginally by 0.5

Much like my decorating I make no claims of perfection in method or result.

 
Re: Better than Decorating
TonyTaff (IP Logged)
12 October, 2017 11:07
I am minded to set a maths test in the portal to this thread!



676.97 (*) donated to the Saracens Foundation due to visits to the Sarries frontpage [www.rugbynetwork.net]

Please read and submit articles for publication. (*) As at October 31, 2017.

 
Re: Better than Decorating
Barty II (IP Logged)
12 October, 2017 11:35
Sara'sman, you might be able to help me with the question that ruined my Masters year fluid mechanics exam...

a) What is the average air speed velocity of an unladen swallow? (choose African or European)
b) What is the largest sized coconut the swallow could conceivably carry to England?

 
Re: Better than Decorating
AP (IP Logged)
12 October, 2017 11:57
Quote:
Barty II
Sara'sman, you might be able to help me with the question that ruined my Masters year fluid mechanics exam...
a) What is the average air speed velocity of an unladen swallow? (choose African or European)
b) What is the largest sized coconut the swallow could conceivably carry to England?

Clearly you are not a king. You have to know these things when you are a king.



Successful hills are here to stay
Everything must be this way
Gentle streets where people play
Welcome to the Soft Parade

 
Re: Better than Decorating
Sara'sman (IP Logged)
12 October, 2017 12:49
a) is easy - zero, assuming it has returned home!
b) also none since the swallow has no arms?

Perhaps I should warn you that whenever asked at interview why I only achieved 3rd class I always replied that I was too intelligent for a second but didn't work hard enough for a first. Seemed to work!


Sorry, only registered users may post in this forum.
We record all IP addresses on the Sportnetwork message boards which may be required by the authorities in case of defamatory or abusive comment. We seek to monitor the Message Boards at regular intervals. We do not associate Sportnetwork with any of the comments and do not take responsibility for any statements or opinions expressed on the Message Boards. If you have any cause for concern over any material posted here please let us know as soon as possible by e-mailing abuse@sportnetwork.net
 
 

Who is online?

Total users online:  

Most users online:  

Users on this site:  

Where are they?