MotoMkali Posted February 1, 2021 Share Posted February 1, 2021 (edited) 5 hours ago, GENTLEMAN said: Show your working please. LW = 38 games, 12 goals, 6 assists (excluding Liverpool game as it is an anomaly). Mean 0.47 CAM/CM= 15 games, 6 assists. Mean 0.4. Important to note half those assists came this season in 5 games as CAM (which is 0.6 - however that is much too small a sample size so included the games he played at centre mid last season as well). Basically he played in a much deeper role for 6/15 games and in a worse team for 10/15 games and is putting up basically the same numbers (I want him further forward than he played last season and 3 of his goals this season have come from a central position so he can evidently score from there but he just hasn't in such a small sample of games). For this season alone. Jack has 3 assists in 5 games (0.6) and 4 goals and 3 assists in 13 games which is 0.54 (again excluding the Liverpool game - I have done a mathematical proof to determine whether it is OK to exclude it but essentially if jack score or assists more than 3 goals in a game currently it is alright to exclude it - he has only scored or assisted 3 or more goals once in his career and that is against Liverpool). Again too small a sample size to be meaningful but I'd argue we played more difficult games in the 5 he played as CAM, one with a red card, one against united and one against Chelsea. Also in that period was the only time we have picked up points when conceding first. Edit: Correction he won a penalty against West Brom. Which was included as an assist. Which is incorrect on my part. So the correct values should be 0.33 in the 15 and 0.4 in the 5 games. Although does show the example of such a small sample size quite well. Edited February 1, 2021 by MotoMkali Link to comment Share on other sites More sharing options...
CosmicVillan Posted February 1, 2021 Share Posted February 1, 2021 31 minutes ago, useless said: I thought Smith would be an interesting choice when at Walsall, did good job to get them out of relegation trouble, had them playing good football, and a Villa fan. I hope that he can win us a major trophy in the coming seasons. Ironically they dubbed him The Ginger Mourihino....thankfully much more progressive Link to comment Share on other sites More sharing options...
Rightdm00 Posted February 1, 2021 Share Posted February 1, 2021 (edited) 23 minutes ago, MotoMkali said: LW = 38 games, 12 goals, 6 assists (excluding Liverpool game as it is an anomaly). Mean 0.47 CAM/CM= 15 games, 6 assists. Mean 0.4. Important to note half those assists came this season in 5 games as CAM (which is 0.6 - however that is much too small a sample size so included the games he played at centre mid last season as well). Basically he played in a much deeper role for 6/15 games and in a worse team for 10/15 games and is putting up basically the same numbers (I want him further forward than he played last season and 3 of his goals this season have come from a central position so he can evidently score from there but he just hasn't in such a small sample of games). For this season alone. Jack has 3 assists in 5 games (0.6) and 4 goals and 3 assists in 13 games which is 0.54 (again excluding the Liverpool game - I have done a mathematical proof to determine whether it is OK to exclude it but essentially if jack score or assists more than 3 goals in a game currently it is alright to exclude it - he has only scored or assisted 3 or more goals once in his career and that is against Liverpool). Again too small a sample size to be meaningful but I'd argue we played more difficult games in the 5 he played as CAM, one with a red card, one against united and one against Chelsea. Also in that period was the only time we have picked up points when conceding first. 23 minutes ago, MotoMkali said: Double quote How are you only getting 5 games in the middle? It's at least 7 Brighton West Ham Chelsea Crystal Palace Man United West Brom Burnley Wolves is the only game we played a 4-3-3 while Barkley was out so he was LW there. Brighton and Crystal Palace he spent at least half the time in the middle due to injury/suspension. Not sure how you can just toss them in the LW pile. Just check his game log at FBref.com. In the end look how much work you have to do just to equalize his play between the two positions. It's obvious he has performed at one position better than the other. Edited February 1, 2021 by Rightdm00 Double post 1 Link to comment Share on other sites More sharing options...
MotoMkali Posted February 1, 2021 Share Posted February 1, 2021 15 minutes ago, Rightdm00 said: How are you only getting 5 games in the middle? It's at least 7 Brighton West Ham Chelsea Crystal Palace Man United West Brom Burnley Wolves is the only game we played a 4-3-3 while Barkley was out so he was LW there. Brighton and Crystal Palace he spent at least half the time in the middle due to injury/suspension. Not sure how you can just toss them in the LW pile. Just check his game log at FBref.com. In the end look how much work you have to do just to equalize his play between the two positions. It's obvious he has performed at one position better than the other. 5 hours ago, MotoMkali said: Grealish played left wing up until the Burnley game. We played traore cam, Ramsey CAM and went 3 in the middle with hourihane. Grealish scored a goal in those 3 games. Got 3 assists in the 5 he played as CAM. Link to comment Share on other sites More sharing options...
nick76 Posted February 1, 2021 Share Posted February 1, 2021 42 minutes ago, MotoMkali said: LW = 38 games, 12 goals, 6 assists (excluding Liverpool game as it is an anomaly). Mean 0.47 CAM/CM= 15 games, 6 assists. Mean 0.4. Important to note half those assists came this season in 5 games as CAM (which is 0.6 - however that is much too small a sample size so included the games he played at centre mid last season as well). Basically he played in a much deeper role for 6/15 games and in a worse team for 10/15 games and is putting up basically the same numbers (I want him further forward than he played last season and 3 of his goals this season have come from a central position so he can evidently score from there but he just hasn't in such a small sample of games). For this season alone. Jack has 3 assists in 5 games (0.6) and 4 goals and 3 assists in 13 games which is 0.54 (again excluding the Liverpool game - I have done a mathematical proof to determine whether it is OK to exclude it but essentially if jack score or assists more than 3 goals in a game currently it is alright to exclude it - he has only scored or assisted 3 or more goals once in his career and that is against Liverpool). Again too small a sample size to be meaningful but I'd argue we played more difficult games in the 5 he played as CAM, one with a red card, one against united and one against Chelsea. Also in that period was the only time we have picked up points when conceding first. Somebody else has provided different figures to you showing him much better on LW...I’m also not sure some of your categorisation. Why are you excluding Liverpool game anyway? If you do that surely are skewing the data. Why would you not exclude where he doesn’t do anything because that would also be an anomaly. Link to comment Share on other sites More sharing options...
nick76 Posted February 1, 2021 Share Posted February 1, 2021 25 minutes ago, Rightdm00 said: How are you only getting 5 games in the middle? It's at least 7 Brighton West Ham Chelsea Crystal Palace Man United West Brom Burnley Wolves is the only game we played a 4-3-3 while Barkley was out so he was LW there. Brighton and Crystal Palace he spent at least half the time in the middle due to injury/suspension. Not sure how you can just toss them in the LW pile. Just check his game log at FBref.com. In the end look how much work you have to do just to equalize his play between the two positions. It's obvious he has performed at one position better than the other. Fitting his narrative! Link to comment Share on other sites More sharing options...
Kiwivillan Posted February 1, 2021 Share Posted February 1, 2021 2 minutes ago, nick76 said: Somebody else has provided different figures to you showing him much better on LW...I’m also not sure some of your categorisation. Why are you excluding Liverpool game anyway? If you do that surely are skewing the data. Why would you not exclude where he doesn’t do anything because that would also be an anomaly. He's thrown 7 games 3 assists at CM whereas I left out 4 games 2 goals 2 assists at LM. Full PL positional record LW 49 games 13G 10A CAM 19 games 0G 3A CM 7 games 0G 3A LM 4 games 2G 2A SS 1 game no returns RW 1 game no returns 1 Link to comment Share on other sites More sharing options...
useless Posted February 1, 2021 Share Posted February 1, 2021 (edited) I don't really wan to say this but Dean Smith couldn't care less where the uninformed herd on here thinks he should play. So coming out with this and that is of no consequence whatsoever, Smith will just play him where he thinks is best for team whether we like it or not. Edited February 1, 2021 by useless 1 Link to comment Share on other sites More sharing options...
Kiwivillan Posted February 1, 2021 Share Posted February 1, 2021 Just last 2 seasons LW/LM 38 games 14 goals 11 assists CAM/CM 16 games 0 goals 6 assists Link to comment Share on other sites More sharing options...
MotoMkali Posted February 1, 2021 Share Posted February 1, 2021 5 minutes ago, nick76 said: Somebody else has provided different figures to you showing him much better on LW...I’m also not sure some of your categorisation. Why are you excluding Liverpool game anyway? If you do that surely are skewing the data. Why would you not exclude where he doesn’t do anything because that would also be an anomaly. Nope, including such an anomalous game skews the results. It outside of 3 standard deviations from the mean which basically means you need to exclude the data to make any sort of analysis representative of the future. Link to comment Share on other sites More sharing options...
Kiwivillan Posted February 1, 2021 Share Posted February 1, 2021 1 minute ago, MotoMkali said: Nope, including such an anomalous game skews the results. It outside of 3 standard deviations from the mean which basically means you need to exclude the data to make any sort of analysis representative of the future. Why are you excluding data to make yours look better? Link to comment Share on other sites More sharing options...
MotoMkali Posted February 1, 2021 Share Posted February 1, 2021 6 minutes ago, Kiwivillan said: He's thrown 7 games 3 assists at CM whereas I left out 4 games 2 goals 2 assists at LM. Full PL positional record LW 49 games 13G 10A CAM 19 games 0G 3A CM 7 games 0G 3A LM 4 games 2G 2A SS 1 game no returns RW 1 game no returns Including 10 games at the minimum from before we got related for the first time when he was playing LW doesn't make any sense. I just used the last 2 years of premier league football. Link to comment Share on other sites More sharing options...
Rightdm00 Posted February 1, 2021 Share Posted February 1, 2021 13 minutes ago, MotoMkali said: Grealish played left wing up until the Burnley game. We played traore cam, Ramsey CAM and went 3 in the middle with hourihane. Grealish scored a goal in those 3 games. Got 3 assists in the 5 he played as CAM. This is why I split it up with and without Barkley. You say he played LW vs Brighton but his heatmap says differently. Same for Crystal Palace. Started middle but actually spent most of the game out wide, especially after the red to Mings. Those are just 2 examples. He drifts far more without a rigid #10 in the middle. It's not as easy to say he was LW one day and a CAM on another. Still, his best play this year and last has come with Barkley in the #10 role and Jack out wide left. We can debate why that is but to say he has been better from the CAM position is a weak, unsupported opinion. Link to comment Share on other sites More sharing options...
MotoMkali Posted February 1, 2021 Share Posted February 1, 2021 Just now, Kiwivillan said: Why are you excluding data to make yours look better? No. If something is outside of 3 standard deviations from the mean it is excluded. This is common knowledge. It is acknowledged as an outlier than not Included in any further analysis. 1 Link to comment Share on other sites More sharing options...
Kiwivillan Posted February 1, 2021 Share Posted February 1, 2021 1 minute ago, MotoMkali said: Including 10 games at the minimum from before we got related for the first time when he was playing LW doesn't make any sense. I just used the last 2 years of premier league football. This is my last update. There's 1 game at SS missing. No returns. Just last 2 seasons LW/LM 38 games 14 goals 11 assists CAM/CM 16 games 0 goals 6 assists Link to comment Share on other sites More sharing options...
MotoMkali Posted February 1, 2021 Share Posted February 1, 2021 (edited) 6 minutes ago, Kiwivillan said: Just last 2 seasons LW/LM 38 games 14 goals 11 assists CAM/CM 16 games 0 goals 6 assists He has only got 15 assists in that time period in the prem. Transfermarkt includes pens won and converted as assists so this data is wrong. Edited February 1, 2021 by MotoMkali Link to comment Share on other sites More sharing options...
MotoMkali Posted February 1, 2021 Share Posted February 1, 2021 Just now, Kiwivillan said: This is my last update. There's 1 game at SS missing. No returns. Just last 2 seasons LW/LM 38 games 14 goals 11 assists CAM/CM 16 games 0 goals 6 assists You exclude the Liverpool game as it doesn't not reflect traditional returns and cannot be counted on again in the future. Including it skews his performances to LW by a stupid amount. If he had 4 goals as a CAM that would also be excluded because it is well outside the acceptable deviation from the mean. Link to comment Share on other sites More sharing options...
MotoMkali Posted February 1, 2021 Share Posted February 1, 2021 4 minutes ago, Rightdm00 said: This is why I split it up with and without Barkley. You say he played LW vs Brighton but his heatmap says differently. Same for Crystal Palace. Started middle but actually spent most of the game out wide, especially after the red to Mings. Those are just 2 examples. He drifts far more without a rigid #10 in the middle. It's not as easy to say he was LW one day and a CAM on another. Still, his best play this year and last has come with Barkley in the #10 role and Jack out wide left. We can debate why that is but to say he has been better from the CAM position is a weak, unsupported opinion. Traore Played CAM that game. Grealish roamed more frequently true but as is evidenced by his defensive work he was playing wide left. Link to comment Share on other sites More sharing options...
Kiwivillan Posted February 1, 2021 Share Posted February 1, 2021 4 minutes ago, MotoMkali said: He has only got 15 assists in that time period in the prem. FPL has him at 18 but they class some assists that others don't. I'm checking variances Link to comment Share on other sites More sharing options...
Rightdm00 Posted February 1, 2021 Share Posted February 1, 2021 2 minutes ago, MotoMkali said: Traore Played CAM that game. Grealish roamed more frequently true but as is evidenced by his defensive work he was playing wide left. Traore didn't even finish the game. Got subbed with 20 minutes left to play including extra time. So no we define attacking players by where they defend on the pitch and not where they play while in attacking positions? Seems backwards and far to simplistic of a view. Link to comment Share on other sites More sharing options...
Recommended Posts