What's new

The Just-Past Half-Term Report

DC_Boy

New Member
May 20, 2005
17,608
5
hi BOF

good report

I'm glad you understand this bit, as you correctly say

'True, all you need to get in the top four is to have one more point (or better GD) than the team in 5th place,'

I was trying to explain this to some people on here, who couldn't/wouldn't grasp it -)

76 points would be nice, but we won't get that IMO

however on our current form 70 points is within reach, and that has been enough to finish above the 5th place team every season since 4 became the magic number

last year IIRC villa only got 63 points - so allowing for a reasonable distribution of the missing points Arsenal only needed about 68 points at most to get 4th place

this season it looks like the top 3 will run away from the pack (tho arsenal really aren't as good as their results show IMHO -but points is points etc etc)


it wouldn't amaze me to see 67 points being enough as the pack below fight it out

we are well capable of that

alas i think city will do us - and if we lose to pool on sat, i think they'll do us too

but we have a chance - anfield is the first of those chances- here's hoping for a result that says 4th is us (ie a draw or win)
 

Son_Of

SC Supporter
Aug 22, 2008
4,260
15
BOF: Four clean sheets on the trot. Not bad, eh? That brings the total to 6 so far this season, compared to last season when at this stage we had - er - 6. So again, perhaps we shouldn't run away with the idea that we've become Fort Knox at the back.
=============================================

but last season we kept more clean sheets than we had for ages. so to be keeping up with that target this season is good :)

please revisit this thread after the next 8 games. we're surely gonna get at least 8 points, so it should be happy reading
 

davidmatzdorf

Front Page Gadfly
Jun 7, 2004
18,106
45,030
Does anyone know how many points we got in 2008 (from Jan to Dec) and how that compares with other teams? I'd guess we'd be 5th.

Did you mean 2008 or 2009?

Between Jan and May 2009 we got 31 points from 18 games; between Aug and Dec 2009 we got 37 points from 20 games. So the answer for 2009 is 68 points from 38 games. That would generally correspond with 4th place over the course of an average season; 5 times, a team with 68-69 points has finished 3rd; once, a team with 69 points finished 2nd.

Between Jan and May 2008 we got 22 points from 18 games; between Aug and Dec 2008 we got 20 points from 20 games. So the answer for 2008 is 42 points from 38 games. That would generally correspond with 14th or 15th place over the course of an average season. A team with 42 points has finished 18th once.

...Quite simply, unless it's a freak season like 2004-2005, 70 points or as near as dammit should secure fourth. That's 1.84 PPG. Currently we're on 1.85, so on course and marginally better off than we were this time four years ago. What ultimately sunk our challenge then was a bad run of nine points from seven games; although we recovered, we couldn't make up that deficit sufficiently in the remaining 11 games—or not sufficiently so that we didn't need a win against the Spammers....

...76 points would be nice, but we won't get that IMO

however on our current form 70 points is within reach, and that has been enough to finish above the 5th place team every season since 4 became the magic number...

it wouldn't amaze me to see 67 points being enough as the pack below fight it out

we are well capable of that

In the history of the Premiership, every team that has obtained 67 or more points has finished 4th or higher. The maximum total for 5th place has been 66. It could be as high as 69 or 70 required to finish 4th this season, depending on whether the presence of a more competitive league drives the required total higher or lower (I can make a good case for either and I'm not sure which is correct, or even if there is a "correct" version).

But I'm finding it hard to see how the league's balanced schedule, with each team playing each other team twice, can permit five teams to finish with more than 70 points. It has never happened and the points totals that correspond with finishing places have been remarkably consistent since the league was reduced to 20 clubs.
 

Jenko

Well-Known Member
Mar 18, 2004
5,290
4,157
Statistics will never prove anything about the future. Its a lovely well thought out article - but as has been said - is only useful to a certain point.

I think the very best use you could make of it is to hand it to Harry and tell him to use it as a motivational tool for the next 8 matches. I'd imagine they are already aware of it but you never know. If I was a player I would put a lot of focus on it and it would only help continue the seasons pace.
 

smu01mhw

New Member
Jul 19, 2008
6
0
If we had 2 points after the first 8 games of last season, why does the graph show us at around the 18 point mark?
 

Jaispurs

Well-Known Member
May 11, 2005
733
351
If we had 2 points after the first 8 games of last season, why does the graph show us at around the 18 point mark?

You have an amazing memory. How did you know we only had 2 points from 8 games ? Where did you hear that ?
 

chrissivad

Staff
May 20, 2005
51,646
58,072
If we had 2 points after the first 8 games of last season, why does the graph show us at around the 18 point mark?


Because you didn't read the post, did you?

The graph is comparing the games we have played this season the the same games we played last season.

And as the first 8 games we played this season were not the same as the first 8 we played last season those first 8 games last season have been spread out over this season.
 

davidmatzdorf

Front Page Gadfly
Jun 7, 2004
18,106
45,030
If we had 2 points after the first 8 games of last season, why does the graph show us at around the 18 point mark?

Helps to read the article :wink:.

It's a graph comparing this year's matches against the matches against the same teams last year, so the 2008/09 matches are out of order.
 

Geoff

New Member
May 11, 2005
16
0
But I'm finding it hard to see how the league's balanced schedule, with each team playing each other team twice, can permit five teams to finish with more than 70 points. It has never happened and the points totals that correspond with finishing places have been remarkably consistent since the league was reduced to 20 clubs.[/QUOTE]

I have enjoyed reading this discusion, feel compelled to reply to the above with regards the 70 points issue. I feel we will need 71/72 points to finish 4th this season as the spread of points will be influenced by the fact that all 4 of the big 4 seem to have dropped more points this year than previously (as a group) and hence Villa, City, Liverpool and ourselves will attain more points than is normal for sides finnishing 4th 5th 6th and 7th.
 

robbiek

Member
Mar 3, 2006
98
97
I have enjoyed reading this discusion, feel compelled to reply to the above with regards the 70 points issue. I feel we will need 71/72 points to finish 4th this season as the spread of points will be influenced by the fact that all 4 of the big 4 seem to have dropped more points this year than previously (as a group) and hence Villa, City, Liverpool and ourselves will attain more points than is normal for sides finnishing 4th 5th 6th and 7th.

You are missing the obvious counter-argument -- that they will all take points off each other, so perhaps *less* points than usual will be required to obtain a CL place this year. I think that is what Davidmatzdorf meant when he said:

It could be as high as 69 or 70 required to finish 4th this season, depending on whether the presence of a more competitive league drives the required total higher or lower (I can make a good case for either and I'm not sure which is correct, or even if there is a "correct" version).
 

davidmatzdorf

Front Page Gadfly
Jun 7, 2004
18,106
45,030
I have enjoyed reading this discusion, feel compelled to reply to the above with regards the 70 points issue. I feel we will need 71/72 points to finish 4th this season as the spread of points will be influenced by the fact that all 4 of the big 4 seem to have dropped more points this year than previously (as a group) and hence Villa, City, Liverpool and ourselves will attain more points than is normal for sides finnishing 4th 5th 6th and 7th.

Your argument for requiring more points to finish 4th-6th is equally sustainable as an argument for requiring fewer points. If the clubs at the top of the league have fewer points than in previous years, then it should require fewer points to overtake one or two of them.

There is no reason to assume that the points that the "top 4" clubs will be losing will be lost to the 4-5 clubs who are trying to catch up to them. In fact, so far this season, the "top 4" clubs have tended to lose to the likes of Burnley at often as to the likes of Villa. The successful clubs have been less able to get consistent results against teams that previously would have been 3-point bankers. The league seems to be more competitive, from top to bottom, and the established CL clubs are not able to deliver consistent results.

You can "feel" what you like, but the records show that 67 points have always been enough to get 4th, sometimes 3rd. I can see how it might be a bit higher this year, but unless there is a huge, 10-15 point gap somewhere in the middle of the table, thus creating two widely-separated mini-leagues, with 8-9 clubs miles ahead of everyone else, I can't see how one can justify an argument that it will require 5 more points than ever before to finish 4th.

The balanced schedule doesn't produce results like that. I haven't checked all the tables comprehensively, but generally there is never a gap wider than about 5 points between adjacent clubs in the final standings.
 

davidmatzdorf

Front Page Gadfly
Jun 7, 2004
18,106
45,030
I haven't checked all the tables comprehensively, but generally there is never a gap wider than about 5 points between adjacent clubs in the final standings.

I have now. The results are very striking. The maximum gaps between adjacent places, since the league was reduced to 20 clubs, are very wide at the ends of the table and very narrow in the middle. There is a similar pattern to the average gaps. That confirms what one would expect from anedcdote and memory, which is that there are sometimes 1 or 2 clubs which are vastly superior or vastly inferior to the others. These outliers are bending the averages out of shape.

1st/2nd: Max 18, Avg 7
2nd/3rd: Max 15, Avg 5
3rd/4th: Max 16, Avg 7
4th/5th: Max 11, Avg 5

Between 7th/8th and 17th/18th, the maximum is 9 (once) and the averages are all 1, 2 & 3.

Then:

18th/19th: Max 16, Avg 3
19th/20th: Max 24, Avg 6
 

Geoff

New Member
May 11, 2005
16
0
You are missing the obvious counter-argument -- that they will all take points off each other, so perhaps *less* points than usual will be required to obtain a CL place this year. I think that is what Davidmatzdorf meant when he said:

Maybe Davidmatzdorf was right, but I was thinking they always take points from each other, didn't Fulham, Villa, Everton, take points from each other last year? This year Villa & City and maybe ourselves have taken points from the 'Big 4' plus each other. Hence less points to win the title and come second & third but more points to come 4th 5th & 6th & 7th?

After 19 games we had 34 points (City & Villa both on 35 points after 19 games) doubling this up puts us 6th if we only get 68 points. Hence I think we need 71/72 for 4th place.

I guess these are all mute points ultimately and their are many valid arguments, until we get to game number 38.
 

Geoff

New Member
May 11, 2005
16
0
You can "feel" what you like,
.

THANKS!!


Quite an aggressive response to somebody else's opinion. I thought I was entitled to an opinion of my own.

I suspect you will be a very rich man, come May, with your superior knowledge.
 

davidmatzdorf

Front Page Gadfly
Jun 7, 2004
18,106
45,030
Quite an aggressive response to somebody else's opinion. I thought I was entitled to an opinion of my own.

Wasn't meant that way, sorry, just being flip about the difference between feelings and figures. Should have put a :wink:.

I suspect you will be a very rich man, come May, with your superior knowledge.

Except that I don't bet...
 

minesadouble

Drove my Chevy to the Levy
Jul 27, 2006
749
2,933
Good article BOF and great thread everybody. I agree that the date v date comparison is more significant than the club v club one, but the former makes interesting reading.

What will be very interesting is the next 7-8 games. Looking at the fixture list, we have Liverpool next and then the run-in against the 'top 3' before the final match against Bolton. Frankly, all the matches in between are 'winnable' (although of course I'm not saying we will win them all). But we look well positioned and the end of the Season may be a good time to play the CL teams; and we have Chelsea and Arsenal at the Lane.

What I think puts us in a better position than for many years is the depth of the squad we have (compared with past Spurs and most current competitors). Our morale appears better than for years too with a greater sense of 'squad' than even in 2005/06.

What ultimately cost us in the Lasagna-gate Season was not just our poor run in Jan/Feb (mentioned above) but that Arsenal had the balls to win their final 3 games (after salvaging the 1-1 draw with us) when they had to take 9 points. I think it may go down to the wire again this Season but I have a feeling that this time we have the squad, quality and morale to win the crucial points, where in the past we've fallen short. But time will tell !
 

jolsnogross

Well-Known Member
May 17, 2005
3,730
5,431
the records show that 67 points have always been enough to get 4th, sometimes 3rd. I can see how it might be a bit higher this year, but unless there is a huge, 10-15 point gap somewhere in the middle of the table, thus creating two widely-separated mini-leagues, with 8-9 clubs miles ahead of everyone else, I can't see how one can justify an argument that it will require 5 more points than ever before to finish 4th.

It looks like this year could be anomalous, with the huge point gaps at the top and bottom not materializing. The top 3 may create a gap, but I wouldn't predict they'll be 10 or more points clear. It may turn out that the top points total this year wont be anywhere near 90 and may struggle to pass 80.

At the bottom, Pompey have had terrible results but are currently only 4 points adrift and then a 4 point spread separates 19th (Hull) from 11th (Everton).

I noticed a couple of weeks ago that the mid-table split you described above was developing - I can't find a table now but I think before the Xmas schedule it was 7 or 8 points between 9th and 10th, the largest gap in the league. It's now been cut back to 4 points between Fulham in 9th and Sunderland in 10th.

But because the spread of 4 points between almost all of the bottom-half teams, the scenario you outline of 8 or 9 top-half teams vastly superior (in points) could materialize. The top half have not bunched up as much as the bottom half though - we're 5 places but 10 points clear of Fulham, while Sunderland are 9 places and only 5 points clear of Hull.

It'll be interesting to see if the usual, fairly consistent pattern of Premier League points distribution can be bucked. And even more interesting to see if Spurs can stick with the pace - we're 6th in the form table (last 6 matches) at the moment; below City and Villa but just above Liverpool. Sunday could be pivotal (and it seems like an eternity away).
 

JimmyG2

SC Supporter
Dec 7, 2006
15,014
20,779
Davidmatzdorf

'It's just bullshit when people state that you can prove anything with statistics. It doesn't help to understand anything, it just throws a very important baby out with a few petty pints of bathwater. You can blow a load of smoke and obscure reality by misusing statistics. They're very vulnerable to exploitation by propagandists.'

It is not 'bullshit', because you can and people do prove whatever they want with statistics.The particular set of statistics used in the original article have been organised in a certain way to try to make a comparison.

Its an interesting method of comparing results from this to last season but because it presents the figures out of sequence it obscures the effect on performance of form and momentum.

There is no such thing as 'like for like' performances because of personnel changes in the two squads, weather conditions change and can effect the results; bad decisions, and injuries can further obscure the comparison., and a good result in say the first two or three games may effect the performance in the next.And visa versa.

A lot of bad science is based on the selective presentation of statistics, omitting negative results which are of no use to marketing agencies, selecting the sample to produce the desired effect and so on. But is there ever a set of complete statistical information and would it be useful if there was? As soon as you organise the figures you invite interpretation.

This has certainly happened here and an extremely interesting discussion has ensued. Proving 'anything' by statistics means exactly what you admit can happen, that the statistics are misused by selection.

'Pure' statistics are a mathematical fiction although I have been known to use and misuse them from time to time.

My seats of my pants tell me that we are having a very good season and I defy any statistics no matter how presented to persuade me otherwise,

I am writing this on behalf of the mathematically illiterate of whom I am ashamed to be a fully qualified member.:)
 

davidmatzdorf

Front Page Gadfly
Jun 7, 2004
18,106
45,030
It is not 'bullshit', because you can and people do prove whatever they want with statistics.The particular set of statistics used in the original article have been organised in a certain way to try to make a comparison.

I was getting at something specific about people, not something general about maths. When I see the hackneyed phrase "you can prove anything you want with statistics", or any variant on it, it is almost always backed by a hidden agenda, which is that numbers and statistics are useless. That's bullshit. They're highly useful and they can be used to learn things, through analysis and observation. They aren't the be-all and end-all, nor are they the gospel, and they have to be approached with a healthy degree of scepticism - but I mean "scepticism" in its true meaning, not as a synonym for "cynicism".

Saying "you can prove anything you want with statistics" is often code for "I'm frightened of numbers so I'm going to diss anyone who's using them", or worse, "I'm so cynical that I can't be arsed to make the effort to tell the difference between propaganda and careful analysis".

Well, I can tell the difference. One of the several ways I make a living involves designing, maintaining, developing and using extremely complicated spreadsheets and I don't want my carefully-designed and continually-under-improvement football spreadsheet dismissed with the phrase "you can prove anything you want with statistics". You can't. You can fool and manipulate some people with bad statistics. But that doesn't prove anything. It just shows that some people can't spot propaganda and bullshit.
 

ethanedwards

Snowflake incarnate.
Nov 24, 2006
3,377
2,498
Probably the most thought provoking and interesting thread on SC for many a day. Some great posts well argued and statistics analysed. We are in a great position league wise, no Europa Cup diversions and have an excellent squad of players. We need a little bit of luck combined with good stewardship at the helm to achieve a top four finish. We have had some great and good managers in the past twenty odd years, but as that old warhorse Patton said " Give me a lucky general anytime". Harry it is up to you, destiny calls make it happen.
 
Top