Statistics show

paddy20
Hob Nob Regular
Posts: 1251
Joined: 18 Apr 2007 17:50
Location: Wokingham

Statistics show

by paddy20 » 13 Aug 2013 09:03

Statistics show that the best way to improve your team is to sort out its weakest link. I.e. if you have say your best player that rates at 80% (100% being brilliant) but the weakest player at 25% they are less likely to perform as well as a team that has their best player at 70% and their weakest at 35%. Who would you say in our present team is the weakest link and how would you improve it?

User avatar
stickywhisky
Member
Posts: 396
Joined: 27 May 2011 11:01
Location: Southampton : (

Re: Statistics show

by stickywhisky » 13 Aug 2013 09:14

Not sure if serious??

Colin Cheeselog
Member
Posts: 515
Joined: 15 Apr 2004 16:47

Re: Statistics show

by Colin Cheeselog » 13 Aug 2013 09:28

So in order to improve, football teams should replace their poorest performers with players that will play better? This is a revolutionary new way of thinking.


Dear Football Association,

I have an idea that will change football forever...

Cureton's Volley
Hob Nob Regular
Posts: 1629
Joined: 08 Jan 2013 23:58

Re: Statistics show

by Cureton's Volley » 13 Aug 2013 09:40

Our weak link in 5-4-1 is our striker (whoever plays).

Otherwise our weak link would be Gunter at right back... but he isn't that bad so it shows we are reasonably strong elsewhere.

User avatar
Fox Talbot
Hob Nob Regular
Posts: 1116
Joined: 09 Nov 2009 16:07
Location: Left Back.

Re: Statistics show

by Fox Talbot » 13 Aug 2013 09:47

stickywhisky Not sure if serious??


Read 'The Numbers Game' - all in there.


paddy20
Hob Nob Regular
Posts: 1251
Joined: 18 Apr 2007 17:50
Location: Wokingham

Re: Statistics show

by paddy20 » 13 Aug 2013 10:36

Colin Cheeselog So in order to improve, football teams should replace their poorest performers with players that will play better? This is a revolutionary new way of thinking.


Dear Football Association,

I have an idea that will change football forever...


Not sure that you understood. If the best player is slightly worse and the worst player is slightly better statistics show that you will get better results

paddy20
Hob Nob Regular
Posts: 1251
Joined: 18 Apr 2007 17:50
Location: Wokingham

Re: Statistics show

by paddy20 » 13 Aug 2013 10:37

Fox Talbot
stickywhisky Not sure if serious??


Read 'The Numbers Game' - all in there.


Yep. Thats right. Brilliant book

User avatar
Platypuss
Hob Nob Moderator
Posts: 8203
Joined: 14 Apr 2004 21:46
Location: No one cares about your creative hub, so get your fukcin' hedge cut

Re: Statistics show

by Platypuss » 13 Aug 2013 10:39

Statistics also show that the average person has fewer than 2 legs, so I'm not sure that baldly asserting "statistics show x" is necessarily helpful.

Davezk
Member
Posts: 769
Joined: 05 Sep 2012 21:10

Re: Statistics show

by Davezk » 13 Aug 2013 10:40

So in theory, the optimum team is one where all the players are as average as each other?

Then there is no weak link.

Yeah this idea is flawed...


User avatar
blueroyals
Hob Nob Regular
Posts: 2153
Joined: 02 Sep 2010 02:11

Re: Statistics show

by blueroyals » 13 Aug 2013 10:59

#Pray4Paddy

User avatar
Geekins
Hob Nob Regular
Posts: 2477
Joined: 07 Aug 2007 12:33
Location: Back on a kayak

Re: Statistics show

by Geekins » 13 Aug 2013 11:05

I read the thread title as Spastics Show... oh it is right when you click into it.

User avatar
PistolPete
Hob Nob Regular
Posts: 1344
Joined: 01 Mar 2005 06:38
Location: 1871

Re: Statistics show

by PistolPete » 13 Aug 2013 11:40

paddy20
Colin Cheeselog So in order to improve, football teams should replace their poorest performers with players that will play better? This is a revolutionary new way of thinking.


Dear Football Association,

I have an idea that will change football forever...


Not sure that you understood. If the best player is slightly worse and the worst player is slightly better statistics show that you will get better results


Neuer
Dante
Van Buyten
Lahm
Alaba
Schweinsteiger
Muller
My 62 year old dad
Robben
Ribery
Mandzukic

Vs

Aldershots Team Vs Grimbsby

01 Morris
02 Oastler
05 Goodman
06 Webster
15 Barker
04 Stanley
07 Mekki
08 Rowlands
19 Molesley
09 Williams
10 Paterson

Who would win? Or have I got it wrong? Surely the idea is a bit more complex??

Better to rename the thread "Who is the weakest link"? But then again, it's a flawed question as it depends on formation etc!

User avatar
Ian Royal
Hob Nob Legend
Posts: 35156
Joined: 15 Apr 2004 13:43
Location: Playing spot the pc*nt on HNA?

Re: Statistics show

by Ian Royal » 13 Aug 2013 12:43

Statistically (in other words, not necessarily reliably) the two weakest performers so far, iirc, have been Blackman against Bolton (definitely) followed by ALF against Ipswich (I think). I'll check when I'm home.

But the sample size is really too small to draw any firm conclusions yet, and in any case, performance has far too many unmeasurable or subjective variables to be based solely on statistics, which is where snowstat consistently goes wrong. It must also be based on observation and opinion.


User avatar
NewCorkSeth
Hob Nob Addict
Posts: 9519
Joined: 05 Jul 2013 00:17
Location: Wherever Nameless may be.

Re: Statistics show

by NewCorkSeth » 13 Aug 2013 13:01

PistolPete


Neuer
Dante
Van Buyten
Lahm
Alaba
Schweinsteiger
Muller
My 62 year old dad
Robben
Ribery
Mandzukic

Vs

Aldershots Team Vs Grimbsby

01 Morris
02 Oastler
05 Goodman
06 Webster
15 Barker
04 Stanley
07 Mekki
08 Rowlands
19 Molesley
09 Williams
10 Paterson

Who would win? Or have I got it wrong? Surely the idea is a bit more complex??

Better to rename the thread "Who is the weakest link"? But then again, it's a flawed question as it depends on formation etc!


Hmmmm well your 62 year old dad is most likely the worst player on the pitch sooo Grimbsby win 2-0.
Last edited by NewCorkSeth on 13 Aug 2013 13:16, edited 1 time in total.

User avatar
PistolPete
Hob Nob Regular
Posts: 1344
Joined: 01 Mar 2005 06:38
Location: 1871

Re: Statistics show

by PistolPete » 13 Aug 2013 13:14

Possibly the worst use of quoting ever.

User avatar
NewCorkSeth
Hob Nob Addict
Posts: 9519
Joined: 05 Jul 2013 00:17
Location: Wherever Nameless may be.

Re: Statistics show

by NewCorkSeth » 13 Aug 2013 13:17

Don't quite know how that happened..

paddy20
Hob Nob Regular
Posts: 1251
Joined: 18 Apr 2007 17:50
Location: Wokingham

Re: Statistics show

by paddy20 » 13 Aug 2013 14:07

PistolPete
paddy20
Colin Cheeselog So in order to improve, football teams should replace their poorest performers with players that will play better? This is a revolutionary new way of thinking.


Dear Football Association,

I have an idea that will change football forever...


Not sure that you understood. If the best player is slightly worse and the worst player is slightly better statistics show that you will get better results


Neuer
Dante
Van Buyten
Lahm
Alaba
Schweinsteiger
Muller
My 62 year old dad
Robben
Ribery
Mandzukic

Vs

Aldershots Team Vs Grimbsby

01 Morris
02 Oastler
05 Goodman
06 Webster
15 Barker
04 Stanley
07 Mekki
08 Rowlands
19 Molesley
09 Williams
10 Paterson

Who would win? Or have I got it wrong? Surely the idea is a bit more complex??

Better to rename the thread "Who is the weakest link"? But then again, it's a flawed question as it depends on formation etc!


No. In your example the %'s of the top team would add up to far more than Aldershot. In the example I gave they both added up to 105. I.e. its better to try and improve your weakest link when comparing teams of a close skill similarity

paddy20
Hob Nob Regular
Posts: 1251
Joined: 18 Apr 2007 17:50
Location: Wokingham

Re: Statistics show

by paddy20 » 13 Aug 2013 14:14

Ian Royal Statistically (in other words, not necessarily reliably) the two weakest performers so far, iirc, have been Blackman against Bolton (definitely) followed by ALF against Ipswich (I think). I'll check when I'm home.

But the sample size is really too small to draw any firm conclusions yet, and in any case, performance has far too many unmeasurable or subjective variables to be based solely on statistics, which is where snowstat consistently goes wrong. It must also be based on observation and opinion.


The sample size was massive in this example. I think they looked at all the European major leagues over a long period. I agree statistics can be made to mean what you want but where it is done on large scale using the pertinent mathematical theories it can be very accurate. This approach was used in Baseball in America to great effect and still is. Its only time before it is every day part of life in football in Europe. One of the greatest proponent in Uk is Martinez and see how Wigan have over performed for a number of years

User avatar
NewCorkSeth
Hob Nob Addict
Posts: 9519
Joined: 05 Jul 2013 00:17
Location: Wherever Nameless may be.

Re: Statistics show

by NewCorkSeth » 13 Aug 2013 14:15

Is that not what QPR tried?

User avatar
creative_username_1
Hob Nob Regular
Posts: 1728
Joined: 07 Aug 2012 20:43
Location: 3rd Place Music Comp

Re: Statistics show

by creative_username_1 » 13 Aug 2013 14:24

The downside with these models is that they are unable to (accurately) predict the future (the flaw of every model)

How are the %'s measured is it a relative thing to your team or an overall player rating (like a chess ranking)

Who is online

Users browsing this forum: Bing [Bot], Google [Bot], Jammy Dodger, Royals and Racers, WestYorksRoyal and 431 guests

It is currently 28 Mar 2024 08:51