10 Aug 2009

Gerrard and Torres - The ultimate analysis of their partnership. Is it really that effective?

Steven Gerrard and Fernando Torres will be absolutely vital to Liverpool's FC's hopes of winning the Premiership this season. We all know how well the two players work together, but just how effective is the Gerrard/Torres partnership?

I have analysed every LEAGUE game Gerrard and Torres have started together since the Spaniard joined the club in 2007. I think you'll agree that the results are quite intriguing.

Season 2008-2009


Gerrard + Torres starting together (t=Torres g=Gerrard a=Assist)

Aug 16, 2008 - Sunderland - W, 0 - 1 (t1)
Aug 23, 2008 - Middlesbrough - W, 2 - 1 (g1)
Sep 20, 2008 - Stoke City - D, 0 - 0
Sep 27, 2008 - Everton - W, 0 - 2 (t2)

Oct 5, 2008 - Manchester City - W, 2 - 3 (t2 g1a)
Jan 19, 2009 - Everton - D, 1 - 1 (g1)
Jan 28, 2009 - Wigan Athletic - D, 1 - 1
Feb 1, 2009 - Chelsea - W, 2 - 0 (t2)

Mar 14, 2009 - Manchester United - W, 1 - 4 (t1 – g1)
Mar 22, 2009 - Aston Villa - W, 5 - 0 (g3)
Apr 4, 2009 - Fulham - W, 0 - 1
May 9, 2009 - West Ham United - W, 0 - 3 (g2 – t1a)
May 17, 2009 - West Brom - W, 0-2 (g1 - g1a)
May 24, 2009 - Spurs - W, 3-1 (t1 - g1a)

TOTALS

P14 W11 D3 F30 A7

Torres: 10 goals + 2 assist
Gerrard: 9 goals + 3 assists
Total: 19 goals + 5 assists

ANALYSIS

* G+T: 24 goals scored/created in 14 games started together
* G+T: Average of 1.7 goals scored/created per game
* Unbeaten in all 14 games together
* 80% of games won with G + T starting together


SEASON 2007 - 2008

Gerrard + Torres starting together

Aug 11, 2007 - Aston Villa W, 1 - 2 (g1)
Aug 19, 2007 - Chelsea D, 1 -1 (t1)
Sep 29, 2007 - Wigan Athletic W, 0 - 1
Oct 7, 2007 - Tottenham Hotspur D, 2 - 2 (t1)

Oct 28, 2007 - Arsenal D, 1 - 1 (g1)
Nov 24, 2007 - Newcastle United W, 0 - 3 (g1)
Dec 2, 2007 - Bolton Wanderers W, 4 - 0 (g1 t1)
Dec 8, 2007 - Reading L, 3 - 1 (g1)

Dec 16, 2007 - Manchester United L, 0 - 1
Dec 22, 2007 - Portsmouth W, 4 - 1 (t2)
Dec 26, 2007 - Derby County W, 1 - 2 (g1 t1)
Dec 30, 2007 - Manchester City D, 0 - 0

Jan 2, 2008 - Wigan Athletic D, 1 - 1
Jan 21, 2008 - Aston Villa D, 2 - 2
Jan 12, 2008 - Middlesbrough D, 1 - 1 (t1)
Jan 30, 2008 - West Ham United L, 1 - 0

Feb 2, 2008 - Sunderland W, 3 - 0 (t1 – g1)
Feb 23, 2008 - Middlesbrough W, 3 - 2 (t3)
Mar 2, 2008 - Bolton Wanderers W, 1 - 3
Mar 5, 2008- West Ham United W, 4 - 0 (g1 t3)

Mar 8, 2008 - Newcastle United W, 3 - 0 (t1 g1)
Mar 15, 2008 - Reading W, 2 - 1 (t1)
Mar 23, 2008 - Manchester United L, 3 - 0
Mar 30, 2008 - Everton W, 1 - 0 (t1)

Apr 13, 2008 - Blackburn Rovers W, 3 - 1 (g1 – t1)
May 4, 2008 - Manchester City W, 1 - 0 (t1)
May 11, 2008 - Tottenham Hotspur W, 0 - 2 (t1)

TOTALS

P27 W16 D7 L4 F50 A21

Torres: 20 goals + 2 assist
Gerrard: 10 goals + 10 assists
Total: 30 goals + 12 assists

ANALYSIS

* G+T: 41 goals scored/created in 27 games started together
* G+T: Average of 1.4 goals scored/created per game
* Unbeaten in 85% of games started together
* 60% of games won with G + T starting together
* 13 league games drawn in 2007-2008: T + G started 7 of those
* Failed to win 40% of games with T + G starting together

GRAND TOTALS (Last 2 seasons combined)

P41 W27 D10 L4 F80 A28

Torres: 30 goals + 4 assists
Gerrard: 19 goals + 13 assists
Total: 49 goals + 17 assists

ANALYSIS

* G+T: 66 goals scored/created in 41 games started together
* G+T: Average of 1.7 goals scored/created per game
* Unbeaten in 93% of games started together
* 66% of games won with G + T starting together

* 24 league games drawn from 2007-2009: T + G started 10 of those
* Failed to win 36% of games with T + G starting together

Clearly, the Gerrard/Torres partnership is vital to Liverpool’s success in the league! However, I do not think it is completely accurate to say that if they had played more games together last season then Liverpool would have won the league.

last season alone, Gerrard and Torres played in the league draws against Stoke City, Everton and Wigan; in 2007-08, they played together in 7 of Liverpool’s 13 league draws.

If the partnership is so potent, why does the team still draw so many games with both players in the starting line up?

Based on last season’s draw stats, it’s conceivable that Gerrard and Torres wouldn’t have made that much difference in the other 8 draws during the campaign.

Of course, it works both ways – they could have had a massive difference, but I’m just making the point that having them both in the starting line up more regularly doesn’t necessarily mean Liverpool draw fewer games.

One thing is for sure: with Gerrard and Torres in the same team you are practically guaranteed goals and a very high probability of not losing the game.

So - to answer the question posed in the title: is the Gerrard-Torres partnership really that effective? The answer is a resounding yes! And then some.

There is still arguably room for improvement though, which can only be a good thing for Liverpool.

*Source for statistics: LFC History + Soccernet

* How effective are Liverpool without Steven Gerrard? Find out here: Debunking Liverpool FC myths - no 2: Liverpool struggle without Steven Gerrard



29 comments:

  1. Hi,
    Your stats only go so far, torres always plays up front gerrard does not.the stats should comare the possition he has played in.

    ReplyDelete
  2. In the first year Torres and Gerrard were getting to know each others play, yet still performed brilliantly.

    Last year, Liverpool scored 36 points out of 14 games they played together. Thats 98 points over a season.

    Gerrard and Torres is a fearsome partnership. we just need a few others to chip in with match winning goals. Bennynoun and Kuyt both scored goals last season, but we need our midfield to score match winners, not 3rd and 4th goals in an easy win.

    ReplyDelete
  3. Great post! Interesting to see exactly how potent this partnership is, especially since they pretty much 'clicked' right away and many strike partnerships take two-three seasons to start reading each other's play. I think it's fair to say that we need these two fit and firing for a whole season to wrest the title off the scum.

    Keep up the good work

    ;)

    ReplyDelete
  4. ;) Nice article Jamie, good to see you've looked at things constructively and objectively..I think what it shows is weakness in other areas of the team. We all know there is nothing wrong with our front line, it's the wide areas that give most concern, too many times our wingers prefer to either cut inside or go backwards.
    Carl

    ReplyDelete
  5. Can you please explain this quote? "<span style="">However, I do not think it is completely accurate to say that if they had played more games together last season then Liverpool would have won the league."???? </span>

    And this quote? "<span style="">If the partnership is so potent, why does the team still draw so many games with both players in the starting line up?" Errr.... there were only 2 goalless draws out of 10. Issue with defence, perhaps?</span>

    ReplyDelete
  6. Cvt - what have goalless draws got to do with anything?  There have been 10 draws overall when G+T have started together. Whether they're goalless or not is irrelevent as I'm not suggesting that scoring goals is an issue.

    Re the following quote: <span><span>"I do not think it is completely accurate to say that if they had played more games together last season then Liverpool would have won the league."</span></span>

    I've hear many fans say that if G+T had played together more often last season we would've won the league.  I'm suggesting that this is not a foregone conclusion.

    ReplyDelete
  7. Thanks Carl.  I agree about the wide areas.  Why is it we can never solve this problem?  The wide areas were a concern theroughout the Houllier era and 5 years into Rafa's reign we still haven't managed to find a top class winger.  I just can't accept that in 10 years, there hasn't been one decent winger out there within Liverpool's budget. Houllier and Rafa just have not prioritised the wide areas, and I think the creativity of the team has suffered as a result.

    ReplyDelete
  8. I thought this was quite interesting, so thought I would look at the numbers in a bit more depth (but not much). I have taken your figures for games where Gerrard and Torres, but just looked at the match outcome.

    If you look at results for games where G&T start and compare it to results where they didn't, you get something like this for the past two years (in the league)...

    With G&T        W27 D10 L4
    Without          W19 D14 L2

    Throwing a basic statistical test like Chi-Squared at those figures (which isn't ideal with such low numbers in the lose columns, but what can you do), it can tell you whether there is any statistical significance behind the difference. Quite simply, computer says "no".

    Obviously, this is a highly crude approach, and to do things properly you would really need to be analysing team performance for every minute every player was on the pitch. Then you could start establishing which players appear to have the most positive outcome on matches.

    However, at the basic level at which it works, you are right to raise some question marks. The numbers provide no evidence that G&T starting together means more games won.

    Silly what you can show with statistics, eh?

    ReplyDelete
  9. Great Article!!!

    Where ever our soft spots may have been last season be it defenders, wingers or lack of reserve strikers a simple stat keeps going round my head and that is the only 2 games we lost last season and we scored OWN GOALS in both!!!
    Had Carragher and Alonso not helped Spurs and Boro to victories then any frailties within the side would have pailled into insignificance! Even just a draw and a win out of those 2 would have done the trick!

    Anyway, interesting stats that made good reading!

    YNWA - The Fox.

    ReplyDelete
  10. Not exactly, JK. I think the signing of Johnson, and the earlier acquisitions of Dossena, who still has to prove himself, and Degen, shows that Rafa likes to get his width from the fullbacks, which is why he has Masch in defensive mid (so that the fullbacks have to do a little less defending and some attacking). Riera hugs the touchline, and Rafa plays the 3 attacking mids, which is perfect for players like Benayoun, Babel, and Kuyt, to a certain extent. I think we're fine with Kuyt and Benayoun on the right this season, and the fact that JOhnson is there makes our right side quite formidable. As for the left side, Riera's pretty decent, and Babel needs to up his game. Insua's up and coming, and who can forget Aurelio, he's a good fullback.

    If you get width from the fullbacks, you don't really need wingers. Look at Brazil in 2002, where they played a 5-2-3 formation, with Carlos and Cafu as fullbacks, 2 controlling mids, and 2 out of Denilson, Rivaldo, Ronaldinho along with Ronaldo who was a fixture as striker. This season, hopefully Johnson and Aurelio/Insua/Dossena can get us the width. 

    ReplyDelete
  11. Jamie: re. Draws - you question why there were so many draws with G&T playing? They are strikers and their main responsibility is to score goals. I just pointed out that in 8 of the 10 draws, they fulfilled their 'duty', so maybe the goals conceded rather than the lack of goals scored should be considered?

    Re winning the league. Given your use of stats, I am surprised that you have not extrapolated the results. 80% win and 20% draw would equal 98 points. 

    ReplyDelete
  12. Guest who posted this - :-D  
    <span style=" color: #3a3a3a; line-height: 16px;"> </span>
    <span style=" color: #3a3a3a; line-height: 16px;">If you look at results for games where G&T start and compare it to results where they didn't, you get something like this for the past two years (in the league)...  
     
    With G&T        W27 D10 L4  
    Without          W19 D14 L2  </span>

    There is actually a huge difference in that the 8 extra games won equates to an additional 24 points  on the table, now that is a lot of points by anyones standard, you certainly won't win the Prem without them  ;)

    ReplyDelete
  13. You can reply directly to the peson who posted by clicking the reply button under their post.  This comment system allows threading ;)

    ReplyDelete
  14. I was just applying some real statistical analysis for amusement value, and it showed that team performance with and without G&T starting is not statistically significant (i.e. the differences are not considered large enough to prove that the variations are not simply down to chance).

    It was an act of curiosity rather than dismissing the value of G&T. However, I expect Rafa with his Prozone software (or whatever it is he uses) can get much more accurate analysis of the team performance with and without them.

    The 8 extra wins = 24 points is only accurate up to a point. For a start, the games with G&T total 41, the games without total 35, so it would never be as simple as simply subtracting one figure from the other If you wanted to do something like that, you'd have to look at the win ratio, i.e. 27/41 vs 19/35. Using that 19/35 would give an expected figure of 22.25 wins in 41 matches.

    If you took these performance figures of the team with and without G&T and applied their stats to a season of 38 games, you would end up with...

    With G&T     W25   D9   L4   PTS 84
    Without       W21   D15   L2   PTS 78

    Of course, some of this reflects the fact that we had fewer points in the 07/08 season anyway.

    Again, all good number fun. Proves little, but quite amusing to investigate.

    ReplyDelete
  15. Apparently the 'Utimate Analysis'... ;)

    ReplyDelete
  16. Insua started 11 games last season, we won 10 of them, drew the other. If we analyse that, we would conclude we end the season with 35 wins and 3 draws, so... Insua is our most important player for sure! :-D

    ReplyDelete
  17. hahahaha thnx Jamie, I wondered what that button was for lol :-[

    Carl

    ReplyDelete
  18. yes it is good fun. on a side note it also just shows you can make numbers tell any story you want them too... ;)

    ReplyDelete
  19. That's not strictly true though, is it?  complex statistical analyses aside, there are certain general things that cannot be denied, such as:

    * Liverpool are unbeaten in 93% of games in which G+T start together
    * Goals (either scored or created) are practically guaranteed when G+T play together.

    No matter how you slice the data, these facts (up to this point) cannot ne refuted.  The same goes for other statistics I've used in other articles.  Yes, you can introduce dozens of different variables but at the base level, certain things cannot be denied.

    ReplyDelete
  20. Jaimie, u seem like u still havent grasped the idea of statistical significance.
    It is absolutely refutable all these things that u just said. Simply because your numbers are too small and it is not statistically significant.
    Look up to the earlier posts regarding the one that took Insua's stats.

    <span style="cursor: pointer;">Guest</span><img src="http://js-kit.com/images/icon10-external-url.png"/> <span>Insua started 11 games last season, we won 10 of them, drew the other. If we analyse that, we would conclude we end the season with 35 wins and 3 draws, so... Insua is our most important player for sure! </span>

    Going by your analysis method... this guy is correct too! So is insua our most important player?
    We of course know that it's not true. But your method would prove it!

    So where is the catch?
    The catch is in that this is not statistically significant.
    I went through this in my other post which you conveniently deleted.
    And you just conveniently googled for some website which said something abt statistical signifcance and paraphrased something without actually understanding it and challenged me.
    Please read up on your statistics first can? :)

    Anyway...
    For something to be proven by statistics, the numbers need to be large enough else most of the time, it cant be proven to be statistically significant. Its not to say though that large numbers are definitely needed. But it certainly helps.
    Also, the difference in the 2 results compared is quite important. A big difference generally is more likely to be statistically significant. BUT again, a small difference may still be statistically significant, but generally its harder to prove, meaning you need bigger sample sizes.
    That's why that site that you paraphrased says what it says.

    ReplyDelete
  21. CK - your analysis of 'statistical significance' has no bearing on the stuff that I post.  This is a football site, not the Office for National Statistics.  You may not think w, y and z has 'statistical significance' and that's your right.  I have a different view, and people will interpret the statistics in their own way, not how you tell them to interpret them.

    ReplyDelete
  22. And once again - it is a FACT that with Torres and Gerrard starting together, Liverpool are unbeaten in 93% of games. yes, there are lots of different variables you can apply but on a general level, that fact remains correct.

    If it is NOT correct, please show HOW and WHY it is not correct ;)

    ReplyDelete
  23. JK, Liverpool only lost 2 games last season anyway--so, it's fairly obvious that the 'Pool don't lose many whether G&T play or not.

    I must say that you contributed to CK's anger by deleting his post. You say that you only delete insulting posts or those that don't deal with the issues. However, it seems that CK's crime was neither of these, but merely to provide some accurate sounding information about statistics.

    Why did you delete?

    (No, I'm not CK, *DONT_KNOW* but see you as a guy who always tries to sound reasonable and scientific, yet gets easily shaken when other people's opinions and ideas seem stronger than your own. In the last analysis, it's not much of a defense to effectively say: "this is my site and I'm entitled to believe whatever I like."

    ReplyDelete
  24. What nonsense.  I'm not 'shaken' at all when people challenge my opinions.  As i said earlier in the thread, I deleted CK's post because of its unpleasant patronising tone.  I don't delete posts that disagree with me just because they disagree - 90% of comments on the site disagree with me (!) but they're still all there.

    If people can't debate without trying to  belittle the other person then that's not my problem.  It's simple: state your case in the right way and we don't have a problem.

    ReplyDelete
  25. In the past two seasons, Liverpool lost 4 (league) games from 41 with G&T starting, so unbeaten in just over 90% of games (i.e. about 90.2). However, in the same period, they lost only two games from 35 where G&T didn't start, so in fact unbeaten in slightly over 94% of matches (so perhaps we should consider them a liability!!!). However, as I highlighted previously, they win more with G&T, so would end up with more points.

    Liverpool don't get beaten much, regardless of G&T, so in reality, it's very difficult to prove that the two of them starting influences anything. It may well be that looking at matches in which neither played at all would show a significant difference. However, that would be a whole further exercise.

    ReplyDelete
  26. Actually, it's not really all that accurate to suggest that G&T starting "guarantees goals". The team score goals anyway. Comparing the with and without figures, you only get a slight increase in the goals per game where G&T started. The boost amounts to less than 7%, which doesn't sound like a huge amount.

    Under normal circumstances, I would work out if the increase was statistically significant. I know you say you don't want to go down that avenue, but the trouble is that if you don't, I can "prove" that Insua guarantees more goals that G&T...

             Games    F     A    GF per    GA per
    s-2009    38    77    27    2.03    0.71
    s-2008    38    67    28    1.76    0.74
    Total       76   144   55    1.89    0.72
                       
    G&T         41    80    28    1.95    0.68
    Without    35    64    27    1.83    0.77

    Insua       13    27      6      2.08    0.46
    Without    63    117    49    1.86    0.78

    Insua is the magic ingredient that guarantees goals!!! Look at that - more than 2 goals a game when he plays. Forget G&T, Insua is the hero of LFC.

    Plainly ridiculous, I realise, but if you simply look at the numbers, rather than anaylyse them, that's the result you get.

    ReplyDelete
  27. Kuyt and Benayoun did score match winners.
    Yossi saved us from a draw at Fulham, and Kuyt salvages points for us all the time

    ReplyDelete
  28. Great work Statto!
    Looks like we need a new headline.
    Ultimate analysis of the liverpool team's most effective player - INSUA!
    Since its widely thought that G&T are the most impt influence on the team, then if Insua guarantees more goals 2.08 per game compared to G&T 1.95, INSUA is the BEST!!!!
    You agree JK?

    Is this a FACT? :)

    ReplyDelete
  29. Great work Statto! 
    Looks like we need a new headline. 
    Ultimate analysis of the liverpool team's most effective player - INSUA! 
    Since its widely thought that G&T are the most impt influence on the team, then if Insua guarantees more goals 2.08 per game compared to G&T 1.95, INSUA is the BEST!!!! 
    You agree JK? 
     
    Is this a FACT? :)

    ReplyDelete