Topic: Stats, Sambermetrics & Liverpool.

[BMW]

So as we all know John Henry is a fan of the use of Sabermetrics but what is it excatly?

Sabermetrics is the mathematical and statistical analysis of baseball records. To understand the field of sabermetrics, one first should be familiar with the game of baseball. The game is played between two teams, each consisting of nine players. The nine players are a pitcher, a catcher, first baseman, second baseman, shortstop, third baseman, left fielder, center fielder and right fielder. It consists of nine innings. One inning is divided into two halves; in the top half of the inning, one team plays in the field and the second team comes to bat, and in the bottom half, the teams reverse roles. The team that is batting during a particular half-inning is trying to score runs. The team with the higher number of runs at the end of the nine innings is the winner of the game.

There are many forms of data that can be taken away from any given match.

Now back to football.

Passing and Tackling Stats from the Man United 3-0 in March 2011

Kuyt: Passes attempted 47, successful 33. Tackles attempted 13,  won 4.
Luis Suarez: Passes attempted 45, successful 28. Tackles attempted 12, won 10.

25 attempted / 14 successful @ a 57% success rate. Kuyt and Luis made 37% of the total tackles attempted.


Lucas: Passes attempted 56, successful 47, Tackles attempted 10, won 1.
Gerrard: Passes attempted 56, successful 42. Tackles attempted 9, won 5.
Maxi: Passes attempted 34, successful 27. Tackles attempted 5, won 2.
Raul: Passes attempted 29, successful 24. Tackles attempted 2, won 1.

26 attempted / 9 successful @ a 35% success rate. Lucas, Gerrard, Maxi and Raul made 39% of the total tackles attempted.


Skrtel: Tackles attempted 2, won 1.
Carragher: Tackle attempted 5, won 4.
Johnson: Tackles attempted 8, won 5.
Kyrgiakos: Tackles attempted 2, won 1.

17 attempted / 11 successful  @ a 65% success rate.

Skrtel, Carragher, Johnson and Kyrgiakos made only 25% of the total tackles attempted.

Tackling Stats.

We defended from the front and the main front two attacking players in Suarez and Kuyt together attempted nearly as many tackles as the four midfield players in Lucas, Gerrard, Maxi and Raul. They (Kuyt and Suarez) made more tackles than all the defenders combined. Also the success rate for tackles attempted by Kuyt and Suarez was higher than the success rate for tackles attempted by Lucas, Gerrard, Maxi and Raul, which if you follow general opinion that attacking players (strikers and wide forwards) have poor when it comes to tackling and scoring goals is there job not breaking up play in the opposition's final third - it goes against general opinion at the very least.

I have the stats for Man United's attacking two in Berbatov and Rooney (I gave them the benefit of the doubt and put down Hernandez's stat's) and it's laughable compared to that of Kuyt and Suarez's and that's a understatement... here they are:

Rooney - Passes attempted 45, successful 30. Tackles attempted 5,  won 2.
Berbatov  - Passes attempted 30, successful 24. Tackles attempted 3, won 1.
J Hernandez's 45 mins - Passes attemped 14, successful 10. Tackles attempted  2, won 0

So that's 10 attempted tackles / 5 successful @ a 50% success rate between three players!

Also when you consider out back four only needed to make 11 successful tackles compared to the 25 successful tackles the players infront of them made it again show how much work was put in to winning the ball back as high up the pitch as possible.

Man United's back four was forced to attempt to make a total of 24 tackles succesful tackles, thier players infront of them made a total of only 7 successful tackles. In other words we put the hammer down from the word go and they folded, we went thur them with ease it was only due to good defending by Man United's back four we didn't beat them 5 or 6 nil with ease...

Smalling - Tackles attempted 10, won 8
Rafael- Tackle attempted 6, won 2
Evra- Tackles attempted 11, won 10
Brown - Tackles attempted 7, won 4

@ a success rate of 71%, which is pretty good considering the pressure we put them under.

He didn't play against Man United but...



Newcastle 3-1 Liverpool: Liverpool unable to cope with aerial power of Carroll, you understand where going with this...

Passing Stats.

Even though Man United's midfield wasn't winning the ball back when they lost possession of the ball they held onto the ball pretty well without being a goal threat bar the odd attempt, Carrick attempted 85 passes, with 66 being successful. Scholes attempted 72 passes, with yes 66 being successful. Giggs attempted 58 passes, with 40 being successful. That said Kuyt and Suarez attempted and completed more passes than Berbatov, Rooney and Hernandez. So just looking at both the stats and the way everything panned out we were more focused on closing them down as far up the pitch as possible and once we got possession we moved the ball forward to Suarez so he could link up with Kuyt and other players making runs from deep, rather than passing the ball side ways keeping possession of the ball.

Because we had our attacking players with their heads down winning the ball back on a regular basis, to any extent it didn't matter if we lost the possession of the ball. In any case Lucas attempted 56 passes, 47 being successful @ 84% success rate. Carrick attempted 85 passes, with 66 being successful @ a 77% success rate. Only Scholes' with passing success rate @ 90% being higher but again I'd have it a  guess quiet a few of those passes were side ways due to the fact that our attack received the ball more and were  a lot more of a goal threat.  Also if  you are winning possession in the opposition's half 9 times of 10,  logically it would take you less passes to get the ball into the  opposition's final third.


Lastly General Goals Stats.

They are a bit easier statstically breakdown if you like, personally for me the way they were scored is far more interesting than the shots on compared to off target as all they show is that the striker was having a bad day or season depending on you look at it. Take the Torres transfer, Abramovich decided to pay the amount he did based on how successful Torres was since he signed for Liverpool. I've heard that in the time Torres played under Rafa he had the highest goals per minutes on the pitch ratio in the history of the Premier League history, higher than Ruud Van Nistelroy and Henry who are rated as the best ever strikers in the history of the Premier League by many.

What Abramovich should have been looking at is how he scored those goals and what type of support he needs to score them. It's obvious that if your a striker who scores a lot your goal's from quick counter attacks that lead to the type of through balls Gerrard used to provide Torres on a regular basis, you can't expect that same striker to score at that rate in a team that looks play a slow build up, possession based game - see Torres lack off success with the Spain National Team. I read some Torres score 40% goals from through balls where he could use his pace to beat defenders, some may say that it was a means to an end (scoring 33 goals in his 1st season with us) other might call it being one dimensional unlike other striker's... say Luis Suarez.



Luis Suarez's 49 goal at Ajax 2009/2010

Penalty: 10
Freekick's: 2
1 on 1's with the goalkeeper, thur balls or over the top: 5
Out side the box: 5
Receiving the ball inside the box: 11
Individual effort (something out of nothing): 5
Header: 3
Tap in's, poacher goals, follow ups: 8

(I'm not saying I'm 100 percent correct.)

It's clear that he can do it all and that isn't an understatement, I didn't realize how many poacher type goals/following up the previous shot type goals he scored. He really does have it all in his locker and that's just his goal's he scored. If you go by the fact he won 10 attempted tackles against Man United which is equal highest amount tackles won along with Evra it goes to show he is a very, very hard worker and at the same time capable of moments of magic. It seems there is a lot to Suarez's game whilst he was playing for Ajax, he provided them with a cutting edge to their 4-3-3 possession based passing game which I assume they played whilst Suarez was at Ajax...  that's what most interesting about a player like Suarez, they are not just goal scorers but equally able to create goalscoring chances for their team mates and themselves. In Suarez's case prior to signing for Liverpool he had a scoring rate of a goal every 1.78 games.

Now for me I find these statistic's very interesting, if you go by terms used when describing Sabermetrics you would file these stats under situational data.

...match situations like the hitting performance of batters is recorded for a number of different situations, such as day versus night games, on grass fields and artificial turf fields, against pitchers who throw right handed and left handed, and during home and away games.

[mulhiem]

Thats all fine but there is much more to a game of football than passes and tackles - how do you measure zonal positioning? Where the attacking team dont have an positional output because of the zonal positioning of a defensive player?

[stevedo]

Nice thorough OP. It was my understanding that other principles included manipulating the acceptable stats, ie a pitcher getting an easy innings at the end to boost his perceived worth, and having a punt on bad penny players who were undervalued.

[nocturnalvin]

I know Royhendo would love this though.

[ThePrufeshanul]

Interesting read - many thanks for posting.


Despite the fact thst the whole sport is based on statistics - goals for, goals against, points won at home, points away etc. - no doubt you will have the usual brigade moaning about how statistics cant be applied to football and you should watch the games / players with your own eyes instead.

Though it should of course be obvious that no two people see things exactly the same way.

The answer then is to improve the resolution of your statistics not discard them completely.

Sabermetrics is the future.

[BMW]

Nice thorough OP. It was my understanding that other principles included manipulating the acceptable stats, ie a pitcher getting an easy innings at the end to boost his perceived worth, and having a punt on bad penny players who were undervalued.

http://www-math.bgsu.edu/~albert/papers/saber.html

Your right, not the best title I suppose... should have been Stats & Liverpool.

[BMW]

Thats all fine but there is much more to a game of football than passes and tackles - how do you measure zonal positioning? Where the attacking team dont have an positional output because of the zonal positioning of a defensive player?

Of course.

You just can't go into detail the of the whole picture, without  it being a book.

Also it's  not really a case of stats = success but the right stats =  can get you ahead of the pack.

[rappcats]

I consider myself a fan of statisttics graphs in football, but you can't base your whole opinion or compare players just on stats. Sabermetrics work nice in baseball, because of the nature of the game, in football not so much especially considering off ball movement and positioning.

good OP still, nice read

[djschembri]

The key to Sabermetrics was to identify which stats are meaningful in order to determine the quality of a player.

For example, in your post you mentioned the passes success rates. However that stat alone does not determine the effect the player had on the game. If half of the passes were made between defender and defensive midfielder, then they can be classified as not being important.

However what would be more interesting is the number of successful passes were made to a player in the final third. A high passing rate here would be of value, as it would mean that the player has the ability to find another player in the most dangerous part of the field, and were most of our moves break down.

[nocturnalvin]

Great effort but sorry dude, doesn't say anything new.

[mulhiem]

Also it's  not really a case of stats = success but the right stats =  can get you ahead of the pack.

Some of the right stats cannot be measured leaving you with a slightly warped statistical impression of the game.  Most of the game is played off the ball.

The only stats that get you ahead of the pack are the ones that prove the goals for are higher than the goals against.  Measuring everything else is futile.  I really couldnt give a rats ass if Lucas made 100 completed passes and 100 completed tackles if we get beat 3 nil.

[andy in warrington]

I think I'll leave this for other people to understand. I'll just watch the match.

[BMW]

The key to Sabermetrics was to identify which stats are meaningful in order to determine the quality of a player.

For example, in your post you mentioned the passes success rates. However that stat alone does not determine the effect the player had on the game. If half of the passes were made between defender and defensive midfielder, then they can be classified as not being important.

However what would be more interesting is the number of successful passes were made to a player in the final third. A high passing rate here would be of value, as it would mean that the player has the ability to find another player in the most dangerous part of the field, and were most of our moves break down.

I figure that a high tackling rate by our attacking players meant the ball was contained in the opponents half of the pitch, I might be wrong but it makes sense. Like if the back 4 were doing alot of the defensive work with the players infront ball watching but at the same time keeping possession of the ball it's safe to say it's counter productive. As for finding how many successful passes into the final 3rd were completed that would be a valuable stat I agree... but that's the problem it's know what stats are the valuable ones that's the key.... make of it what you will.

[frikkibj]

there is a great thread about this here: http://www.redandwhitekop.com/forum/index.php?topic=264503.0

everyone that hasn't read it should do so

it made me understand and accept the Torres transfer at the time

[nocturnalvin]

Some of the right stats cannot be measured leaving you with a slightly warped statistical impression of the game.  Most of the game is played off the ball.

The only stats that get you ahead of the pack are the ones that prove the goals for are higher than the goals against.  Measuring everything else is futile.  I really couldnt give a rats ass if Lucas made 100 completed passes and 100 completed tackles if we get beat 3 nil.

Disagree on that. At least it means Lucas is having a proper game.
The complication of stats in football presents the headache.  The qualitative stuff is hard to quantify.

[nocturnalvin]

there is a great thread about this here: http://www.redandwhitekop.com/forum/index.php?topic=264503.0

everyone that hasn't read it should do so

it made me understand and accept the Torres transfer at the time

Precisely which part does it make u understand the El Ninny transfer ?

[Danyaals Kop]

Interesting

[djschembri]

I figure that a high tackling rate by our attacking players meant the ball was contained in the opponents half of the pitch, I might be wrong but it makes sense. Like if the back 4 were doing alot of the defensive work with the players infront ball watching but at the same time keeping possession of the ball it's safe to say it's counter productive. As for finding how many successful passes into the final 3rd were completed that would be a valuable stat I agree... but that's the problem it's know what stats are the valuable ones that's the key.... make of it what you will.

I agree. If I remember correctly there was an article from a professor from Leeds Uni around the time of the FSG takeover, regarding how statistics may apply to football ... Was an interesting read too

EDIT: this is the link: http://www.independent.co.uk/sport/football/premier-league/secret-to-beanes-success-study-of-data-that-no-one-else-had-thought-to-study-2137957.html

At the end of the article it mentions some of the useful stats. It mentions that missed tackles stats may be very useful, whereas pass completion are not as effective, and mentions Fabregas as an example.

[BMW]

there is a great thread about this here: http://www.redandwhitekop.com/forum/index.php?topic=264503.0

everyone that hasn't read it should do so

it made me understand and accept the Torres transfer at the time

I just ordered the book... just waiting on it..

[xerxes1]

The increasing use of statistics to measure a player's value is inevitable, and welcome.

The transfer regime under G&H was a shambles, they had no idea, Rafa had no guidance,that he did so well with the majority of his key purchases wasa tribute to the man, it was a recipe for disaster and the bloated squad of not good enough players we now have.
The appointment of a DOF is the right move, whether Le Danny is the right appointment remains to be seen.

[nocturnalvin]

Well maybe...just maybe, we RAWKites should take the initiative and crack our brains over what can and should be measured.

How about "involvement" or "direct contribution to goals" ?
Off the top of my mind is instead of purely assists, how about involvement in, say, 3 passes before each goal.
Or % of passes in relation to the team for that game.

There are more i'm sure.

[xerxes1]

There are some basics to be addressed. Long expensive contracts to thirty something players are unwise (Carra)  annual contracts are fine.

Long contracts to injury prone players who even when fit at the end of theor contract, have little resale value ( AA) are similarly unwise.

[Greg]

Some of the right stats cannot be measured leaving you with a slightly warped statistical impression of the game.  Most of the game is played off the ball.

The only stats that get you ahead of the pack are the ones that prove the goals for are higher than the goals against.  Measuring everything else is futile.  I really couldnt give a rats ass if Lucas made 100 completed passes and 100 completed tackles if we get beat 3 nil.

It doesn't really work like that though does it? If you look at any football match in isolation, then the statistics won't prove much. But look at the collated statistics over 38 games and you should see things very differently. You'll probably find that the the team who have had the most shots, passes, tackles and possession etc are the ones who have just won the league.

As usual, some people see the word "statistics" and come in to say that statistics are a load of bollocks. I suspect that the people who say this just struggle to understand the concept of statistics. As yet, I've never met anyone who has a good understanding of statistics to believe they are worthless and irrelevant bollocks.

[jaffod]

The only stat I'm interested in is how much money Henry & Co are going to cough up in the summer.

[BMW]

Well maybe...just maybe, we RAWKites should take the initiative and crack our brains over what can and should be measured.

How about "involvement" or "direct contribution to goals" ?
Off the top of my mind is instead of purely assists, how about involvement in, say, 3 passes before each goal.
Or % of passes in relation to the team for that game.

There are more i'm sure.

In terms of match 'contributions', more value given to a 'match' winning 'contribution' rather than a 'contribution' that helps round off a 3-0 win?

It's like Kuyt doesn't score often but when he does they are often Key goals, I'm sure there are midfielders who play well against the smaller team or in easier match and go hiding against the bigger teams in the more bigger matches... to have stats that could weed out these types of players would be very useful.. It like how we often call Man United flat track bullies and we found we find playing smaller teams often tricky, I'm sure I've seen a table Liverpool vs the bottom 10 teams in the league that showed this, along with the goals scored/conceded before..  stuff  like that is interesting but no doubt others out there know more about stats in football than me haha, I'm just an active supporter...

[DanA]

People have to remember these stats are not taken in isolation. It's a case of finding good candidates like Suarez and figuring out if statistically he's better than the scout first thought.

[Malaysian Kopite]

Good read.

[its cold in the stands]

It doesn't really work like that though does it? If you look at any football match in isolation, then the statistics won't prove much. But look at the collated statistics over 38 games and you should see things very differently. You'll probably find that the the team who have had the most shots, passes, tackles and possession etc are the ones who have just won the league.

As usual, some people see the word "statistics" and come in to say that statistics are a load of bollocks. I suspect that the people who say this just struggle to understand the concept of statistics. As yet, I've never met anyone who has a good understanding of statistics to believe they are worthless and irrelevant bollocks.

but what do the stats prove that the league table doesnt over those 38 games?
i think stats are useless in footy, i`m just thankful they werent around years ago, for instance kenny dalglish`s pass completion rate would often be shite compared to most players now because dalglish was always trying to unlock defences with imaginative passes.
quite often the whole ground would applaud a pass that didnt quite come off because they could see it took incredible vision to even spot that the pass was on, and yet that would go down as a negative stat against the player, until stats can understand the context of a pass or even a tackle then they wont work.
jamie redknapp and john barnes when they were in center mid together must have re-wrote the stat records but what did they win as a midfield partnership?

[carling]

What if one tackle is a combination of excellent reading of the game and exquisite timing to prevent what would otherwise be a certain goal.  Then another is when the opponent takes a bad touch in a nothing area of the pitch and the player gets a lucky toe in to put it out for a throw.

Do they both count as the same?

[SP]

It doesn't really work like that though does it? If you look at any football match in isolation, then the statistics won't prove much. But look at the collated statistics over 38 games and you should see things very differently. You'll probably find that the the team who have had the most shots, passes, tackles and possession etc are the ones who have just won the league.

As usual, some people see the word "statistics" and come in to say that statistics are a load of bollocks. I suspect that the people who say this just struggle to understand the concept of statistics. As yet, I've never met anyone who has a good understanding of statistics to believe they are worthless and irrelevant bollocks.

But statistics are only useful if interpreted the right way. Any set of statistics will have flaws in them. To draw meaningful conclusions from them requires an understanding of their limitations.

As an example take the tacking statistics? The key to understanding them would be to understand what is a successful tackle. For instance Kuyt attempted 13 tackles but won 4. How many of the unsuccessful tackles resulted in the opposition player having to play a hurried ball, that put the receiver in trouble? Stats will only measure what they are designed to do. They can certainly tell a lot about a players contribution, but they are still just a tool. In the wrong hands, they can be horrible. Ask the FA.

[SP]

What if one tackle is a combination of excellent reading of the game and exquisite timing to prevent what would otherwise be a certain goal.  Then another is when the opponent takes a bad touch in a nothing area of the pitch and the player gets a lucky toe in to put it out for a throw.

Do they both count as the same?

It depends upon the stats. You could design a system that measures the importance of the tackle - and then you would be able to differentiate. A system that just records the number of tackles would be of limited utility. But I suspect JWH would be looking at something far more sophisticated...

[its cold in the stands]

What if one tackle is a combination of excellent reading of the game and exquisite timing to prevent what would otherwise be a certain goal.  Then another is when the opponent takes a bad touch in a nothing area of the pitch and the player gets a lucky toe in to put it out for a throw.

Do they both count as the same?

exactly mate, there`s no context, a player could make up yards of ground on an opponent clean through on goal, put in a tackle and lose the tackle but force the opponent wide or force him to take another couple of seconds to get the ball under control again and in the mean time the rest of the defence has got back.
is that `lost tackle` a negative stat that should go against the player or a tremendous piece of defensive cover?
it`s like assists as well, quite often the best pass in a move isnt the one just before the goal is scored.
the game is too fluid to turn into a science.

[its cold in the stands]

It depends upon the stats. You could design a system that measures the importance of the tackle - and then you would be able to differentiate. A system that just records the number of tackles would be of limited utility. But I suspect JWH would be looking at something far more sophisticated...

but that is just tackles, you`d have to have similar systems that accurately measure everything in the game from the distance covered by players to wether a pass was neccessary or if the player had failed to spot a better pass that was on, and so on and so on.
you could be there for months analysing a single game when it will probably just confirm what you thought at the time anyway.
traditionally the biggest champions of these stats tended to be the long ball merchants, they are all made up that over half the goals in the league come from set pieces or whatever the stat is, i`d rather the sport remained unscientific and watchable myself.
these negative coaches who champion tactics have already turned pele`s `beautiful game` into boring stalemates (world cups for example), we dont need more negative crap in the game.

[WorldChampions]

Good opening post and thanks for taking the time to compile the stats.

I'm not sure how sambermetrics like the ones used in baseball could be translated to football. Baseball is alot less complicated than football.

[horne]

said it before,say it again.....youre own football brain if you know youre game,will be more effective than the stats.
someone said earlier about midfield passes not having real stat value as oposed to final third passes....the midfield passes could be wearing down the opposition in terms of energy in the tank,psychologically,or even ticking down the clock....how does s'metrics account for things like this?

science says the human brain is better than the computer.....sabermetrics is trying its best to prove the computer is better than the human brain....or have i got this wrong?

[Sangria]

What if one tackle is a combination of excellent reading of the game and exquisite timing to prevent what would otherwise be a certain goal.  Then another is when the opponent takes a bad touch in a nothing area of the pitch and the player gets a lucky toe in to put it out for a throw.

Do they both count as the same?

They're all shown in a map of the tackles made. The statistical look may be significant over a longer run, while for a shorter period the map may be more informative. The tackling maps of Lucas and Mascherano, during the periods when each was the main tackler in the team, describes how they approach the role of DM differently. As always, it's not the data which is insignificant, but precisely which data is being looked at, and what conclusions are drawn. A blanket dismissal of such analyses is as wrong as drawing the wrong conclusions from the wrong set of data.

[Vulmea]

exactly mate, there`s no context, a player could make up yards of ground on an opponent clean through on goal, put in a tackle and lose the tackle but force the opponent wide or force him to take another couple of seconds to get the ball under control again and in the mean time the rest of the defence has got back.
is that `lost tackle` a negative stat that should go against the player or a tremendous piece of defensive cover?
it`s like assists as well, quite often the best pass in a move isnt the one just before the goal is scored.
the game is too fluid to turn into a science.

sounds a bit like thinking the worlds flat or man will never fly , I dont want it to be true but it is they will model football

the sabermetrics link is sound because the big advantage that early users gained was they looked at things that actually made a difference rather than basic stats e.g there used to be a set number of things people looked for in baseball - speed of foot, number of home runs, a player needed to look like a player - this was all old school thinking

what happened pre-sabermetrics was that Bill James discovered that actually all the old time scouts had it largely wrong, they were looking at what they thought mattered and not what actually mattered and there was massive resistance to it because nobody in the establishment wanted to listen despite the stats - James looked at different things than the traditional statto's - it was those different stats that mattered not the actual measurement -player performance has always been judged - its how you do it that makes the difference

not until one lad Billy Beane (the lad who recommended Comolli to fsg) actually listend and implemented what James was saying did anybody actually sit up and listen. Beane was able to put together  a good team on a very very limited budget and get them to massively over perform - the equivalent of getting Blackpool into the CL.

What we need to think about is not the number of tackles won or lost or the distance somebody runs but the key aspects of the game and what they are - what makes a good player, a good team, a good performance and how do you measure it.

A great winger playing with a poor forward may get 5 assists, the forward may get 5 goals so how do you know who is the better player - you'd need to count the number of chnaces missed - say the winger supplied 20 crosses and only 25% were converted - then you'd want to look at crosses as  key stat....... but a crap winger with  a great forward may
get 5 assists from 20 crosses but the chances he created were all of far poorer quality and only the ability of the forward made the goals....... so then it becomes a question of quality crosses - so how do you measure a quality cross......a cross that gives the forward a high percentage chance to score ? Eventually somebody clverere than me with the patience will figure the answer out.

the answer isn't trusting your own eyes though because they are completely biased.












just like the people on here who want to 'trust their- own eyes' -

[Sangria]

But statistics are only useful if interpreted the right way. Any set of statistics will have flaws in them. To draw meaningful conclusions from them requires an understanding of their limitations.

As an example take the tacking statistics? The key to understanding them would be to understand what is a successful tackle. For instance Kuyt attempted 13 tackles but won 4. How many of the unsuccessful tackles resulted in the opposition player having to play a hurried ball, that put the receiver in trouble? Stats will only measure what they are designed to do. They can certainly tell a lot about a players contribution, but they are still just a tool. In the wrong hands, they can be horrible. Ask the FA.

And Lucas's tackling technique means the success percentage is less significant a figure than for players who commit themselves more. As long as he doesn't get caught out of position, the number of challenges he attempts is the most significant figure, as his tackling isn't solely dependent on winning the ball, but can also contribute to the player in possession having less freedom to decide what to do with it. In contrast, for a player who relies on sliding in and winning the ball cleanly or not at all, the success percentage is all-important, as once he's made that attempt, he's out of the game.

Arsenal measure the amount of time a player spends on a ball, and weed out those who dwell too much. Chalkboards doesn't have this unfortunately, and no other free online analysts offer this either. But it's one stat I'd like to see.

[carling]

They're all shown in a map of the tackles made. The statistical look may be significant over a longer run, while for a shorter period the map may be more informative. The tackling maps of Lucas and Mascherano, during the periods when each was the main tackler in the team, describes how they approach the role of DM differently. As always, it's not the data which is insignificant, but precisely which data is being looked at, and what conclusions are drawn. A blanket dismissal of such analyses is as wrong as drawing the wrong conclusions from the wrong set of data.

Well the maps obviously help but for me there's still a long, long way to go before you can draw any meaningful conclusions.

[its cold in the stands]

sounds a bit like thinking the worlds flat or man will never fly , I dont want it to be true but it is they will model football

the sabermetrics link is sound because the big advantage that early users gained was they looked at things that actually made a difference rather than basic stats e.g there used to be a set number of things people looked for in baseball - speed of foot, number of home runs, a player needed to look like a player - this was all old school thinking

what happened pre-sabermetrics was that Bill James discovered that actually all the old time scouts had it largely wrong, they were looking at what they thought mattered and not what actually mattered and there was massive resistance to it because nobody in the establishment wanted to listen despite the stats - James looked at different things than the traditional statto's - it was those different stats that mattered not the actual measurement -player performance has always been judged - its how you do it that makes the difference

not until one lad Billy Beane (the lad who recommended Comolli to fsg) actually listend and implemented what James was saying did anybody actually sit up and listen. Beane was able to put together  a good team on a very very limited budget and get them to massively over perform - the equivalent of getting Blackpool into the CL.

What we need to think about is not the number of tackles won or lost or the distance somebody runs but the key aspects of the game and what they are - what makes a good player, a good team, a good performance and how do you measure it.

A great winger playing with a poor forward may get 5 assists, the forward may get 5 goals so how do you know who is the better player - you'd need to count the number of chnaces missed - say the winger supplied 20 crosses and only 25% were converted - then you'd want to look at crosses as  key stat....... but a crap winger with  a great forward may
get 5 assists from 20 crosses but the chances he created were all of far poorer quality and only the ability of the forward made the goals....... so then it becomes a question of quality crosses - so how do you measure a quality cross......a cross that gives the forward a high percentage chance to score ? Eventually somebody clverere than me with the patience will figure the answer out.

the answer isn't trusting your own eyes though because they are completely biased.












just like the people on here who want to 'trust their- own eyes' -

that wont work in football mate, people have tried it since the days of charlie hughes and the only thing that came out of it was long ball footy.
footy is way too fluid and has way too many variables to turn into a science.