Which MLS players take "good" shots? (And does Lovel Palmer take all the bad ones?)
/By Drew Olsen (@drewjolsen)
I have been a frequent and vocal critic of Lovel Palmer, and his all too frequent tendency to launch shots from 40 yards into the 40th row.
Ask around, and the journeyman is hardly beloved by the fans of any team he’s played for. When I unscientifically surveyed a few friends who are separately fans of his former teams, Palmer did not once get described using the words “quality”, “ability” or “skill.” Indeed, nearly everyone described him with a different variation of some four-letter word, but “frustration” is an apt synonym that sums them up. And most had a similar complaint; for every solid defensive play he makes, it always feels like he kills his own team’s attack because of his overconfidence in his long-range shot. A shot that has earned him four goals in his MLS career, and only one in the last three seasons. While my memory of his time with the Timbers may be unreliable, it seemed he was good for at least one cringe-worthy long distance shot per game. The below video encapsulates what I’m talking about:
But that is all anecdotal evidence, and this is an analytics website! To see if my criticism of Palmer has been justified, I decided come up with a way of determining who takes the worst (and best) shots in MLS. It’s really quite simple; I took all the players who have taken at least 40 total shots (there was lots of noise for players with fewer shots than that) since the first year of our dataset in 2011 and divided the sum of each player’s expected goals by the number of shots they’ve taken.
Our expected goals (xG) metrics assign a value to each shot based on where and how it was taken which tell us how much each shot is worth; that is, how many goals can be expected from that shot. Finding the average xG per shot for each player allows us to see how valuable those shots are. A high number means that the player took high quality shots; when he shot, it was usually close to the goal and it was usually with his foot (headers have a lower conversion rate). Conversely, a low number means the player usually shoots from far away, and the odds of his shots going in are slim.
| Player | Pos | Min | Shots | G | xG | G - xG | xG per shot | G per shot | G/s - xG/s |
|---|---|---|---|---|---|---|---|---|---|
| Charlie Davies | F | 3692 | 71 | 19 | 15.09 | 3.91 | 0.213 | 0.268 | 0.055 |
| Bradley Wright-Phillips | F | 4024 | 159 | 33 | 32.44 | 0.56 | 0.204 | 0.208 | 0.004 |
| Tesho Akindele | M | 2718 | 44 | 10 | 8.87 | 1.13 | 0.202 | 0.227 | 0.026 |
| Jaime Castrillon | M | 2952 | 62 | 9 | 11.8 | -2.8 | 0.19 | 0.145 | -0.045 |
| Brian Ching | F | 3104 | 71 | 10 | 12.65 | -2.65 | 0.178 | 0.141 | -0.037 |
| Luke Rodgers | F | 1624 | 45 | 9 | 7.97 | 1.03 | 0.177 | 0.2 | 0.023 |
| Peguy Luyindula | F | 2442 | 44 | 6 | 7.48 | -1.48 | 0.17 | 0.136 | -0.034 |
| Patrice Bernier | M | 7102 | 68 | 13 | 11.55 | 1.45 | 0.17 | 0.191 | 0.021 |
| Alejandro Moreno | F | 3741 | 48 | 7 | 8.1 | -1.1 | 0.169 | 0.146 | -0.023 |
| Nat Borchers | D | 12466 | 58 | 9 | 9.78 | -0.78 | 0.169 | 0.155 | -0.013 |
| Luke Moore | F | 2726 | 63 | 7 | 10.51 | -3.51 | 0.167 | 0.111 | -0.056 |
| Obafemi Martins | F | 5424 | 140 | 31 | 23.21 | 7.79 | 0.166 | 0.221 | 0.056 |
| Fanendo Adi | F | 2574 | 84 | 13 | 13.86 | -0.86 | 0.165 | 0.155 | -0.01 |
| Alvaro Saborio | F | 8381 | 278 | 51 | 45.38 | 5.62 | 0.163 | 0.183 | 0.02 |
| Jermain Defoe | F | 1726 | 58 | 11 | 9.42 | 1.58 | 0.162 | 0.19 | 0.027 |
| Danny Koevermans | F | 2145 | 70 | 16 | 11.11 | 4.89 | 0.159 | 0.229 | 0.07 |
| Antoine Hoppenot | F | 2062 | 41 | 7 | 6.44 | 0.56 | 0.157 | 0.171 | 0.014 |
| Landon Donovan | M | 9368 | 211 | 41 | 33.08 | 7.92 | 0.157 | 0.194 | 0.038 |
| Vicente Sanchez | F | 2134 | 48 | 7 | 7.51 | -0.51 | 0.156 | 0.146 | -0.011 |
| Omar Bravo | F | 2368 | 60 | 9 | 9.38 | -0.38 | 0.156 | 0.15 | -0.006 |
| Erick Torres | F | 4090 | 140 | 22 | 21.76 | 0.24 | 0.155 | 0.157 | 0.002 |
| Robbie Keane | F | 8231 | 296 | 55 | 45.95 | 9.05 | 0.155 | 0.186 | 0.031 |
| Matt Hedges | D | 9282 | 44 | 10 | 6.81 | 3.19 | 0.155 | 0.227 | 0.073 |
| Jerry Bengtson | F | 2381 | 46 | 4 | 7.04 | -3.04 | 0.153 | 0.087 | -0.066 |
| Juan Agudelo | M | 5706 | 146 | 22 | 22.3 | -0.3 | 0.153 | 0.151 | -0.002 |
Above are the 25 players with the highest xG per shot in MLS since 2011. The results bear out mostly as you would expect; the players who score lots of goals in MLS tend to do so because they take high quality shots; they can get into places where the shots they take are more likely to go in than those of the average player. I’ve also included the number of actual goals per shot to give some perspective; for the most part, the xG/shots and G/shots numbers are pretty close together, but that’s where the elite have separated from the average. The top strikers have actually scored more goals than the average player would have expected to. In the cases of players like Davies, Martins, Donovan, Keane and Hedges (?!), they've shown they’re especially skilled finishers.
But it wasn’t the logical desire to see which MLS players take the highest quality shots that led me down this rabbit hole. Nay, it was a desire to confirm my previously held belief that Lovel Palmer was sent from the future to wreak havoc on the attack of every team he plays for by shooting from distance at every opportunity.
And boy were my beliefs confirmed.
| Player | Pos | Min | Shots | G | xG | G - xG | xG per shot | G per shot | G/s - xG/s |
|---|---|---|---|---|---|---|---|---|---|
| Lovel Palmer | D | 9037 | 113 | 2 | 4.3 | -2.3 | 0.038 | 0.018 | -0.02 |
| Adam Moffat | M | 7182 | 99 | 7 | 4.35 | 2.65 | 0.044 | 0.071 | 0.027 |
| Roger Espinoza | M | 5482 | 65 | 2 | 3.1 | -1.1 | 0.048 | 0.031 | -0.017 |
| Vitor Gomes Pereira Junior | M | 12747 | 211 | 13 | 10.31 | 2.69 | 0.049 | 0.062 | 0.013 |
| David Beckham | M | 4454 | 69 | 9 | 3.46 | 5.54 | 0.05 | 0.13 | 0.08 |
| Jamie Smith | M | 2062 | 42 | 6 | 2.11 | 3.89 | 0.05 | 0.143 | 0.093 |
| Sam Cronin | M | 11578 | 108 | 6 | 5.43 | 0.57 | 0.05 | 0.056 | 0.005 |
| Kyle Beckerman | M | 11987 | 176 | 14 | 9.06 | 4.94 | 0.051 | 0.08 | 0.028 |
| Paulo Nagamura | M | 6282 | 94 | 4 | 4.84 | -0.84 | 0.051 | 0.043 | -0.009 |
| Tristan Bowen | F | 1677 | 44 | 2 | 2.27 | -0.27 | 0.052 | 0.045 | -0.006 |
| Gershon Koffie | M | 9687 | 123 | 9 | 6.46 | 2.54 | 0.053 | 0.073 | 0.021 |
| Kalif Alhassan | M | 5430 | 125 | 5 | 6.64 | -1.64 | 0.053 | 0.04 | -0.013 |
| Julian de Guzman | M | 3485 | 48 | 3 | 2.58 | 0.42 | 0.054 | 0.063 | 0.009 |
| Daniel Hernandez | M | 4975 | 41 | 2 | 2.23 | -0.23 | 0.054 | 0.049 | -0.006 |
| Russell Teibert | M | 5351 | 50 | 2 | 2.76 | -0.76 | 0.055 | 0.04 | -0.015 |
| Chris Birchall | M | 3306 | 44 | 2 | 2.44 | -0.44 | 0.055 | 0.045 | -0.01 |
| Alexandre De Lima | M | 4472 | 91 | 3 | 5.2 | -2.2 | 0.057 | 0.033 | -0.024 |
| Bobby Convey | M | 5634 | 44 | 3 | 2.52 | 0.48 | 0.057 | 0.068 | 0.011 |
| Osvaldo Alonso | M | 12447 | 164 | 4 | 9.42 | -5.42 | 0.057 | 0.024 | -0.033 |
| Abdoulie Mansally | D | 4011 | 46 | 0 | 2.65 | -2.65 | 0.058 | 0 | -0.058 |
| Khari Stephenson | M | 5793 | 169 | 10 | 9.78 | 0.22 | 0.058 | 0.059 | 0.001 |
| Eric Avila | M | 8333 | 100 | 7 | 5.86 | 1.14 | 0.059 | 0.07 | 0.011 |
| Stefan Ishizaki | M | 3145 | 52 | 5 | 3.07 | 1.93 | 0.059 | 0.096 | 0.037 |
| Ben Speas | M | 1929 | 43 | 2 | 2.56 | -0.56 | 0.06 | 0.047 | -0.013 |
| Marco Pappa | M | 7118 | 244 | 22 | 14.58 | 7.42 | 0.06 | 0.09 | 0.03 |
Above are the 25 players with the lowest average xG per shot since 2011. These are some names that will feel familiar; they’re players not afraid to shoot from distance. Players like Moffat, Juhnino, Beckham*, and Beckerman have been feared for their long-range abilities for years, and you’ll see they’ve overperformed their xG/shots. They’ve all scored more goals than the average player would expect to given where they take their shots from; that makes sense because our eyes tell us that these players are better shooters from distance than the average player.
But Palmer is the true standout. His xG of only 0.038 goals per shot is the lowest in the dataset. That is 0.046 goals fewer per shot than the league median of 0.084. Put another way, Palmer’s average shot is worth less than half of the median MLS shot. And he’s actually under-performed that conversion rate, meaning he’s shot worse than the average shooter from those distances.
I’ve been piling on Palmer, but there is reason for hope, Fire fans! It may be no coincidence that of all the teams Palmer has played for, Chicago fans seem to have the most apathy about his shooting. That’s because he rarely does it anymore. After averaging a high of one shot every 52 minutes in the 2011 season, he seems to be shooting less and less. Last season he took a shot only every 94 minutes, and this season he has only three shots in almost 600 minutes played (196 min. per shot). Couple that with the fact that he's a perfectly serviceable MLS defender, and there is reason to be optimistic.
Is this a foolproof way of determining the best or worst scorers in MLS? Of course not. Is it a handy tool for seeing what types of shots players take, and how good they are at converting them? You bet. And for that, I would like to thank Mr. Lovel Palmer.
*Beckham’s G/shot–xG/shot of 0.080 is second best of all players. He must have been good at free kicks from distance or something.