Football betting site Betway

Stats

Have a look at this superb piece of analysis of expected goals versus actual goals for all Championship sides over the past 5 seasons:

https://experimental361.com/2020/02/09/championship-trends-9-feb-2020/

It's interesting to see how our expected goals performance over the years has more often than not remained fairly constant when our actual points tally has varied significantly - especially in seasons 2015/16 and 2017/18. This suggests that the expected goals league might be a more useful long-term predictor of success than the actual table. This is being borne out this season, where the fantastic start we had to the season massively exceeded our expected performance, but now we're modestly under-achieving, as was evident on Saturday. Overall, we're probably are where we deserve to be.

No surprise that in the last two thirds of last season, under Potter, that we consistently under-achieved. I can't help but think that it we had the excellent set-piece defence that we have now last season that we would have come close to winning automatic promotion. C'est la vie.

This post is also illuminating: https://experimental361.com/2020/02/09/scatter-graphics-championship-9-feb-2020/

What's remarkable about Swansea's position on these graphs is how unremarkable our position is - basically a middle of the pack team. Last season, especially when it came to possession and pass completion stats we were very much the footballing outlier. How times have changed.

Comments

  • Not sure I'd put much weight on expected goals.

    According to expected goals Man City should be 5 points ahead of Liverpool.
  • No one stat is going to give you a full picture, discernment is the order or the day. Teams who over-perform stats wise, such as Liverpool, are particularly sharp where it matters - at both ends. Teams that under-perform (e.g. Stoke - a scarcely believeable 3rd in the expected goals table) clearly have issues at both ends. Expected goals tables, as with any stats, can help to cut through the noise and focus attention on problem areas. 

    Stats are wonderful, and the more stats the better as far as I'm concerned. Just ask Matthew Benham (Brentford second in expected goals league, 5th in actual league, two points off 2nd place).
  • Success hasn't happened overnight for Brentford while the following from 2015 is a good read.
    https://www.theguardian.com/football/2015/oct/15/brentford-owner-matthew-benham-hates-moneyball

    Two excerpts from the article from Benham.

    “I used to be very sceptical about using individual stats for players,” he added. “I am now slightly more open to them. There are pluses and minuses. The good things is that sometimes a player might be very good at tackles and interceptions but you don’t really realise it because some players who make a lot of tackles by getting in quickly don’t tend to stick in the mind as much as a Stuart Pearce-type player. On the other hand, is a player making a lot of tackles because he is badly positioned in the first place? So these can have a lot of use but they always need context.”

    “And there’s one journalist who interviewed me for a book about four years ago, and in the book I talk at length about the unreliability of any maths model – and how I also believed in scouting with the eyes.
    Wyn
  • barry said:
    Not sure I'd put much weight on expected goals.

    According to expected goals Man City should be 5 points ahead of Liverpool.
    When you see that you can either say stats are rubbish and move on, or examine the reasons why Liverpool overperform their xG compared to City. Better finishers? Goalkeeper making more saves? Man on the back post on corners making vital goal line clearances? VAR conspiracy?

    According to xG stats Man City are better at making quality chances than Liverpool. This is interesting, not incorrect because the league table says they're in second place.
    Wyn
  • Success hasn't happened overnight for Brentford while the following from 2015 is a good read.
    https://www.theguardian.com/football/2015/oct/15/brentford-owner-matthew-benham-hates-moneyball

    Two excerpts from the article from Benham.

    “I used to be very sceptical about using individual stats for players,” he added. “I am now slightly more open to them. There are pluses and minuses. The good things is that sometimes a player might be very good at tackles and interceptions but you don’t really realise it because some players who make a lot of tackles by getting in quickly don’t tend to stick in the mind as much as a Stuart Pearce-type player. On the other hand, is a player making a lot of tackles because he is badly positioned in the first place? So these can have a lot of use but they always need context.”

    “And there’s one journalist who interviewed me for a book about four years ago, and in the book I talk at length about the unreliability of any maths model – and how I also believed in scouting with the eyes.
    Good post. Grimes is a good case in point. It's amazing how many Swansea fans seem to think he's not very good defensively when he's up their with the best in the league on tackle count. Perhaps because he's not physically imposing he doesn't seem like the sort of player who would be good defensively, but the stats don't lie. Equally Gallagher and Byers put themselves about, but I'm not sure either are particularly good tacklers, nor read danger well. Leon was the perfect exemplar of a player who made up for his slight physique by his excellent reading of the game, and consequent ability to make interceptions. 

    Stats are very useful for sure, but the smartest people in football will understand their limitations, and that's where the edge is. Again referring to Leon, his stats, apart from pass completion, weren't that impressive, but anyone with any sort of discernment will have known how fundamental he was to our success. 
    Jackareme
  • EFLstats for this week make for interesting reading: https://twitter.com/EFLStats?ref_src=twsrc^google|twcamp^serp|twgr^author

    4th on expected goals, 3rd on completed passes (behind Derby - it was a quality game to be fair), 1st on passing accuracy (85%), equal 1st on shots on target, equal 3rd on total shots. 

    I think there are some grounds for cautious optimism regarding the performance on Saturday, but the games against Derby and Stoke show that we're in danger of developing a vulnerability in front of the back four. Address that and we should progress until the end of the season. 
  • Wyn
    good to see that a guy like Benham still thinks there is a place for the ‘eyes’ when spotting players.
    Wyn
  • Wyn
    good to see that a guy like Benham still thinks there is a place for the ‘eyes’ when spotting players.
    Agreed. I imagine it's a bit like an interview process. Stats like psychometric testing are great for narrowing down the field of suitable candidates, but the final judgement has to be a subjective human one, as with an interview panel. Having said, that in a decade or so it wouldn't surprise me if AI were to make subjective judgements redundant. How clubs find an edge over their competitors at that point is a moot point.
  • Wyn,
    I wonder how much Frankie Burrows's fabled contact book would be worth in today's game.
  • Wyn,
    I wonder how much Frankie Burrows's fabled contact book would be worth in today's game.
    Probably not much. The players in it would be far too old by now ;)
    Wyn
  • Wyn,
    I wonder how much Frankie Burrows's fabled contact book would be worth in today's game.
    Good question. I guess a lot less valuable than it would have been back in the day. 
Sign In or Register to comment.