When Matejovsky Starts

[poohs pure]

is it just me or is snowball giving us newburyites a bad name

Oh I'm not REALLY from Newbury. I'm from Swindon.... (or not)

if you knew anything about reading football club you'd have said aldershot not swindon. go away now before you get even more annoying.

thread hijack....

yes or no, is snowball starting to become a cock?

[AlexY25]

is it just me or is snowball giving us newburyites a bad name

Oh I'm not REALLY from Newbury. I'm from Swindon.... (or not)

if you knew anything about reading football club you'd have said aldershot not swindon. go away now before you get even more annoying.

thread hijack....

yes or no, is snowball starting to become a cock?

I don't think he is, but he seems to annoy a few people on here.

[Platypuss]

Again Snowball your stats ignore the fact that in a full Premiership season, we don't play those same 10 games again and again. We don't get to play Derby or Birmingham 4 times each, and we do have to play teams like United, Arsenal, and Spurs, who Marek did not feature against. It is completely wrong to claim that '1.3 points per game' equates to anything over a season when the rest of the season does not mirror those 10 games.

JEEZ you are thick. I told you I did the 38 multiplier for not-very-clever people. (Hi!)

Marek's games DID however include 4th, 5th. 6th, 7th, 9th... five from the top half and five from the bottom

In a season we play 38 matches against an average position of 10.5.

Marek's games were against an average position of 11.6, MARGINALLY easier

04
05
06
07
09
13
16
17
19
20

11.6 average position

Any particularly reason why you assume that the "strength" of the division is normally distributed?

[Snowball]

I just get p!ssed with cmonurz inparticular, reading every post, responding in black and white, with the insight of a house-fly
and then I find myself having to deal in minutiae just to make a point.

I find it mind-numbingly stupid that people read those stats and decided I actually meant
yes we would get precisely 49 points.

I have posted maybe a dozen times and shown a points-per-game followed by "101 point season"

I did it yesterday when looking at form and odds. Nobody THEN thought I meant someone was going to get those points.



It seems to me that at least 95% of the list has a fixed view of every player, the manager, the coaches,
the price of burgers. It's fixed and woe betide anybody who challenges mere assertions and gut feelings.

This thread is a classic example. Marek staring in those ten games and we won four and had a positive GD
and when you look at the points achieved, the goal-difference achieved, it seems to me utterly clear
that a major reason for us going down last year was not playing Marek enough. We were relegated on GD.

Our home win v Liverpool was brilliant, but very unexpected. When Liverpool came back to 1-1 I fully expected 1-3
and not the other way round. Fantastic that we won but I do think it was a freak result.

Copps does seem to underplay some quality people. I wouldn't mind IF HE DIDN'T BUY THEM

[Snowball]

Again Snowball your stats ignore the fact that in a full Premiership season, we don't play those same 10 games again and again. We don't get to play Derby or Birmingham 4 times each, and we do have to play teams like United, Arsenal, and Spurs, who Marek did not feature against. It is completely wrong to claim that '1.3 points per game' equates to anything over a season when the rest of the season does not mirror those 10 games.

JEEZ you are thick. I told you I did the 38 multiplier for not-very-clever people. (Hi!)

Marek's games DID however include 4th, 5th. 6th, 7th, 9th... five from the top half and five from the bottom

In a season we play 38 matches against an average position of 10.5.

Marek's games were against an average position of 11.6, MARGINALLY easier

04 /05/06/07/09/13/16/17/19/20

11.6 average position

Any particularly reason why you assume that the "strength" of the division is normally distributed?

Funny, isn't it? When I talk about the difference between the top and bottom of the Championship I get beaten over the head with, "All the teams are the same."

We all know that there are three leagues in the Premiership. The top 4 (maybe top-five this year), the next 5-6-7 fighting for the Eufa spot and then the ten or so clubs whose main aim is to accrue 40+ points and stay up for another year.

That's why Coppell has said more than once that the results against the top four "don't matter" and that the results against the bottom 6-7-8-9-10 DO matter.

It's absolutely true we should take into account the fact that the top four last season were "a league apart" and we should not consider them when assessing stats, just presume we lose eight matches... but then we get a 0-0 at Old Trafford and beat Liverpool, and the year before got 2-2 at Chelsea

All I can do to compare is look at average position of the opposition (1 place different) or compare matches against the same teams. When we do that Marek looks a lot better. Take out the Liverpool results and the ga is much bigger

W4 D1 L4 Marek 13 points from 9 games
W2 D2 L5 Without 08 points from 9 games

[londinium]

Without reading every post am I right in thinking that we are better off points wise when Marek does NOT start?

1.75 points per game when he starts versus whatever it is this season when he doesnt start, which is surely better than 1.75

[cmonurz]

I just get p!ssed with cmonurz inparticular, reading every post, responding in black and white.

Annoying when people take your stats at face value, isn't it? More insults though, thank you.

You are the one who stated that we almost certainly would have stayed up had Marek played more games, and then used those 10 games as a basis for that opinion. I countered that those 10 games are in no way representative of our whole season, and so to extrapolate probable survival purely from those 10 games was, in my opinion, incorrect.

You then tried to give your stats a bit more meat by comparing those games to the reverse fixtures, which mean nothing at all in relation to Marek's starting/not starting. You are still doing it in your post above - 5 points difference in games Marek has/hasn't started. But what is that acutally showing? Not much at all - the reverse fixtures are completely different games. It's an illogical comparison.

It is you who chooses to take a stat-based approach to every argument. You can't keep throwing a hissy fit when people question those stats.

For the record, I know very well how stats work, I've used them for a long time. And yours sometimes don't work.

[Platypuss]

Snowball appears to have difficulties with grasping the difference between data and information.

[Snowball]

This is very funny. I taught stats to undergrads at Liverpool

[Snowball]

I just get p!ssed with cmonurz inparticular, reading every post, responding in black and white.

Annoying when people take your stats at face value, isn't it?

Stats at face-value? You never do! You immediately extrapolate and put word sin my mouth

You are the one who stated that we almost certainly would have stayed up had Marek played more games, and then used those 10 games as a basis for that opinion.

Yes. It's a basis, not a proof. It's a reasonable assumption because it's a random sample.

I worked for a couple of years for BRMB (The British Market Research Bureau) and 26.31% is a monstrously-big sample. One weakness of the stats is the fact that the games come in a block,
but is it a coincidence that Marek comes in when we have lost games, we lose another badly, then have two better results, though still defeats, and then start winning?

LOST 4-6 A Tottenham
LOST 0-2 H Portsmouth
LOST 1-3 A Aston Villa
LOST 0-2 H Man Utd
LOST 0-1 A Chelsea
MAREK GETS HIS CHANCE
LOST 0-2 H Bolton
LOST 0-1 A Everton
LOST 1-2 H Aston Villa

WON 1-0 A Middlesbrough
WON 2-0 H Manchester City
LOST 1-2 A Liverpool Marek scores)
WON 2-1 H Birmingham
Drew 0-0 H Blackburn (Marek sent off)

Is it "random" that we immediately lose badly 3-0 at Newcastle?

I countered that those 10 games are in no way representative of our whole season, and so to extrapolate probable survival purely from those 10 games was, in my opinion, incorrect.

And I have shown that the average league position was 11.6th when the EXPECTED league position is 10.5th. The difference is tiny

The 10, over 25% of 38 is a randomly-selected sample from 38. (I have to explain ever damn word now, but the "randomness" is because I can have no idea which games Marek has played in. That's Coppell's decisions.No sample is "perfect" but sampling does not normally use as much as 25% of the "population" to get its results.

You then tried to give your stats a bit more meat by comparing those games to the reverse fixtures, which mean nothing at all in relation to Marek's starting/not starting. You are still doing it in your post above - 5 points difference in games Marek has/hasn't started. But what is that actually showing? Not much at all - the reverse fixtures are completely different games. It's an illogical comparison.[/qoute

That's the best bit of squirming I've ever seen. I can't believe you can be so "illogical"

The absolute BEST comparison we can make is one where Reading play the opposite fixtures,
how could there be a better comparison. So if Ronaldo signs and we win eight games on the trot
against eight sides who beat us in the previous home/away fixture, you would say. "That doesn't
tell us anything about the presence of Ronaldo." That is stupid beyond belief.

You can't keep throwing a hissy fit when people question those stats.

I don't. I complain when they are deaf dumb and blind.

For the record, I know very well how stats work, I've used them for a long time. And yours sometimes don't work.

SOMETIMES?

[cmonurz]

26.31% (LOL - why can't you just say around a quarter like the rest of us?) is a reasonable sample size, when that sample is genuinely random. However, for the reasons I have alluded to in this thread, that sample is not random when that 26.31% is a block of fixtures in a football season. How statistically governed do you seriously believe football is?

You only have to look at one individual game to see the problems in comparing Marek's starting the match to the reverse fixture.

Let's take Blackburn.

20th Oct 07 Blackburn 4-2 Reading
29th Mar 08 Reading 0-0 Blackburn

You claim that it is perfectly logical to compare these matches as a basis for Marek's improving the team's results. This ignores the following:

- There were three other differences in the starting line-up between those two games. Murty, Sonko, Duberry and Fae replaced by Ingimarsson, Oster, Marek and Doyle. Given that we kept a clean sheet at home, maybe its Ingi who won us the point, and not Marek?
- Matejovsky was sent off after 72 minutes. We achieved the 0-0 with 10 men for the last 18 minutes of the game.
- Reading were at home for the 0-0 draw, and away for the defeat. It is generally an advantage to play at home.
- Blackburn came into the 4-2 win on the back of 4 straight wins. Their form was marginally worse coming into the game at the Madejski.

Now for me, much of this is bullshit. I wouldn't attach much significance to any of this except to argue your points re the effectiveness of Marek. You have gone back into the middle of last season, pulled together 10 matches, then simply said 'see, he's great' all on the basis on a marginal statistical improvement during his spell in the side, and irrespective of the nature of the games involved, who we played, what form they were in, who else played, etc.

Great data analysis, but your interpretation of it is one-sided. All I am doing is highlighting some ways in which that is the case. You can crunch numbers on this as much as you like, but ultimately appraising Marek's contribution is an individual thing, and most will do this quantitatively and qualitatively, and come to a conclusion.

And will you desist with your insulting me, it's getting tiresome. For an adult you have a real problem with people disagreeing with you.

[Ian Royal]

A thread where snowball talks shit and uses manipulated stats to pove his point.

Shocker.

[Snowball]

You two working in relays now?

Keep taking the tablets. They WILL work, eventually.

[Rex]

Malaria tablets??????

[Snowball]

26.31% (LOL - why can't you just say around a quarter like the rest of us?)

Because it's MORE than a quarter

is a reasonable sample size, when that sample is genuinely random.[/quote

It's random from my perspective,, as I didn't choose them.

However, for the reasons I have alluded to in this thread, that sample is not random when that 26.31% is a block of fixtures in a football season. How statistically governed do you seriously believe football is?

A fair bit. Which is why, for example only two sides in our league do less well (and fractionally)
against sides in the bottom half. And one of those is bottom the bottom club.
That's statistically predicted and actually happens.

You only have to look at one individual game to see the problems in comparing Marek's starting the match to the reverse fixture. Let's take Blackburn.

20th Oct 07 Blackburn 4-2 Reading
29th Mar 08 Reading 0-0 Blackburn

You claim that it is perfectly logical to compare these matches as a basis for Marek's improving the team's results. This ignores the following:

- There were three other differences in the starting line-up between those two games. Murty, Sonko, Duberry and Fae replaced by Ingimarsson, Oster, Marek and Doyle. Given that we kept a clean sheet at home, maybe its Ingi who won us the point, and not Marek?
- Matejovsky was sent off after 72 minutes. We achieved the 0-0 with 10 men for the last 18 minutes of the game.
- Reading were at home for the 0-0 draw, and away for the defeat. It is generally an advantage to play at home.
- Blackburn came into the 4-2 win on the back of 4 straight wins. Their form was marginally worse coming into the game at the Madejski.

In other words "IT IS IMPOSSIBLE TO COMPARE PLAYERS EVER." There is no way of judging players. If Doyle scores 30 in a season, it MIGHT be, because the service was better, the defenses were weaker, the goalies wee being bribed, the wind was in his favour...

You are making a classic error (why am I not surprised?) I do NOT take a single result, or two results, or three results or 4,5,6,7,8,9 results. I took TEN results, every game he started in. SURE, the Blackburn result might be tilted one way, but then another result might be tilted in the opposite way. These things balance out.

Now we scraped a 1-0 at home to Derby, but slaughtered them 4-0 away, so that's the opposite to what you've just said. (Now you bring in a new factor and say Derby were demoralised...)

Or how about Middlesboro? We only managed a 1-1 draw at home, but with Matty playing we WON away


So that's 2-1 to me. Add in Fulham, 02 v 1-3 is more or less the same.


Liverpool we played well up there, Matty scored, but we lost. Then we had a big result at the MadStad but that was home so we "should" do better than a 2-1 defeat.
Also Liverpool were awesome at home 12-6-1 43 13 and dodgier away from home 9-7-3 24 15

You could add in that the 3-1 home win was BEFORE MATTY HAD JOINED THE CLUB so how we played might have been different

Point is EVERY game has a load of variables pulling and pushing

but when you lump TEN GAMES together then the luck evens out, the weather evens out, sendings off even out etc

and 2 points and an 8-goal GD over ten games IS significant

Now for me, much of this is bullshit. I wouldn't attach much significance to any of this except to argue your points re the effectiveness of Marek. You have gone back into the middle of last season, pulled together 10 matches, then simply said 'see, he's great' all on the basis on a marginal statistical improvement during his spell in the side, and irrespective of the nature of the games involved, who we played, what form they were in, who else played, etc.

Um, I said "He's great"? WHERE exactly? I said he made the side play marginally better. I even used the word marginal, and .2 points is "marginal" but over a season it makes a difference. You seem also to have overlooked the fact that he was only eligible for 13 games from his full debut and he started in ten of them. So in that actual season, at the rate he was gaining points, he could have got us an extra .6 (POINT which of course cannot be done, it would be zero or one.

And of course, 4 starts this year, no defeats, again "randomly selected" the 1st, 22nd, 23rd, 26th DWWD = 2 points a game, better than the season's average 28 17-8-7 51 1.82 points a game

I can do no more than look at all his starts. Two different seasons, 14 different sets of excuses, yet he keeps that little plus-average... and if he starts tomorrow and we win it'll be DWWDW = 11 from 5 = 2.2 points per game

Great data analysis, but your interpretation of it is one-sided. All I am doing is highlighting some ways in which that is the case. You can crunch numbers on this as much as you like, but ultimately appraising Marek's contribution is an individual thing, and most will do this quantitatively and qualitatively, and come to a conclusion.

You think I don't think "qualitatively"? I didn't mention Karacan's equisite pass, or talk about (like many others)
of Marek's regular intention to play the "winning" ball rather than the safe one?

Do you think I statiistically measure "Noel Hunt is a gutsy player.."?


Point is. Someone could start a thread which says... We should play Marek, he's great, I love to watch him play, he's a Czech international, he scored against Germany, we would be much better off with him in the team, etcerabladeeblah

and people would come back and just troll out random negatives.

exactly like the Harps is shit, Harps is vital waste-of-space argument



The way to bring at least SOME rigour to arguments is to find ways to measure.


Let's use an example. If Marek had played those ten games (after a run of 5 defeats) and we won all ten, then he got an injury and we lost four on the trot
that would mean "without Marek" 9 defeats, "with Marek" ten wins.

SURELY, then you would see that the presence of Marek in the team is beneficial?


Well that's exactly what we had, except the improvement, real enough, wasn't that big.

[Platypuss]

This is very funny. I taught stats to undergrads at Liverpool

Are we supposed to be impressed?

[cmonurz]

Snowball lecturing about stats: Blah blah blah lecture stat stat blah
Student: But sir, what about.......
Snowball: What?! (stare of death)
Student: Nothing (pisses himself).

[Platypuss]

Let's use an example. If Marek had played those ten games (after a run of 5 defeats) and we won all ten, then he got an injury and we lost four on the trot
that would mean "without Marek" 9 defeats, "with Marek" ten wins.

SURELY, then you would see that the presence of Marek in the team is beneficial?


Well that's exactly what we had, except the improvement, real enough, wasn't that big.

Please show that this real improvement was down to Marek alone and is statistically significant.

[West Stand Man]

if you knew anything about reading football club you'd have said aldershot not swindon. go away now before you get even more annoying.

thread hijack....

yes or no, is snowball starting to become a cock?

No, I think you need to take a long look in the mirror if you want to see the cock in this instance.

[Ian Royal]

if you knew anything about reading football club you'd have said aldershot not swindon. go away now before you get even more annoying.

thread hijack....

yes or no, is snowball starting to become a cock?

No, I think you need to take a long look in the mirror if you want to see the cock in this instance.

Presumably it's a magic mirror with an image of snowball after being asked who is the biggest cock of them all.