The Snowball stat thread

[Bandini]

This rather basic error, and the attempt to gloss over it, does call into question the wider issue of the integrity of the stats.

[Platypuss]

I'm not sure where we go from here. Like many other nobbers, I had factored in Snowball's various biases in the way in which he presented his stats, and the data that he chose to put into them, but I had assumed that underlying all of that was a basic mathematical rigour upon which we could rely.

This is all very alarming.

Yeah, right. Like higher maths is REALLY important in totalling
goals scored and minutes played.

LOL @ "higher maths"

[Wycombe Royal]

So if when you were at school you were taught that the earth was flat would you still believe that, even though it had been proved otherwise?

Just because you were taught something at school (and that was a lot longer ago than nearly everyone on here) it doesn't mean that what you were taught is still correct.

I've explained myself. Don't need to explain it again.

So why have you then?

I think you also have the belief that if you say something enough times eventually everyone will believe you (or are you just trying to convince yourself?)

[sandman]

Anyway, chat among yourselves. I'm interested in when strikers score.

Is it when Saturday comes?

[floyd__streete]

This thread is car-crash Hobnobbing at its best.

[Basingstoke Royal]

Church has scored 6 goals in 1,620mins this season and Le Fondre has 6 goals in 1,606.

The stats make it look like Church is almost as good as ALF.

In reality though, anyone can see that Le Fondre offers much more to the team than Church and a third of Church's goals were against a 9 man West Ham side that were already on the coach home.

[floyd__streete]

^ and that is why using stats to 'prove' your opinion is utterly, utterly bogus.

[Ian Royal]

[floyd__streete]

If stats and spreadsheets had any bearing on football results then I am off to put a tenner on Carol Vorderman as next England manager.

[Snowball]

Church has scored 6 goals in 1,620mins this season and Le Fondre has 6 goals in 1,606.

The stats make it look like Church is almost as good as ALF.

In reality though, anyone can see that Le Fondre offers much more to the team than Church and a third of Church's goals were against a 9 man West Ham side that were already on the coach home.

So apart from getting the minutes wrong, what other bits of the stats did you miss?

I have ALF as CLEARLY our best striker on stats, with a major contribution every 140 minutes played versus 210 for Church

If both players were to play 46 Games of 90 minutes that would put ALF on 30 Goals or Assists in the season and Church on 20

That's a pretty big difference IN FAVOUR OF ALF. And Church comes out in FIFTH best (not including Roberts who would make him sixth)

But, as usual, some sloppy HobNobber doesn't bother with the actual posted data.

Here is the table posted in "Attacking Stats"

Update after BRISTOL

1,543 Minutes 11 Major Contributions one every 140 Minutes 6 Goals 5 Assists ALF (RFC Games)
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


The other problem is, that neither has yet really taken off in terms of goals-scored so a single goal
(or in Church's case 2 goals v West Ham), MASSIVELY alter the stats. Church's 2 goals improved his
stats by 50% in 22 minutes.

And nobody (except you) says that goal-scoring minutes is the only criteria for assessing a striker.

The table has assists also (and I've had other HC's like winning a penalty or taking a large part in a goal)
and, although there's no proper stat for it (unless you have OPTA) I've also mentioned the quality of
interplay, and some brilliant, extra-quality passing from ALF

[Snowball]

If stats and spreadsheets had any bearing on football results then I am off to put a tenner on Carol Vorderman as next England manager.

Classic dumb error.

Nobody (especially me) says

stats affect the play
or
stats affect the game
or
stats affect the result

the stats are a semi-objective tool of ASSESSMENT, AFTER THE FACT


If this is too difficult a concept let me know and perhaps we
can limit discussion to "Completely Gash", so-so and "Good"

[Maguire]

If this is too difficult a concept let me know and perhaps we
can limit discussion to why 276/0 is not infinity no matter how many times I say it is

[Platypuss]

I don't consider this a case of right-wrong OR poor education" or "out of date" views.

I consider it a trivial difference of opinion over terminology.

If you're going to play Humpty Dumpty on as straightforward a matter as division by zero then what point is there in taking anything else you may say on face value?

[Snowball]

If you're going to play Humpty Dumpty on as straightforward a matter as division by zero then what point is there in taking anything else you may say on face value?

Because division by zero and such arguments are a lot different from
"How many goals did Shane Long score"

You could just as easily say, (if you were an atheist)

"You believe in God, so are irrational, so what's the point of hearing your views on football."

[Platypuss]

the point
|
|
500 miles
|
|
your post

[Extended-Phenotype]

Snowball; in response to your earlier post, I think you need to read what I said again as you have failed to grasp anything with any degree of comprehension.

Secondly, perhaps more simple for you to understand, as minutes are played against a different opponent, at different times of the game, under different instructions, in different formations, with different players… I fail to see how goals per minute has any real value as a statistic.

Thirdly though, you are incorrect about zero infinities.

Zero has no multiplicative inverse, so can’t be defined.

Quite simply, think of 1/x as x increases towards infinity. You get closer to zero, sure but as maximum infinity can’t be reached, neither can zero. Hence, the inverse of that equation will give an undefined result, not infinity.

And man, if asked for an honest show of hands of who in here googleraped thier answers it would be like Neremburg.

A bit. Well, not really. But there would be a lot of hands shooting into the air.
And p|ss being drunk.

[Ian Herring]

Snowball: a strange case.

That is all.

[Ian Royal]

Someone who is fundamentally incapable of admitting to being wrong when proved to be beyond doubt, is not worth engaging with on anything even vaguely subjective.

[Snowball]

Secondly, perhaps more simple for you to understand, as minutes are played against a different opponent, at different times of the game, under different instructions, in different formations, with different players… I fail to see how goals per minute has any real value as a statistic.

The three main strikers have:

almost identical total minutes played
almost identical minutes played when starting
almost identical minutes played coming on as a sub

and they have been on the pitch together an average of 20+ times in 20 games

Sorry to disappoint you again but the three main strikers have had virtually identical minutes
EQUALLY DISTRIBUTED across games. What's more these are when they have been on the pitch together

(It's actually MORE times than this but I didn't bother recording the odd ten minute patches)

Strikers on the pitch together (not including changeover times where overlap might be 1-5 minutes)

24 Alf with Hunt or Church
20 Church with Hunt or Alf
16 Hunt with Church or Alf


Ah well, perhaps it's starts versus sub appearances?

Err, NO, not that either. They are incredibly close.

24 Total Appearances 18 Starts 6 Sub Appearances Simon Church
22 Total Appearances 16 Starts 6 Sub Appearances Adam Le Fondre
24 Total Appearances 16 Starts 8 Sub Appearances Noel Hunt

And not "Total Minutes" either

1,469 Minutes Church
1,543 Minutes Le Fondre
1,513 Minutes Hunt

Big differences, huh?


And no significant difference in minutes when starting

1,384 Start Church
1,382 Start Le Fondre
1,342 Start Hunt

Wow, 42 minutes from highest to lowest!

And no significant difference in minutes as sub

085 Sub Minutes for 2 Goals = 043 Mins per Goal Church
161 Sub Minutes for 3 Goals = 054 Mins per Goal Le Fondre
171 Sub Minutes for 1 Goals = 171 Mins per Goal Hunt

Wow, an extra 86 minutes for Hunt, enough time to score a fifth of a goal.


Look, it's elementary. All three have had virtually identical minutes on the pitch
spread right across 28 games, virtually identical minutes played when starting,
and very, very similar times as a sub. On top of that the three possible pairings
have played together in between 60% and 86% of the games, against all teams.

(It's more than that actually)


Thus "minutes played" is EXTREMELY relevant and not biased in any way

[Snowball]

Someone who is fundamentally incapable of admitting to being wrong when proved to be beyond doubt, is not worth engaging with on anything even vaguely subjective.

Well oxf*rd off then blood sniffer