The Snowball stat thread

[chilipepper91]

oh gawd

[themadstad]

Interesting Le Fondre's minutes per contribution compared Church's. Church still features more than Le Fondre though.

[Snowball]

Interesting Le Fondre's minutes per contribution compared Church's. Church still features more than Le Fondre though.

I think Alf has done a few more minutes (which I find hard to believe)
It may be that the minutes stats i got elsewhere are wrong or totaled wrong.

But Church's "started when we won" stats are good (10 from 13) so may be all that running around actually does do something.
I watch him and think a lot of it is ineffectual and that he lacks aggression, doesn't hurt opponents, but it's nevertheless a fact
that he started in 10 of our 13 wins, also scored twice as a sub when we won 3-0... so you have to say he has taken a big part
in 11 of 13 wins... And this is a guy who when I eyeball him think he's meh. Maybe McDermott knows what he's doing.

1,543 Minutes 11 Major Contributions one every 140 Minutes 6 Goals 5 Assists ALF (RFC Games) (EDIT, Minutes Updated)
1,513 Minutes 08 Major Contributions one every 187 Minutes 3 Goals 5 Assists HUNT (Set up Pen Chance for Roberts)
1,469 Minutes 07 Major Contributions one every 210 Minutes 6 Goals 1 Assists CHURCH




Update after BRISTOL
1,543 Minutes 11 Major Contributions one every 140 Minutes 6 Goals 5 Assists ALF (RFC Games) (EDIT)
0,446 Minutes 03 Major Contributions one every 147 Minutes 3 Goals 0 Assists MANSET
1,549 Minutes 09 Major Contributions one every 171 Minutes 2 Goals 7 Assists KEBE
1,513 Minutes 08 Major Contributions one every 187 Minutes 3 Goals 5 Assists HUNT (Set up Pen Chance for Roberts)
1,469 Minutes 07 Major Contributions one every 210 Minutes 6 Goals 1 Assists CHURCH
1,266 Minutes 06 Major Contributions one every 211 Minutes 3 Goals 3 Assists HRK
1,075 Minutes 05 Major Contributions one every 215 Minutes 0 Goals 5 Assists HARTE (INCL 2 half-assists)
2,115 Minutes 09 Major Contributions one every 235 Minutes 4 Goals 5 Assists McAnuff
1,936 Minutes 03 Major Contributions one every 655 Minutes 1 Goals 2 Assists KARACAN
0,745 Minutes 01 Major Contributions one every 745 Minutes 3 Goals 0 Assists GRIFFIN
2,620 Minutes 03 Major Contributions one every 871 Minutes 3 Goals 0 Assists PEARCE
0,876 Minutes 01 Major Contributions one every 876 Minutes 3 Goals 0 Assists TABB
1,089 Minutes 01 Major Contributions one every 1,089 Minutes 0 Goals 1 Assists MILLS
2,248 Minutes 02 Major Contributions one every 1,124 Minutes 1 Goals 1 Assists L-WOOD
1,964 Minutes 00 Major Contributions one every XXX Minutes 0 Goals 0 Assists CUMMINGS
2,159 Minutes 00 Major Contributions one every XXX Minutes 0 Goals 0 Assists HRK
=================================================================================
0,008 Minutes 01 Major Contributions one every 008 Minutes 0 Goals 1 Assists CYWKA (Won Penalty)
0,078 Minutes 02 Major Contributions one every 039 Minutes 1 Goals 1 Assists Roberts
0,244 Minutes 02 Major Contributions one every 122 Minutes 1 Goals 1 Assists CYWKA (RFC & Derby, League)
0,644 Minutes 04 Major Contributions one every 161 Minutes 1 Goals 1 Assists Roberts (incl Blackburn)
1,903 Minutes 14 Major Contributions one every 136 Minutes 10 Goals 4 Assists Alf (all league games)

[chilipepper91]

I almost can't be bothered to write this but what the hell, nothing better to do with my day.

"Form" isn't 22 games. "Form" is 5 games, 6 games, 8 games, 10 games at most. Players' "form" does not last half a season. If you're going to use such a huge sample size for your "form" table, why not just use the whole table

[themadstad]

The message Snowball is trying to convey is our first 6 games were a write off with us settling down with departures etc.

Now in my opinion we will not make up that difference. Had we, however, picked up points like we have since game 7 for the first 6, we'd be on 52 points.

So it seems like same old Reading to me tbh though lol.

[Snowball]

Why do you arbitrarily choose Sept 17th?

I've only written this post six or seven times in this thread.

I explained I think (again) on the previous page

We were in disarray in early season. We started with Long, lost Long.

We brought in Gorks Game 5. We brought in Alf Game 6

and we brought in Mills also. He came on for 8 minutes Game 5, started game 6

The side was a mess as a team. We were playing Khumalo, for God's sake!!

If you check back you will see that griffin was injured, came in, broke down. We even play Gunnarson
at RB and in those first 6 games we played 5 different FBs (and I vaguely think we played a winger or midfielders there briefly

Just check the first six team sheets

And also throw in the Wembley Hangover



From Game 7 we have been FAR more settled and gradually improving and our poor form
games 1-6 OBSCURES OUR GENUINE COMPETITIVENESS.

Our form "Since Gorks arrived" would be one reasonable form table.
Our form "Since Alf arrived would be one reasonable form table."

Starting at Game 7 is where we started to play our new players

It was Gorkss third-ever game for us, Alf's second, Mills second start

We were a long way from brilliant, but we had "turned the corner"

I started to record last 6.7.8.9.10 and so on and about once a week SOMEONE
will raise the same issue and I'll say, its X now and next game it'll be X+1

It's 22 games now and after Coventry it'll be 23. Half a season, the most recent 23 games

It is a perfectly valid statistic.

[Snowball]

The message Snowball is trying to convey is our first 6 games were a write off with us settling down with departures etc.

Exactly, and any NORMA:L person can seer that.

it's not exactly rocket science.


STRANGE, but when McDermott took over from Rodgers, nobody felt there was anything wrong
with writing off the first few games while he turned the ship round. Few blame the manager
for those early defeats and draws.

Only stupid people expect a manager to change a side in 24 hours.

[Snowball]

I almost can't be bothered to write this but what the hell, nothing better to do with my day.

"Form" isn't 22 games. "Form" is 5 games, 6 games, 8 games, 10 games at most. Players' "form" does not last half a season. If you're going to use such a huge sample size for your "form" table, why not just use the whole table

UTTER, UTTER rubbish.


So why do we talk about Arsenal's "Invincibles", or sides going a season undefeated at home,
or recently we were talking about Huddersfield's league run.

As long as someone uses a number of matches from 5 upwards and aren't
taking matches out in the middle of the run, that's fine.

It is a RELEVANT statistic that we are second in the table in points-per-game since September 17

If those games were the first 22/23 games of the season we'd be shouting from the roof-tops

[cmonurz]

The message Snowball is trying to convey is our first 6 games were a write off with us settling down with departures etc.

Exactly.

And yet we compare ourselves to the other clubs in the division who over our selected 22 game run may or may not themselves have been 'settling down' or struggling with injuries or departures, or a cup run, or a series of games against the top 6, or their best player getting a harsh red and a suspension or so on and so on.

[winchester_royal]

Quite, Snowball chooses that sample size because it produces the results that best illustrate his point. No other reason.

[Snowball]

The message Snowball is trying to convey is our first 6 games were a write off with us settling down with departures etc.

Exactly.

And yet we compare ourselves to the other clubs in the division who over our selected 22 game run may or may not themselves have been 'settling down' or struggling with injuries or departures, or a cup run, or a series of games against the top 6, or their best player getting a harsh red and a suspension or so on and so on.

WHAT? It's not "a selected 22 games" it's "THE LATEST 22 games, and increases by 1 every game."

It allows for all clubs "settling down, surely?"


And if you check you will find that IF YOU SHORTEN THE TIME SPAN OUR FORM IS EVEN BETTER

If I was "bending statistics" surely I would choose these later dates that make us clear top?

Since November 6th 9 wins in 13

1 13 9-0-4 20 12 +08 27 2.08 Reading

2 13 8-1-4 18 17 +01 25 1.92 West Ham
3 14 8-2-4 22 11 +11 26 1.86 Burnley
4 14 8-1-5 15 13 +02 25 1.79 Brighton
5 13 6-5-2 22 17 +05 23 1.77 Blackpool
6 13 6-5-2 17 13 +04 23 1.77 Cardiff
7 13 7-1-5 14 10 +04 22 1.69 Hull
8 14 7-2-5 22 17 +05 23 1.64 Leeds
9 16 7-5-4 28 14 +14 26 1.62 Birmingham

or Nov 26 (8 wins in 11)

1 11 8 0 3 5 2 18 10 +08 24 2.18 Reading

2 12 8 2 2 7 2 20 7 +13 26 2.17 Burnley
3 12 6 4 2 3 3 20 15 +05 22 1.83 Blackpool
4 12 7 1 4 5 4 13 10 +03 22 1.83 Brighton <<<<<
5 12 7 1 4 4 1 16 16 +00 22 1.83 West Ham
6 11 5 4 2 5 2 14 11 +03 19 1.73 Cardiff
7 13 6 4 3 4 3 25 11 +14 22 1.69 Birmingham
8 12 6 1 5 5 4 12 10 +02 19 1.58 Hull

This table is interesting in a worrying way. Look st Brighton. On the whole table not that much of a danger
but remember they were awesome at the start of the season, and over this last 12 games they are showing
high-play-off form. They could yet be a real danger.

Just as a matter of interest I think we'd've won at Doncaster (say 2-0)

The form table (Last 12 games, who is going to argue with THAT?)

1 12 9 0 3 6 2 20 10 +10 27 2.25 Reading
2 12 8 2 2 7 2 20 07 +13 26 2.17 Burnley
3 12 6 4 2 3 3 20 15 +05 22 1.83 Blackpool
4 12 7 1 4 5 4 13 10 +03 22 1.83 Brighton <<<<<
5 12 7 1 4 4 1 16 16 +00 22 1.83 West Ham
6 11 5 4 2 5 2 14 11 +03 19 1.73 Cardiff
7 13 6 4 3 4 3 25 11 +14 22 1.69 Birmingham
8 12 6 1 5 5 4 12 10 +02 19 1.58 Hull

If we'd drawn 1-1 ?


1 12 8 2 2 7 2 20 07 +13 26 2.17 Burnley
2 12 8 1 3 6 2 19 11 +08 25 2.08 Reading
3 12 6 4 2 3 3 20 15 +05 22 1.83 Blackpool

I we'd lost 1-0? We would still be second in a conventional 12-game form table

1 12 8 2 2 7 2 20 07 +13 26 2.17 Burnley
2 12 8 0 4 6 2 18 12 +07 22 2.00 Reading
3 12 6 4 2 3 3 20 15 +05 22 1.83 Blackpool

[Snowball]

Quite, Snowball chooses that sample size because it produces the results that best illustrate his point. No other reason.

Except that...

We are far better measured over the last 11 games (8 wins, top of table) or the last 13 games (wins)


In fact, if you look at the first 11 games of this run and the second 11 games of this run...


A 7 Reading 11 4-5-2 12 09 +03 17 Points 1.55 ppg. - Games 01-11 A step up from 0.67 ppg at the start
B 1 Reading 11 8-0-3 18 10 +08 24 Points 2.18 ppg. - Games 12-22 A HUGE step up from the above 11 games

C 2 Reading 22 12-- 5 30 19 +11 41 Points 1.86 ppg. - Games 01-22 All 22. My statistic of choice, NOT the one which shows RFC at their best

[Snowball]

You there, Winch?

[cmonurz]

It allows for all clubs "settling down, surely?"

Only if they've settled down in the same manner and at the same time as Reading FC. Some may have settled sooner, some slower. There will be teams who got more ppg in those first 6 games than any other set of 6 games this season (think Wolves in the PL, for example).

That our form is better if you shorten the time span isn't relevant to me. What these stats show isn't important any more so than a six-game form guide in this week's Sun, or an 8-game form guide on football365, but where those tables simply show all club's last 6 or 8 games, and that stays the same each week, your sample increases by 1 each week from your starting point of Reading FC 'settling down', and that for me makes the sample flawed.

[winchester_royal]

You use the 22 because you think that more games add extra credibility to your argument.

As cmonurz pointed out above, there are a number of issues with your stats. You are looking at it purely from the Reading point of view, while ignoring the mitigating circumstances facing other teams that may have led to our victories.

There's a reason that no sensible commentator uses a 22 game form guide. There are a number of variables involved in football, they can be limited over a 5 game period, but not 22.

[Snowball]

You use the 22 because you think that more games add extra credibility to your argument.

You people make me laugh.

I have X people saying "28 games is a better table"

Others saying "It must be 6 or 9" (cos THAT'S the only "reasonable form guide")


Listen. Listen VERY carefully. I am using THE CHAMPIONSHIP, EVERY GAME PLAYED
since September 17th. That is the greatest measure of points-gaining in the Championship
from Sept 17th 2011 to the current date, ever devised in the history of humanity.

It cannot be bettered. It is perfect. It is factual. It is real.

YOU may want games since August 6th. THEN GO AND oxf*rd LOOK THEM UP.
YOU may want the last 6 games. THEN GO AND oxf*rd LOOK THEM UP.
You may want the last 9 games. THEN GO AND oxf*rd LOOK THEM UP.
You may want the last 12 games. THEN GO AND oxf*rd LOOK THEM UP.
You may want only games played on Tuesdays. THEN GO AND oxf*rd LOOK THEM UP.

I want, and will continue to want, and will not change what I want,
that time span which I think is most representative.

I, me, myself, the person residing in my body, the person typing this
IS NOT GOING TO CHANGE HIS MIND. I do not (not even slightly)
give a monkey's oxf*rd what you think.

I have stated my reasons for choosing September 17th. They remain the same and will remain the same until the end of game 46.

It was the first time I thought maybe we had got over (psychologically) losing at Wembley
selling our Captain, and mostly, losing Shane Long. I have stuck to that date ever since
and posted probably 100 posts based on the games since that date. It was Gorkss 3rd start
and Alf's second start, and Mills second start, and Khumalo was OUT. 3/11 is a big deal.

May I suggest that every time I post a table since September 17th, every one of you posts thirteen posts
of complaint, writes to your MP, an protests in the Madejski Car Park. NOTHING-WILL-CHANGE

I've decided I'd like a REALLY LONG THREAD and would welcome your posts to extend it.

As cmonurz pointed out above, there are a number of issues with your stats.
You are looking at it purely from the Reading point of view, while ignoring the mitigating circumstances facing other teams that may have led to our victories.

No, I am looking at it from MY point of view.

But all the other clubs have played the same amount of games, in the same weather
and some have played more.

There's a reason that no sensible commentator uses a 22 game form guide. There are a number of variables involved in football, they can be limited over a 5 game period, but not 22.

Have you any idea exactly how many "correct" number of games = form guide you geniuses have suggested?

5.6.8.9.10.12.23.46 at the vary least

The longer, repeat LONGER a "form" guide gets, the LESS short-term variables have an effect.

Surely even you know that?

Five games could be against the bottom 5 clubs
and the opposition might have had two players sent off in every game

[Snowball]

It allows for all clubs "settling down, surely?"

Only if they've settled down in the same manner and at the same time as Reading FC. Some may have settled sooner, some slower. There will be teams who got more ppg in those first 6 games than any other set of 6 games this season (think Wolves in the PL, for example).

That our form is better if you shorten the time span isn't relevant to me. What these stats show isn't important any more so than a six-game form guide in this week's Sun, or an 8-game form guide on football365, but where those tables simply show all club's last 6 or 8 games, and that stays the same each week, your sample increases by 1 each week from your starting point of Reading FC 'settling down', and that for me makes the sample flawed.

it does?

I have a good idea.

We start a form table with ONE game, randomly, let's say "the first game of the season"

and we increase it, every time a club plays. 1 > 2 > 3 (you get the idea?)

Please select from the below which number of games will give the best indication of final league position

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46

You DO notice that this is a form guide with +1 every game do you not?

IMO a six-game form line is "interesting" but ONLY from a confidence POV
9-10-11-12 and we start to get a measure of the squad's ability. The shorter
the time-span the dodgier it gets. eg Ipswich 5-1 West Ham, so Ipswich are better than West Ham

[Snowball]

Winchester, you tried to show that I chose 22-games as that was the most-favourable
number of games.


BUT
IT
ISN'T

[cmonurz]

I think Snowball just had a mental breakdown.

Still, after polite responses from myself and winch, I think it's pretty evident now where Snowball sits in the grand scheme of this discussion board, simultaneously illustrating his failure to grasp what such a forum involves (it's not about trying to change your mind pal!) as well as showing the contempt he has for anyone who questions him.

I, me, myself, the person residing in my body, the person typing this
IS NOT GOING TO CHANGE HIS MIND. I do not (not even slightly)
give a monkey's oxf*rd what you think.

What a total asshole.

[winchester_royal]

Well then use the whole 28 games.

You can't just base your decision based on when Reading started playing well it's ridiculous.

Next you'll be telling us the Palace game doesn't count because we had a good goal ruled out, or that these last few games are void because Ledge hasn't been playing.

And no, that's a cliche that is just wrong. Variables do not always equal out in football, and using a smaller sample ensures that the form guide gives a good account of the current state of those variables, and hence a better idea of what will happen in the next few games.