| New Statistical Information about the Season 23:37 - May 21 with 2401 views | Davillin |
I ask you to please read these short paragraphs in order to understand some things about the following statistical information in two graphs based upon the same data.
The statistics attempt to show how different players contribute to the team's points total, based on number of games he played.
Some statistical studies can be iron-clad and literal, while some, by the nature of the data cannot. The latter, however, can have value in more of an comparison fashion, as do these. In other words, comparing players who are all measured by the same definition seems to level the playing field.
The statistical data here includes the number of games a player played, and the total number of points the team earned in those games in which he played.
At the outset, the statistician must decide what "a game" is — at one extreme it would be one in which a player played every minute, and at the other extreme one in which a player played even a single minute. The definition of "a game" for purposes of such a study would have to be somewhere between "every minute" and "even a single minute."
I've produced this study for a number of years and have laboured over this decision, something like this:
- more than 45 minutes probably is a bit draconian because experience tells me that a player can influence an outcome at fewer than 45. In my mind, that number is more like 30 minutes, so I have made a statistical decision that for purposes of this study, a game is "more than 30 minutes played."
Experience also tells me that scoring a goal or making an "assisst" is not even close to the sole criterion for "contribution," as a player in any position has as much to contribute, as the Swans' season has demonstrated.
The two graphs below are both based on the same data, with one arranged by "games played" and the other by "team points earned" [highlighted in yellow].
I give both graphs so that you can decide for yourself how much weight to put on the contribution to team output of a player who played only one game in which the team got all three points, compared, for example with a player who played closer to 38 games.
 |  |
| |  |
| New Statistical Information about the Season on 20:44 - May 22 with 1726 views | Davillin |
A brief analysis of the second chart, and my choice for most contribution to the team's points.
First, two observations. Any player whose games led to a PPG-P [points per games played] higher than the team's [1.08] can be said to have made a more positive contribution when he played. That is not to say that players with a PPG-P lower than 1.08] did not make a valuable contribution.
Then of those above 1.08, I think that the ones who played more games had more impact on points earned.
For me, then the contribution of Fernandez of 1.37 PPG out of 27 games played seems to me to be best, followed closely by Naughton, Mawson, and Sigurdsson. [Sigurdsson could count his higher number of 37 games played as something of a drag on his production.]
Not including Nordfeldt, the top three in the chart [Carroll, Olsson, and Rangel] plus Britton, each of whom had high PPGs on a low number of games played, also made valuable contributions. [Post edited 22 May 2017 20:45]
| |
| |
| New Statistical Information about the Season on 20:48 - May 22 with 1712 views | Nirvana |
Great stuff, Davillin.
It's an honor to be your alter ego.
Can we meet for a drink at some point? |  | | |
| New Statistical Information about the Season on 20:52 - May 22 with 1701 views | jack247 |
Am I reading it right that Montero has been involved in 13 games and we have lost them all? |  | | |
| New Statistical Information about the Season on 20:53 - May 22 with 1698 views | NeathJack |
Montero  |  | | |
| New Statistical Information about the Season on 20:58 - May 22 with 1677 views | Nirvana |
Neath_Jack, is your head exploding right now?
I'll take your non-answer as a yes. | | | |
| New Statistical Information about the Season on 21:00 - May 22 with 1653 views | Neath_Jack |
| New Statistical Information about the Season on 20:58 - May 22 by Nirvana |
Neath_Jack, is your head exploding right now?
I'll take your non-answer as a yes. |
Beth? |  |
| |
| New Statistical Information about the Season on 21:02 - May 22 with 1643 views | morningstar |
| New Statistical Information about the Season on 21:00 - May 22 by Neath_Jack |
Beth? |
Stop it. His head will explode!  |  |
| |
| New Statistical Information about the Season on 21:03 - May 22 with 1636 views | Neath_Jack |
| New Statistical Information about the Season on 21:02 - May 22 by morningstar |
Stop it. His head will explode! |
 | |
| |
Login to get fewer ads
| New Statistical Information about the Season on 21:08 - May 22 with 1616 views | longlostjack |
| New Statistical Information about the Season on 20:52 - May 22 by jack247 |
Am I reading it right that Montero has been involved in 13 games and we have lost them all? |
How many of those games has he come on as an "impact" sub when we were chasing the game though? |  |
| |
| New Statistical Information about the Season on 21:11 - May 22 with 1604 views | jack247 |
| New Statistical Information about the Season on 21:08 - May 22 by longlostjack |
How many of those games has he come on as an "impact" sub when we were chasing the game though? |
Most of them I'd guess. Hasn't made much "impact" has he? | | | |
| New Statistical Information about the Season on 21:35 - May 22 with 1562 views | longlostjack |
| New Statistical Information about the Season on 21:11 - May 22 by jack247 |
Most of them I'd guess. Hasn't made much "impact" has he? |
Can't argue with that. | |
| |
| New Statistical Information about the Season on 21:36 - May 22 with 1561 views | SPboy |
Very interesting Dav. If you were to break down the numbers by games played by manager, it is clear that the low scoring players i.e. Those who scored less than 1ppg played most of their games under Bradley |  | | |
| New Statistical Information about the Season on 21:55 - May 22 with 1529 views | grfn |
| New Statistical Information about the Season on 20:52 - May 22 by jack247 |
Am I reading it right that Montero has been involved in 13 games and we have lost them all? |
What?
thats not true, Montero played in 3 wins (Burnley - opening day, Sunderland at home, and Crystal Palace at home) and 1 draw (Man United away). | | | |
| New Statistical Information about the Season on 22:02 - May 22 with 1506 views | grfn |
ok, i see, he hasnt played more 30 minutes in those won or drawn games.
But why did you include those games in "#Games"?
Montero stats including ">30min" games are "5 games - 0 points", and including all of them are "13 games - 10 points".
Same with Borja, its "5 games - 1 point". | | | |
| New Statistical Information about the Season on 22:06 - May 22 with 1491 views | monmouth |
Football is a team game. |  |
| |
| New Statistical Information about the Season on 22:10 - May 22 with 1476 views | jack247 |
| New Statistical Information about the Season on 22:02 - May 22 by grfn |
ok, i see, he hasnt played more 30 minutes in those won or drawn games.
But why did you include those games in "#Games"?
Montero stats including ">30min" games are "5 games - 0 points", and including all of them are "13 games - 10 points".
Same with Borja, its "5 games - 1 point". |
In fairness to him, I was amazed that he had played 30 minutes or more in that many games. 5 sounds about right. | | | |
| New Statistical Information about the Season on 22:19 - May 22 with 1456 views | TheResurrection |
Of all the statistics and intuitive statistical analysis on offer these days, fair play to Tony Montana, his are by far the most pointless and boring.
A job at Opta awaits not. |  |
| |
| New Statistical Information about the Season on 22:27 - May 22 with 1429 views | Cooperman |
| New Statistical Information about the Season on 22:19 - May 22 by TheResurrection |
Of all the statistics and intuitive statistical analysis on offer these days, fair play to Tony Montana, his are by far the most pointless and boring.
A job at Opta awaits not. |
On the contrary, I've not seen his second league table presented elsewhere and I think it makes an interesting comparison with the seasons evolvement (Nordfeldt aside). |  |
| |
| New Statistical Information about the Season on 22:30 - May 22 with 1421 views | Davillin |
| New Statistical Information about the Season on 21:36 - May 22 by SPboy |
Very interesting Dav. If you were to break down the numbers by games played by manager, it is clear that the low scoring players i.e. Those who scored less than 1ppg played most of their games under Bradley |
Thanks for the insights, SPboy. Your point is one I had recognized, but did not include in the calculations simply because of the clear probability that the few players who contributed more while playing under one manager would have contributed more under the other two. Please note that I am not at all "disagreeing" with your point, or "arguing."
Second, the relative influence of managers on points gained is already on record.
Third, there is the factor that Guidolin was faced with more negative conditions [quality of opponents he had to face, international players returning late and in less than top condition, for example], whereas Bradley had an easier row to hoe, but did less. Given all of that, I think the playing field for all players tends to level out by those conditions, rather than becoming more "bumpy" in PPG in games played, especially when compared with what Clement faced, which had similar elements of both.
[p.s. I could actually do a pointed study of your last suggestion if someone thought it worth my while to do so.] | |
| |
| New Statistical Information about the Season on 22:45 - May 22 with 1390 views | TheResurrection |
| New Statistical Information about the Season on 22:27 - May 22 by Cooperman |
On the contrary, I've not seen his second league table presented elsewhere and I think it makes an interesting comparison with the seasons evolvement (Nordfeldt aside). |
Really? How?
It's just a bunch of numbers thrown together showing players with better points who played more games from around January. | |
| |
| New Statistical Information about the Season on 22:50 - May 22 with 1379 views | 34dfgdf54 |
| New Statistical Information about the Season on 22:19 - May 22 by TheResurrection |
Of all the statistics and intuitive statistical analysis on offer these days, fair play to Tony Montana, his are by far the most pointless and boring.
A job at Opta awaits not. |
Well said. |  | | |
| New Statistical Information about the Season on 16:41 - May 23 with 1214 views | Davillin |
| New Statistical Information about the Season on 22:45 - May 22 by TheResurrection |
Really? How?
It's just a bunch of numbers thrown together showing players with better points who played more games from around January. |
I now succumb to the everlasting temptation to show the above poster as a pitiful poseur as a "football expert."
During the period of plummeting down the Table before "around January," Bradley was the manager and he had about as little knowledge about football as the poseur above-referenced. He had no idea about the quality of football players in the Premier League, and couldn't find his best 11 if they wore glowing signs saying, "Pick me."
The fact that the poseur above-referenced does recognize that the players who did well in my study were all "players with better points who played more games from around January" should have tipped him off that they were better players and Bradley didn't recognize that.
It's called "cause and effect." The players weren't successful because they played "from around January," but the team was successful because those players played "from around January."
Class dismissed. [Post edited 23 May 2017 16:42]
| |
| |
| New Statistical Information about the Season on 19:10 - May 23 with 1141 views | pikeypaul |
Hindsight is a wonderful thing.
Back fitting stats to prove a point is a waste of time.
The change of manager in January contributed to the
wins and hence the extra points the players he picked
got. [Post edited 23 May 2017 19:13]
|  |
| |
| New Statistical Information about the Season on 20:24 - May 23 with 1061 views | Davillin |
| New Statistical Information about the Season on 19:10 - May 23 by pikeypaul |
Hindsight is a wonderful thing.
Back fitting stats to prove a point is a waste of time.
The change of manager in January contributed to the
wins and hence the extra points the players he picked
got. [Post edited 23 May 2017 19:13]
|
Thanks for the observation, Pikey, but if I relied on hindsight, your hindsight is not any better.
I didn't "back fit" anything — at least not any more than you did, above.
I agree that "the change of manager in January contributed to the wins and hence the extra points," but like the poseur, you erroneously mix cause and affect because you assume and imply that the previous manager would have done no better had he picked the players that Clement did.
You also use the word "contributed," but actually imply that the manager change "was the sole cause of" the difference.
In sum, you assume that the manager made the difference, whereas I assume that the players did -- or at least, that they share the "credit." | |
| |
| New Statistical Information about the Season on 20:30 - May 23 with 1057 views | Dr_Winston |
The manager clearly made the difference. |  |
| Pain or damage don't end the world. Or despair, or f*cking beatings. The world ends when you're dead. Until then, you got more punishment in store. Stand it like a man... and give some back. |
| |
| |
By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions.
