Liverpool's Transfers Revisited

Duncan Castles writes on Bleacher Report about the dismal transfer record of Liverpool FC under Brendan Rodgers' management. Part of the critic is the overly reliance on an analytic model in scouting that selected bad players or rejected good. For example, Castles points out that Liverpool rejected Sadio Mane on the ground that the model said he isn't good enough for Premier League. Mane now plays for Southampton FC and is proving the opposite. Given that Goalimpact predicted Sadio Mane to be good enough a full year and 4 Mio. Euros before Southampton signed him, let's check what Goalimpact tells us on the other transfers made. Here is the full roster. Rodgers' transfers are marked yellow.

Liverpool FC

PlayerGoalimpactPeak GIAgeLast National TeamNo. GamesNo. Minutes
Goalkeeper
Simon Mignolet105.7114.726.8Belgien24122449
Danny Ward74.1131.521.4Wales [U21]9837
Brad Jones85.686.332.712111212
Average88.5110.827.0
Defense
Lloyd Jones79.2112.319.2England [U19]11970
Mamadou Sakho109.1111.924.8Frankreich21518002
Dejan Lovren118.5119.825.4Kroatien22519680
Glen Johnson125.3135.630.3England40936172
Ryan McLaughlin89.9113.220.2Nordirland [U19]302541
Kolo Touré129.7163.233.7Elfenbeinküste50942986
Michael Williams85.2119.119.1Wales [U21]2138
Alberto Moreno108.4119.022.4Spanien [U21]836965
Manquillo96.7117.520.6Spanien [U21]403270
Jon Flanagan112.7125.321.9England [U20]473714
Martin Škrtel105.5113.530.0Slowakei34330267
José Enrique108.7115.328.8Spanien [U21]25821950
Average105.7122.124.7
Midfield
Steven Gerrard130.6170.434.5England75163042
Suso91.0109.021.0Spanien [U21]805177
Coutinho101.0111.322.5Brasilien [U20]1439443
Cameron Brannagan64.9104.118.64344
Joe Allen112.9115.824.8Wales23716932
Lucas Leiva108.7113.927.9Brasilien28221344
Adam Lallana114.4115.926.6England27521777
Emre Can109.2128.120.9Deutschland [U21]1249880
Jordan Henderson122.5126.124.4England26921103
Jordan Rossiter72.1115.117.77537
Raheem Sterling102.0126.620.0England1218475
Lazar Marković117.3137.120.8Serbien1268203
Average103.9122.823.3
Attack
Daniel Sturridge121.3123.025.3England21713254
Fabio Borini94.7100.423.7Italien [U21]1247798
Rickie Lambert103.5132.132.8England26621835
Mario Balotelli106.9110.824.3Italien25916740
Average106.6116.626.5

First thing to note is that about half of the team was signed by him. Apart from Borini, all seem to be good enough to play Premier League, also Coutinho and Manquillo will need some more development.

Some of the players are quite young and expected to improve significantly in the coming seasons. Given their potential, they were not bad signings as such. Even though they are maybe not good enough to qualify for Champions League this season, some are expected to be develop into that level (>120). Due to this fact, one can argue that the prices paid were too high. You'd expect bigger talent or further developed players for that money. Even in England.
Kolo Toure is a very good player, but well beyond his peak. His aging
is going to accelerate soon.

Especially in attack, the signings were not very convincing. We have no information on the factors that lead to the signings, but the two signings from Italy might point to a malfunctioning adjustment for league strength or opposition in general.

Mignolet is for sure a good keeper, but for 10 million Euros you can easily sign better ones. Brad Jones that played against ManU the other day, doesn't live up to Premier League standards.

Optimal XI

Ignoring all tactical considerations (no, this is not a recommendation), just picking by Goalimpact, leads to the following starting XI plus bench.

PlayerGoalimpactPeak GIAgeLast National TeamNo. GamesNo. Minutes
Daniel Sturridge121.3123.025.3England21713254
Steven Gerrard130.6170.434.5England75163042
Simon Mignolet105.7114.726.8Belgien24122449
Dejan Lovren118.5119.825.4Kroatien22519680
Joe Allen112.9115.824.8Wales23716932
Glen Johnson125.3135.630.3England40936172
Adam Lallana114.4115.926.6England27521777
Kolo Touré129.7163.233.7Elfenbeinküste50942986
Jordan Henderson122.5126.124.4England26921103
Jon Flanagan112.7125.321.9England [U20]473714
Lazar Marković117.3137.120.8Serbien1268203
Average119.2131.526.8
Bench
Mamadou Sakho109.1111.924.8Frankreich21518002
Lucas Leiva108.7113.927.9Brasilien28221344
Emre Can109.2128.120.9Deutschland [U21]1249880
José Enrique108.7115.328.8Spanien [U21]25821950
Average108.9117.325.6
15 Player Average
Average116.4127.726.5

This is good enough to end up sixth in Premier League, so currently Liverpool is underperforming. But even if they weren't, the team isn't good enough to reasonably expect to qualify for Champions League. Comparing this to the Champions League game against Basel, only Škrtel, Sterling, Coutinho and Lambert weren't first choice players according to Goalimpact. Still Basel (113.6) even fielded a slightly better team than Liverpool (113.1) in that game. This again underscores the fact that the team doesn't live up to Champions League ambitions.

Conclusion

If the ambition of Liverpool FC is to qualify for Champions League the players bought don't live up to that objective. Whilst only Borini would be outright rejected as a EPL transfer target by Goalimpact, most are more suitable for a team in the upper midfield of the league table than for a team with Champions League ambitions. In fact, of all new players only Toure and Sturridge are already as of today above the target Goalimpact of 120. Can and Markovic are expected to be well above the target level in future. In case of Markovic that is very soon, Can will take another year or so. Given the money spent on the transfer market, better far options would have been available.

Today's Chelsea FC against Arsenal's "The Invincibles"

Comparing teams across player generations is always a somewhat futile exercise, because they will obviously never play against each other to settle the question which team was better. Today a news paper article comparing appeared in my Twitter timeline comparing Chelsea FC to Arsenal's team in the 2003/2004 season, often called "The Invincibles".


Using a player by player comparison, the newspaper concluded that Arsenal's team at that time was marginally better than today's Chelsea. Just for fun, and knowing that this may raise many critics, we will repeat this comparison using Goalimpact.

Chelsea FC 2014

PlayerGoalimpactPeak GIAgeLast National TeamNo. GamesNo. Minutes
Ramires123.0127.627.7Brasilien35628160
Willian139.2140.126.3Brasilien29921598
Branislav Ivanović121.7136.130.8Serbien38934226
Filipe Luís124.4131.629.3Brasilien31927549
André Schürrle124.8129.224.1Deutschland27919896
Cesc Fàbregas194.6198.927.6Spanien53440327
Petr Čech186.8187.432.5Tschechien64760121
Eden Hazard135.6140.523.9Belgien33524617
Azpilicueta123.1124.825.3Spanien30226299
John Terry164.6199.534.0England68561485
Oscar124.9132.323.3Brasilien22515845
Average142.1149.827.7
Bench
Gary Cahill111.4118.128.9England37233905
Diego Costa114.7115.126.2Spanien25018922
Nemanja Matić121.0121.926.3Serbien16513182
John Mikel108.4112.827.6Nigeria36226165
Average113.9117.027.3
15 Player Average
Average134.5141.127.6

Chelsea has an excellent team on top European level. They will be a tough opponent for the other great teams such as Real Madrid, Barca and Bayern München.

Arsenal FC 2004 ("The Invincibles")

PlayerGoalimpactPeak GIAgeLast National TeamNo. GamesNo. Minutes
Gilberto Silva118.7123.827.8Brasilien1179704
Patrick Vieira149.1155.128.0Frankreich43738684
Jens Lehmann158.4159.134.7Deutschland55251449
Ashley Cole124.8131.423.5England18116173
Sol Campbell141.0148.729.8England44840084
Lauren133.1137.327.4Kamerun26923666
Thierry Henry143.1145.626.8Frankreich43031576
Nwankwo Kanu120.3125.927.9Nigeria31018040
Robert Pirés143.4156.930.7Frankreich47438687
Sylvain Wiltord129.1138.230.2Frankreich42530288
Dennis Bergkamp158.2203.735.2Niederlande63349708
Average138.1147.829.3
Bench
Freddie Ljungberg114.3117.927.3Schweden24618459
Kolo Touré117.3124.823.3Elfenbeinküste1017109
Reyes105.1124.320.8Spanien1107895
Edu112.2112.626.2Brasilien1287845
Average112.2119.924.4
15 Player Average
Average131.2140.428.0

According to Goalimpact, the last Arsenal team to win the Premier League was slightly weaker than today's Chelsea. These numbers are taken at the end of the season, so there are some footnotes. Dennis Bergkamp, at his peak, was the best player of all mentioned in this article. Even better than Cesc. However, Berkamp was aging rapidly during the season and therefore his Goalimpact at the beginning of the season, 167, was significantly higher than at the end.

Although Arsenal 2004 maybe wasn't better than the 2014 Chelsea, it was facing far less competition back then. The second team in the 2004 season was again Chelsea. And they had only a Goalimpact of 114. That is helping in becoming "invincible".


Manchester City vs Liverpool FC: Goalimpact of Lineups

Manchester City has the stronger starting XI and the home field advantage on their side. This makes Liverpool have only outsider chances on a win, though a draw is not very unlikely. All in all, Goalimpact is more optimistic on Cities chances than the betting markets

ResultGoalimpact based on starting XIBetfair
Manchester City59.0%50.5%
Draw25.5%25.8%
Liverpool FC15.5%23.7%

Given that City has also the better bench, the odds are probably not overly optimistic for City fans. And b.t.w. it is a sin that Lazar Markovic is on the bench. There must be a place for him in Liverpool's team somewhere.

Manchester City

PlayerGoalimpactPeak GIAgeLast National TeamNo. GamesNo. Minutes
Joe Hart142.9149.727.3England33631329
Yaya Touré128.6146.431.3Elfenbeinküste44337337
Samir Nasri126.8129.927.1Frankreich43431327
Stevan Jovetic104.1106.824.8Montenegro18113121
Fernando136.8139.727.0Brasilien [U20]23019743
Edin Džeko133.1139.128.4Bosnien-Herzegowina34826033
Martín Demichelis113.6146.033.6Argentinien45037639
Gaël Clichy154.1160.729.0Frankreich38132239
Pablo Zabaleta135.3142.429.5Argentinien42335558
David Silva135.3141.428.6Spanien42032612
Vincent Kompany143.6149.628.3Belgien42637224
Average132.2141.128.6
Bench
Kun Agüero145.7146.126.2Argentinien44733182
Jesús Navas134.2140.528.7Spanien43735231
Aleksandar Kolarov107.0113.428.8Serbien26219789
Bacary Sagna131.6151.431.5Frankreich42137582
Fernandinho142.6149.429.3Brasilien26822227
James Milner117.2123.428.6England52037414
Willy Caballero106.2107.632.825623418
Average126.4133.129.4


Liverpool FC

PlayerGoalimpactPeak GIAgeLast National TeamNo. GamesNo. Minutes
Daniel Sturridge120.9123.224.9England21412975
Coutinho103.4114.622.2Brasilien [U20]1278374
Jordan Henderson122.1126.024.2England24819327
Dejan Lovren120.8122.725.1Kroatien20718094
Raheem Sterling101.6129.419.7England986589
Martin Škrtel106.1113.229.7Slowakei32528593
Glen Johnson125.7133.129.9England39334758
Alberto Moreno111.6123.022.1Spanien695745
Simon Mignolet107.0119.026.4Belgien22220651
Joe Allen112.7116.124.4Wales22315958
Steven Gerrard136.8172.734.2England73361512
Average115.3126.625.7
Bench
Brad Jones86.187.832.312011119
Emre Can107.2127.620.6Deutschland [U21]1149313
Mamadou Sakho110.3113.524.5Frankreich21017528
Kolo Touré131.8162.933.3Elfenbeinküste50242363
Rickie Lambert106.3132.932.4England24921177
Lazar Markovic119.2140.420.4Serbien [U21]1157616
Manquillo98.0120.320.3Spanien [U20]231758
Average108.4126.526.3



Season Prediction

In lack of time, we just publish the pre-season estimates as plain table. For those leagues that started already, the few games so far have been ignored.

Premier League

Rank Team Goalimpact Peak GI Age
1 Manchester City 135.8 146.3 28.8
2 Chelsea FC 133.4 140.5 27.0
3 Arsenal FC 124.6 132.4 25.9
4 Manchester United 123.3 131.2 27.2
5 Tottenham Hotspur 119.6 125.2 26.3
6 Liverpool FC 119.1 129.8 26.1
7 Everton FC 112.7 123.3 27.4
8 Southampton FC 109.2 115.0 26.5
9 Swansea City 107.2 113.6 26.1
10 Newcastle United 106.4 113.5 25.1
11 Leicester City 106.0 115.1 25.9
12 Aston Villa 105.9 110.8 25.4
13 West Bromwich Albion 105.6 113.1 26.5
14 Hull City 104.6 110.7 25.9
15 Sunderland AFC 104.5 110.8 26.3
16 Stoke City 103.8 108.6 26.0
17 West Ham United 103.6 111.4 26.7
18 Queens Park Rangers 103.3 112.9 27.4
19 Burnley FC 102.9 108.4 25.4
20 Crystal Palace 101.7 108.7 27.6

Bundesliga

Rank Team Goalimpact Peak GI Age Bwin Rank TM.de ClubElo EuroClubIndex
1 Bayern München 149.9 157.9 26.8 1 1 1 1
2 Borussia Dortmund 127.8 130.9 26.4 2 2 2 2
3 FC Schalke 04 118.3 127.3 25.5 3 3 3 3
4 VfL Wolfsburg 117.2 124.0 25.3 5 4 5 5
5 VfB Stuttgart 112.6 120.1 26.1 8 8 15 15
6 Bayer Leverkusen 112.0 124.0 25.4 4 5 4 4
7 Hamburger SV 111.7 118.4 25.9 9 9 17 16
8 1899 Hoffenheim 111.7 118.5 25.6 7 6 9 14
9 Bor. Mönchengladbach 110.9 120.4 24.1 6 7 6 6
10 Hertha BSC 110.8 118.2 26.3 14 12 13 14
11 1. FC Köln 108.2 119.8 24.2 13 16 16 17
12 SC Paderborn 07 107.8 116.7 24.6 18 18 18 18
13 1. FSV Mainz 05 107.7 119.2 24.8 11 13 8 7
14 SC Freiburg 106.6 116.1 24.1 17 15 11 12
15 Hannover 96 106.5 113.6 25.7 10 11 10 8
16 FC Augsburg 105.6 113.7 24.6 12 17 7 9
17 Eintracht Frankfurt 103.3 112.0 25.3 16 10 12 10
18 Werder Bremen 100.1 112.0 24.3 15 14 14 13

Development of the Top Five Leagues

Back in February I analyzed the development of the top five leagues in Europe. Both the algorithm and the database has improved since then. Not to mention the football world as such that developed. Time for an update.


Back in 1990, Bundesliga was the strongest league and it stayed like this until throughout the nineties. Serie A was the runner up, followed by the Primera Division and the English Premier League. In the early 2000s they were all equal in Goalimpact terms.

But from this area of high competition on European level, the Premier League and the Bundesliga emerged as new champions while Primera Division and especially Serie A lost ground. Serie A dropped for a short time period even below the French Ligue 1 and recovered only slowly since then. The Primera Division followed the leading England and Germany, but couldn't catch up so far. The Premier League had some edge above Bundesliga between 2006 and 2010, but Bundesliga caught up in the meantime. Certainly, calls that the Premier League fall behind continental football look premature.

Season Predictions: Premier League, Primera Division, Serie A

Just before the season started, I predicted the final standings of the Bundesliga season 2013/2014. With a few match days behind us, I find it amazing how fast the teams got sorted in the table similar to what you'd expect. Obviously, this doesn't say much yet about the real final standings, but the predictions of all methods with the current standings are between 45% (transfermarkt.de market value) and 64% (Goalimpact). There seems to be predictive power in all of them. Worst prediction was the last year's final table with 37%. So some teams that you'd expect up in the table are already there as are the weak teams in the lower part of the table.

Goalimpact Bwin Rank ClubElo Euro Club
Index
tm.de Last Year current
Goalimpact 100% 78% 69% 83% 70% 50% 64%
Bwin Rank 100% 75% 87% 97% 75% 56%
ClubElo 100% 92% 75% 91% 51%
Euro Club Index 100% 84% 82% 62%
tm.de 100% 80% 45%
Last Year 100% 37%
current 100%

As stated in the previous post, one of the issues with using Goalimpact to predict league results is the assumption on the lineups. Goalimpact rates players and not teams and as I had no idea which players were going to be selected to play, I just took the average of all of them for the prediction. Now some match days past and I have another indication. I just take the minuted weighted average of the actual players on the first match days. This has the advantage that I can don't have to manually collect the team roosters for the teams and I can easily produce predictions for other leagues, too.

Premier League

No. Team Goalimpact Min Max
1 Manchester United 140,0 95,3 170,0
2 Chelsea FC 133,0 103,0 178,9
3 Manchester City 132,3 104,7 158,0
4 Arsenal FC 127,1 106,4 150,6
5 Liverpool FC 123,3 94,7 158,5
6 Tottenham Hotspur 121,7 94,8 151,6
7 Everton FC 120,6 93,2 148,5
8 Fulham FC 115,2 79,2 151,2
9 Southampton FC 108,6 88,8 130,5
10 Swansea City 106,6 91,2 125,5
11 Sunderland AFC 105,7 85,8 142,2
12 Newcastle United 105,5 84,2 127,1
13 West Ham United 104,5 87,7 131,1
14 Cardiff City 104,0 85,9 122,4
15 Stoke City 103,6 89,8 133,5
16 West Bromwich Albion 103,2 81,7 146,8
17 Norwich City 101,9 91,4 126,0
18 Hull City 101,6 87,3 137,0
19 Aston Villa 101,3 80,3 116,5
20 Crystal Palace 97,9 86,8 126,3

Sorry, Tottenham, again no Champions League. But actually I should repeat this next month as some of the transfers are not yet reflected in these lineup. I took all matches between the 1st of July and the 1st of September, but this for example doesn't show Özil's impact yet.

Primera Division

No. Team Goalimpact Min Max
1 FC Barcelona 158,5 107,7 190,8
2 Real Madrid 149,8 113,7 197,0
3 Atletico Madrid 122,7 93,5 151,9
4 Valencia CF 112,3 91,8 140,0
5 Sevilla FC 111,8 96,8 129,5
6 Malaga CF 110,0 94,4 142,0
7 Real Sociedad 106,3 94,3 121,8
8 Villarreal CF 105,9 89,8 118,4
9 Athletic Bilbao 103,3 92,1 114,6
10 Celta Vigo 103,2 84,6 124,2
11 Levante UD 103,2 87,1 121,1
12 Elche CF 100,5 82,9 118,7
13 Espanyol Barcelona 99,9 86,4 128,5
14 Real Betis 99,0 87,0 108,2
15 UD Almeria 98,7 87,3 106,5
16 Real Valladolid 98,7 82,8 110,5
17 Granada CF 98,6 82,4 109,8
18 Getafe CF 97,9 87,8 110,6
19 CA Osasuna 97,5 86,9 118,3
20 Rayo Vallecano 95,7 84,6 106,7

As you'd expect, the top clubs are far far away from the bottom clubs. Even Atletico is not really close to Real and Barca. The big unknown in this league is which teams will be relegated. Nearly half the table shouldn't be to sure of themselves.

Serie A

No. Team Goalimpact Min Max
1 Juventus 126,9 92,6 153,3
2 SSC Napoli 120,9 93,6 154,7
3 AC Milan 118,8 95,2 145,9
4 AS Roma 115,8 99,3 150,0
5 Inter 115,7 91,8 146,0
6 Lazio Roma 114,1 94,8 158,8
7 ACF Fiorentina 112,2 99,6 152,1
8 Udinese Calcio 108,1 99,2 123,9
9 Hellas Verona 101,4 88,7 126,2
10 Sampdoria 101,3 87,3 122,8
11 AS Livorno 100,9 92,8 110,3
12 Calcio Catania 100,6 83,7 121,1
13 Bologna FC 99,7 85,7 121,0
14 Genoa CFC 99,7 76,2 121,6
15 Parma FC 99,4 87,6 116,7
16 Atalanta Bergamo 99,3 84,8 113,4
17 Sassuolo Calcio 99,2 14,9 111,2
18 Torino FC 97,5 88,4 105,9
19 Chievo Verona 97,1 86,3 113,8
20 Cagliari Calcio 96,1 88,9 116,0

Similar setup to the Primera Division. Just without the two über-clubs. But again you have a long list of potential relegates. Half of the teams wouldn't be competitive enough to stay in the English Premier League.